diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9590 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9978030025631637,
+  "eval_steps": 500,
+  "global_step": 1364,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014646649578908824,
+      "grad_norm": 11.437485320992703,
+      "learning_rate": 1.4598540145985402e-07,
+      "loss": 1.3349,
+      "step": 1
+    },
+    {
+      "epoch": 0.0029293299157817647,
+      "grad_norm": 11.178701580362661,
+      "learning_rate": 2.9197080291970804e-07,
+      "loss": 1.338,
+      "step": 2
+    },
+    {
+      "epoch": 0.004393994873672647,
+      "grad_norm": 11.35884244830137,
+      "learning_rate": 4.379562043795621e-07,
+      "loss": 1.3216,
+      "step": 3
+    },
+    {
+      "epoch": 0.005858659831563529,
+      "grad_norm": 10.417628767122284,
+      "learning_rate": 5.839416058394161e-07,
+      "loss": 1.3553,
+      "step": 4
+    },
+    {
+      "epoch": 0.007323324789454412,
+      "grad_norm": 11.211576033283285,
+      "learning_rate": 7.299270072992701e-07,
+      "loss": 1.3584,
+      "step": 5
+    },
+    {
+      "epoch": 0.008787989747345295,
+      "grad_norm": 10.861271761101971,
+      "learning_rate": 8.759124087591242e-07,
+      "loss": 1.3293,
+      "step": 6
+    },
+    {
+      "epoch": 0.010252654705236177,
+      "grad_norm": 9.89480118782974,
+      "learning_rate": 1.0218978102189781e-06,
+      "loss": 1.3379,
+      "step": 7
+    },
+    {
+      "epoch": 0.011717319663127059,
+      "grad_norm": 9.16199354912927,
+      "learning_rate": 1.1678832116788322e-06,
+      "loss": 1.304,
+      "step": 8
+    },
+    {
+      "epoch": 0.013181984621017943,
+      "grad_norm": 7.8824339227925,
+      "learning_rate": 1.3138686131386864e-06,
+      "loss": 1.2671,
+      "step": 9
+    },
+    {
+      "epoch": 0.014646649578908825,
+      "grad_norm": 8.125898486086502,
+      "learning_rate": 1.4598540145985402e-06,
+      "loss": 1.2187,
+      "step": 10
+    },
+    {
+      "epoch": 0.016111314536799707,
+      "grad_norm": 10.83717739533965,
+      "learning_rate": 1.6058394160583942e-06,
+      "loss": 1.2298,
+      "step": 11
+    },
+    {
+      "epoch": 0.01757597949469059,
+      "grad_norm": 20.701868436023478,
+      "learning_rate": 1.7518248175182485e-06,
+      "loss": 1.1312,
+      "step": 12
+    },
+    {
+      "epoch": 0.01904064445258147,
+      "grad_norm": 20.601367187466792,
+      "learning_rate": 1.8978102189781023e-06,
+      "loss": 1.0933,
+      "step": 13
+    },
+    {
+      "epoch": 0.020505309410472353,
+      "grad_norm": 23.59333380509113,
+      "learning_rate": 2.0437956204379563e-06,
+      "loss": 1.0839,
+      "step": 14
+    },
+    {
+      "epoch": 0.021969974368363236,
+      "grad_norm": 4.122191454831178,
+      "learning_rate": 2.1897810218978103e-06,
+      "loss": 1.0648,
+      "step": 15
+    },
+    {
+      "epoch": 0.023434639326254118,
+      "grad_norm": 9.470990626696512,
+      "learning_rate": 2.3357664233576643e-06,
+      "loss": 1.0762,
+      "step": 16
+    },
+    {
+      "epoch": 0.024899304284145003,
+      "grad_norm": 5.597117534177985,
+      "learning_rate": 2.4817518248175183e-06,
+      "loss": 0.9529,
+      "step": 17
+    },
+    {
+      "epoch": 0.026363969242035885,
+      "grad_norm": 2.581862089288807,
+      "learning_rate": 2.627737226277373e-06,
+      "loss": 0.8969,
+      "step": 18
+    },
+    {
+      "epoch": 0.027828634199926768,
+      "grad_norm": 5.742679255827132,
+      "learning_rate": 2.7737226277372264e-06,
+      "loss": 0.899,
+      "step": 19
+    },
+    {
+      "epoch": 0.02929329915781765,
+      "grad_norm": 4.2881369714318796,
+      "learning_rate": 2.9197080291970804e-06,
+      "loss": 0.8479,
+      "step": 20
+    },
+    {
+      "epoch": 0.030757964115708532,
+      "grad_norm": 8.043168797174141,
+      "learning_rate": 3.065693430656935e-06,
+      "loss": 0.8139,
+      "step": 21
+    },
+    {
+      "epoch": 0.032222629073599414,
+      "grad_norm": 8.109008525384814,
+      "learning_rate": 3.2116788321167884e-06,
+      "loss": 0.832,
+      "step": 22
+    },
+    {
+      "epoch": 0.033687294031490296,
+      "grad_norm": 3.7973925320214708,
+      "learning_rate": 3.3576642335766425e-06,
+      "loss": 0.793,
+      "step": 23
+    },
+    {
+      "epoch": 0.03515195898938118,
+      "grad_norm": 3.0028103120588856,
+      "learning_rate": 3.503649635036497e-06,
+      "loss": 0.7458,
+      "step": 24
+    },
+    {
+      "epoch": 0.03661662394727206,
+      "grad_norm": 2.9364448534773038,
+      "learning_rate": 3.6496350364963505e-06,
+      "loss": 0.7061,
+      "step": 25
+    },
+    {
+      "epoch": 0.03808128890516294,
+      "grad_norm": 0.9806974298270025,
+      "learning_rate": 3.7956204379562045e-06,
+      "loss": 0.7023,
+      "step": 26
+    },
+    {
+      "epoch": 0.039545953863053825,
+      "grad_norm": 1.064375799560964,
+      "learning_rate": 3.9416058394160585e-06,
+      "loss": 0.6876,
+      "step": 27
+    },
+    {
+      "epoch": 0.04101061882094471,
+      "grad_norm": 0.7759784247428909,
+      "learning_rate": 4.0875912408759126e-06,
+      "loss": 0.6773,
+      "step": 28
+    },
+    {
+      "epoch": 0.04247528377883559,
+      "grad_norm": 0.6478179875434958,
+      "learning_rate": 4.233576642335767e-06,
+      "loss": 0.6657,
+      "step": 29
+    },
+    {
+      "epoch": 0.04393994873672647,
+      "grad_norm": 1.3039225652221051,
+      "learning_rate": 4.379562043795621e-06,
+      "loss": 0.6358,
+      "step": 30
+    },
+    {
+      "epoch": 0.04540461369461735,
+      "grad_norm": 0.6416252297147181,
+      "learning_rate": 4.525547445255475e-06,
+      "loss": 0.5966,
+      "step": 31
+    },
+    {
+      "epoch": 0.046869278652508235,
+      "grad_norm": 0.5640382481221872,
+      "learning_rate": 4.671532846715329e-06,
+      "loss": 0.5848,
+      "step": 32
+    },
+    {
+      "epoch": 0.048333943610399124,
+      "grad_norm": 0.581441699738184,
+      "learning_rate": 4.8175182481751835e-06,
+      "loss": 0.5963,
+      "step": 33
+    },
+    {
+      "epoch": 0.049798608568290006,
+      "grad_norm": 0.5919465771312318,
+      "learning_rate": 4.963503649635037e-06,
+      "loss": 0.5921,
+      "step": 34
+    },
+    {
+      "epoch": 0.05126327352618089,
+      "grad_norm": 0.6061068767118333,
+      "learning_rate": 5.1094890510948916e-06,
+      "loss": 0.5766,
+      "step": 35
+    },
+    {
+      "epoch": 0.05272793848407177,
+      "grad_norm": 0.5757173997249402,
+      "learning_rate": 5.255474452554746e-06,
+      "loss": 0.5234,
+      "step": 36
+    },
+    {
+      "epoch": 0.05419260344196265,
+      "grad_norm": 0.5148147313648633,
+      "learning_rate": 5.401459854014599e-06,
+      "loss": 0.5428,
+      "step": 37
+    },
+    {
+      "epoch": 0.055657268399853535,
+      "grad_norm": 0.5378426318034144,
+      "learning_rate": 5.547445255474453e-06,
+      "loss": 0.5421,
+      "step": 38
+    },
+    {
+      "epoch": 0.05712193335774442,
+      "grad_norm": 0.5096082401040436,
+      "learning_rate": 5.693430656934307e-06,
+      "loss": 0.5447,
+      "step": 39
+    },
+    {
+      "epoch": 0.0585865983156353,
+      "grad_norm": 0.48460009463464304,
+      "learning_rate": 5.839416058394161e-06,
+      "loss": 0.5396,
+      "step": 40
+    },
+    {
+      "epoch": 0.06005126327352618,
+      "grad_norm": 0.49057265665249916,
+      "learning_rate": 5.985401459854016e-06,
+      "loss": 0.5393,
+      "step": 41
+    },
+    {
+      "epoch": 0.061515928231417064,
+      "grad_norm": 0.5380540290567165,
+      "learning_rate": 6.13138686131387e-06,
+      "loss": 0.544,
+      "step": 42
+    },
+    {
+      "epoch": 0.06298059318930795,
+      "grad_norm": 0.4836947048784,
+      "learning_rate": 6.277372262773723e-06,
+      "loss": 0.5291,
+      "step": 43
+    },
+    {
+      "epoch": 0.06444525814719883,
+      "grad_norm": 0.4975286083516887,
+      "learning_rate": 6.423357664233577e-06,
+      "loss": 0.5006,
+      "step": 44
+    },
+    {
+      "epoch": 0.06590992310508971,
+      "grad_norm": 0.4801869620568488,
+      "learning_rate": 6.569343065693431e-06,
+      "loss": 0.5177,
+      "step": 45
+    },
+    {
+      "epoch": 0.06737458806298059,
+      "grad_norm": 0.3973492869700473,
+      "learning_rate": 6.715328467153285e-06,
+      "loss": 0.4788,
+      "step": 46
+    },
+    {
+      "epoch": 0.06883925302087147,
+      "grad_norm": 0.363722763875419,
+      "learning_rate": 6.86131386861314e-06,
+      "loss": 0.5129,
+      "step": 47
+    },
+    {
+      "epoch": 0.07030391797876236,
+      "grad_norm": 0.381635129408091,
+      "learning_rate": 7.007299270072994e-06,
+      "loss": 0.4867,
+      "step": 48
+    },
+    {
+      "epoch": 0.07176858293665324,
+      "grad_norm": 0.3591035550780032,
+      "learning_rate": 7.153284671532848e-06,
+      "loss": 0.4963,
+      "step": 49
+    },
+    {
+      "epoch": 0.07323324789454412,
+      "grad_norm": 0.4186636869053589,
+      "learning_rate": 7.299270072992701e-06,
+      "loss": 0.4986,
+      "step": 50
+    },
+    {
+      "epoch": 0.074697912852435,
+      "grad_norm": 0.41597006530339986,
+      "learning_rate": 7.445255474452555e-06,
+      "loss": 0.4765,
+      "step": 51
+    },
+    {
+      "epoch": 0.07616257781032588,
+      "grad_norm": 0.6427961759723377,
+      "learning_rate": 7.591240875912409e-06,
+      "loss": 0.4694,
+      "step": 52
+    },
+    {
+      "epoch": 0.07762724276821677,
+      "grad_norm": 0.4874587482608775,
+      "learning_rate": 7.737226277372264e-06,
+      "loss": 0.4634,
+      "step": 53
+    },
+    {
+      "epoch": 0.07909190772610765,
+      "grad_norm": 0.3334959408832527,
+      "learning_rate": 7.883211678832117e-06,
+      "loss": 0.4636,
+      "step": 54
+    },
+    {
+      "epoch": 0.08055657268399853,
+      "grad_norm": 0.381153104725579,
+      "learning_rate": 8.029197080291972e-06,
+      "loss": 0.459,
+      "step": 55
+    },
+    {
+      "epoch": 0.08202123764188941,
+      "grad_norm": 0.3340733041790047,
+      "learning_rate": 8.175182481751825e-06,
+      "loss": 0.4851,
+      "step": 56
+    },
+    {
+      "epoch": 0.0834859025997803,
+      "grad_norm": 0.3517427711639141,
+      "learning_rate": 8.32116788321168e-06,
+      "loss": 0.4236,
+      "step": 57
+    },
+    {
+      "epoch": 0.08495056755767118,
+      "grad_norm": 0.3108043883771063,
+      "learning_rate": 8.467153284671533e-06,
+      "loss": 0.4336,
+      "step": 58
+    },
+    {
+      "epoch": 0.08641523251556206,
+      "grad_norm": 0.3334733312721485,
+      "learning_rate": 8.613138686131386e-06,
+      "loss": 0.4365,
+      "step": 59
+    },
+    {
+      "epoch": 0.08787989747345294,
+      "grad_norm": 0.352845389781815,
+      "learning_rate": 8.759124087591241e-06,
+      "loss": 0.4671,
+      "step": 60
+    },
+    {
+      "epoch": 0.08934456243134382,
+      "grad_norm": 0.3423521983704048,
+      "learning_rate": 8.905109489051096e-06,
+      "loss": 0.4405,
+      "step": 61
+    },
+    {
+      "epoch": 0.0908092273892347,
+      "grad_norm": 0.32680792151395816,
+      "learning_rate": 9.05109489051095e-06,
+      "loss": 0.4244,
+      "step": 62
+    },
+    {
+      "epoch": 0.09227389234712559,
+      "grad_norm": 0.35932668840566334,
+      "learning_rate": 9.197080291970804e-06,
+      "loss": 0.4316,
+      "step": 63
+    },
+    {
+      "epoch": 0.09373855730501647,
+      "grad_norm": 0.30409497564005455,
+      "learning_rate": 9.343065693430657e-06,
+      "loss": 0.4404,
+      "step": 64
+    },
+    {
+      "epoch": 0.09520322226290737,
+      "grad_norm": 0.34431142299821804,
+      "learning_rate": 9.48905109489051e-06,
+      "loss": 0.4538,
+      "step": 65
+    },
+    {
+      "epoch": 0.09666788722079825,
+      "grad_norm": 0.33318967367640623,
+      "learning_rate": 9.635036496350367e-06,
+      "loss": 0.4246,
+      "step": 66
+    },
+    {
+      "epoch": 0.09813255217868913,
+      "grad_norm": 0.32969822743310584,
+      "learning_rate": 9.78102189781022e-06,
+      "loss": 0.4185,
+      "step": 67
+    },
+    {
+      "epoch": 0.09959721713658001,
+      "grad_norm": 0.3203953785691397,
+      "learning_rate": 9.927007299270073e-06,
+      "loss": 0.3801,
+      "step": 68
+    },
+    {
+      "epoch": 0.1010618820944709,
+      "grad_norm": 0.32486117695619926,
+      "learning_rate": 1.0072992700729928e-05,
+      "loss": 0.4072,
+      "step": 69
+    },
+    {
+      "epoch": 0.10252654705236178,
+      "grad_norm": 0.35349454533419244,
+      "learning_rate": 1.0218978102189783e-05,
+      "loss": 0.4105,
+      "step": 70
+    },
+    {
+      "epoch": 0.10399121201025266,
+      "grad_norm": 0.32856345829840156,
+      "learning_rate": 1.0364963503649636e-05,
+      "loss": 0.3967,
+      "step": 71
+    },
+    {
+      "epoch": 0.10545587696814354,
+      "grad_norm": 0.29148929214591157,
+      "learning_rate": 1.0510948905109491e-05,
+      "loss": 0.3824,
+      "step": 72
+    },
+    {
+      "epoch": 0.10692054192603442,
+      "grad_norm": 0.35704567954771227,
+      "learning_rate": 1.0656934306569344e-05,
+      "loss": 0.3953,
+      "step": 73
+    },
+    {
+      "epoch": 0.1083852068839253,
+      "grad_norm": 0.31980320150253144,
+      "learning_rate": 1.0802919708029198e-05,
+      "loss": 0.3896,
+      "step": 74
+    },
+    {
+      "epoch": 0.10984987184181619,
+      "grad_norm": 0.34589383334602003,
+      "learning_rate": 1.0948905109489052e-05,
+      "loss": 0.3995,
+      "step": 75
+    },
+    {
+      "epoch": 0.11131453679970707,
+      "grad_norm": 0.3272293236162762,
+      "learning_rate": 1.1094890510948906e-05,
+      "loss": 0.3901,
+      "step": 76
+    },
+    {
+      "epoch": 0.11277920175759795,
+      "grad_norm": 0.3579751175217903,
+      "learning_rate": 1.124087591240876e-05,
+      "loss": 0.3921,
+      "step": 77
+    },
+    {
+      "epoch": 0.11424386671548883,
+      "grad_norm": 0.3239414165797523,
+      "learning_rate": 1.1386861313868614e-05,
+      "loss": 0.3989,
+      "step": 78
+    },
+    {
+      "epoch": 0.11570853167337972,
+      "grad_norm": 0.354305255693745,
+      "learning_rate": 1.1532846715328467e-05,
+      "loss": 0.3983,
+      "step": 79
+    },
+    {
+      "epoch": 0.1171731966312706,
+      "grad_norm": 0.3440015850313245,
+      "learning_rate": 1.1678832116788322e-05,
+      "loss": 0.3994,
+      "step": 80
+    },
+    {
+      "epoch": 0.11863786158916148,
+      "grad_norm": 0.36417507763366563,
+      "learning_rate": 1.1824817518248176e-05,
+      "loss": 0.368,
+      "step": 81
+    },
+    {
+      "epoch": 0.12010252654705236,
+      "grad_norm": 0.3568328603972373,
+      "learning_rate": 1.1970802919708031e-05,
+      "loss": 0.3694,
+      "step": 82
+    },
+    {
+      "epoch": 0.12156719150494324,
+      "grad_norm": 0.33678321301690173,
+      "learning_rate": 1.2116788321167885e-05,
+      "loss": 0.3688,
+      "step": 83
+    },
+    {
+      "epoch": 0.12303185646283413,
+      "grad_norm": 0.36497790299914806,
+      "learning_rate": 1.226277372262774e-05,
+      "loss": 0.3998,
+      "step": 84
+    },
+    {
+      "epoch": 0.12449652142072501,
+      "grad_norm": 0.32377149566716346,
+      "learning_rate": 1.2408759124087593e-05,
+      "loss": 0.3586,
+      "step": 85
+    },
+    {
+      "epoch": 0.1259611863786159,
+      "grad_norm": 0.3633002134475903,
+      "learning_rate": 1.2554744525547446e-05,
+      "loss": 0.3749,
+      "step": 86
+    },
+    {
+      "epoch": 0.12742585133650677,
+      "grad_norm": 0.3719628223211349,
+      "learning_rate": 1.27007299270073e-05,
+      "loss": 0.3811,
+      "step": 87
+    },
+    {
+      "epoch": 0.12889051629439766,
+      "grad_norm": 0.344394951638129,
+      "learning_rate": 1.2846715328467154e-05,
+      "loss": 0.377,
+      "step": 88
+    },
+    {
+      "epoch": 0.13035518125228854,
+      "grad_norm": 0.369488047348081,
+      "learning_rate": 1.2992700729927009e-05,
+      "loss": 0.3663,
+      "step": 89
+    },
+    {
+      "epoch": 0.13181984621017942,
+      "grad_norm": 0.3123562848371242,
+      "learning_rate": 1.3138686131386862e-05,
+      "loss": 0.3561,
+      "step": 90
+    },
+    {
+      "epoch": 0.1332845111680703,
+      "grad_norm": 0.3743377566954952,
+      "learning_rate": 1.3284671532846715e-05,
+      "loss": 0.3488,
+      "step": 91
+    },
+    {
+      "epoch": 0.13474917612596118,
+      "grad_norm": 0.3633285055316245,
+      "learning_rate": 1.343065693430657e-05,
+      "loss": 0.3566,
+      "step": 92
+    },
+    {
+      "epoch": 0.13621384108385207,
+      "grad_norm": 0.31139891879793924,
+      "learning_rate": 1.3576642335766423e-05,
+      "loss": 0.3464,
+      "step": 93
+    },
+    {
+      "epoch": 0.13767850604174295,
+      "grad_norm": 0.3898557118255509,
+      "learning_rate": 1.372262773722628e-05,
+      "loss": 0.3559,
+      "step": 94
+    },
+    {
+      "epoch": 0.13914317099963383,
+      "grad_norm": 0.3497149782228266,
+      "learning_rate": 1.3868613138686133e-05,
+      "loss": 0.3426,
+      "step": 95
+    },
+    {
+      "epoch": 0.1406078359575247,
+      "grad_norm": 0.3646043020546925,
+      "learning_rate": 1.4014598540145988e-05,
+      "loss": 0.3641,
+      "step": 96
+    },
+    {
+      "epoch": 0.1420725009154156,
+      "grad_norm": 0.33631572980864205,
+      "learning_rate": 1.416058394160584e-05,
+      "loss": 0.3325,
+      "step": 97
+    },
+    {
+      "epoch": 0.14353716587330648,
+      "grad_norm": 0.37775436608527085,
+      "learning_rate": 1.4306569343065696e-05,
+      "loss": 0.3615,
+      "step": 98
+    },
+    {
+      "epoch": 0.14500183083119736,
+      "grad_norm": 0.3575269321533561,
+      "learning_rate": 1.4452554744525549e-05,
+      "loss": 0.3404,
+      "step": 99
+    },
+    {
+      "epoch": 0.14646649578908824,
+      "grad_norm": 0.34707886852032405,
+      "learning_rate": 1.4598540145985402e-05,
+      "loss": 0.33,
+      "step": 100
+    },
+    {
+      "epoch": 0.14793116074697912,
+      "grad_norm": 0.36563790351605774,
+      "learning_rate": 1.4744525547445257e-05,
+      "loss": 0.3173,
+      "step": 101
+    },
+    {
+      "epoch": 0.14939582570487,
+      "grad_norm": 0.3723711813860172,
+      "learning_rate": 1.489051094890511e-05,
+      "loss": 0.3216,
+      "step": 102
+    },
+    {
+      "epoch": 0.1508604906627609,
+      "grad_norm": 0.3847201221658523,
+      "learning_rate": 1.5036496350364965e-05,
+      "loss": 0.3344,
+      "step": 103
+    },
+    {
+      "epoch": 0.15232515562065177,
+      "grad_norm": 0.35970372266040174,
+      "learning_rate": 1.5182481751824818e-05,
+      "loss": 0.3289,
+      "step": 104
+    },
+    {
+      "epoch": 0.15378982057854265,
+      "grad_norm": 0.3915138161843807,
+      "learning_rate": 1.5328467153284673e-05,
+      "loss": 0.3361,
+      "step": 105
+    },
+    {
+      "epoch": 0.15525448553643353,
+      "grad_norm": 0.3332723731808281,
+      "learning_rate": 1.5474452554744528e-05,
+      "loss": 0.3147,
+      "step": 106
+    },
+    {
+      "epoch": 0.15671915049432442,
+      "grad_norm": 0.3613636145975632,
+      "learning_rate": 1.5620437956204383e-05,
+      "loss": 0.3432,
+      "step": 107
+    },
+    {
+      "epoch": 0.1581838154522153,
+      "grad_norm": 0.360953722994529,
+      "learning_rate": 1.5766423357664234e-05,
+      "loss": 0.2994,
+      "step": 108
+    },
+    {
+      "epoch": 0.15964848041010618,
+      "grad_norm": 0.44388991308142534,
+      "learning_rate": 1.591240875912409e-05,
+      "loss": 0.3456,
+      "step": 109
+    },
+    {
+      "epoch": 0.16111314536799706,
+      "grad_norm": 0.3537174118941569,
+      "learning_rate": 1.6058394160583944e-05,
+      "loss": 0.3091,
+      "step": 110
+    },
+    {
+      "epoch": 0.16257781032588794,
+      "grad_norm": 0.3713881906195737,
+      "learning_rate": 1.62043795620438e-05,
+      "loss": 0.347,
+      "step": 111
+    },
+    {
+      "epoch": 0.16404247528377883,
+      "grad_norm": 0.35286710247968334,
+      "learning_rate": 1.635036496350365e-05,
+      "loss": 0.3202,
+      "step": 112
+    },
+    {
+      "epoch": 0.1655071402416697,
+      "grad_norm": 0.34725613487477996,
+      "learning_rate": 1.6496350364963505e-05,
+      "loss": 0.3049,
+      "step": 113
+    },
+    {
+      "epoch": 0.1669718051995606,
+      "grad_norm": 0.3362671204409347,
+      "learning_rate": 1.664233576642336e-05,
+      "loss": 0.3197,
+      "step": 114
+    },
+    {
+      "epoch": 0.16843647015745147,
+      "grad_norm": 0.3464409043696399,
+      "learning_rate": 1.678832116788321e-05,
+      "loss": 0.3218,
+      "step": 115
+    },
+    {
+      "epoch": 0.16990113511534236,
+      "grad_norm": 0.3480542874556784,
+      "learning_rate": 1.6934306569343066e-05,
+      "loss": 0.3239,
+      "step": 116
+    },
+    {
+      "epoch": 0.17136580007323324,
+      "grad_norm": 0.3748150806007165,
+      "learning_rate": 1.708029197080292e-05,
+      "loss": 0.3157,
+      "step": 117
+    },
+    {
+      "epoch": 0.17283046503112412,
+      "grad_norm": 0.34345916023717576,
+      "learning_rate": 1.7226277372262773e-05,
+      "loss": 0.3051,
+      "step": 118
+    },
+    {
+      "epoch": 0.174295129989015,
+      "grad_norm": 0.34387241825105747,
+      "learning_rate": 1.737226277372263e-05,
+      "loss": 0.3179,
+      "step": 119
+    },
+    {
+      "epoch": 0.17575979494690588,
+      "grad_norm": 0.33782864040586463,
+      "learning_rate": 1.7518248175182482e-05,
+      "loss": 0.2856,
+      "step": 120
+    },
+    {
+      "epoch": 0.17722445990479677,
+      "grad_norm": 0.37444408915373373,
+      "learning_rate": 1.7664233576642337e-05,
+      "loss": 0.3251,
+      "step": 121
+    },
+    {
+      "epoch": 0.17868912486268765,
+      "grad_norm": 0.4116316036207847,
+      "learning_rate": 1.7810218978102192e-05,
+      "loss": 0.3048,
+      "step": 122
+    },
+    {
+      "epoch": 0.18015378982057853,
+      "grad_norm": 0.3781332494432965,
+      "learning_rate": 1.7956204379562047e-05,
+      "loss": 0.3008,
+      "step": 123
+    },
+    {
+      "epoch": 0.1816184547784694,
+      "grad_norm": 0.4297088346511556,
+      "learning_rate": 1.81021897810219e-05,
+      "loss": 0.3094,
+      "step": 124
+    },
+    {
+      "epoch": 0.1830831197363603,
+      "grad_norm": 0.4029065646398177,
+      "learning_rate": 1.8248175182481753e-05,
+      "loss": 0.3113,
+      "step": 125
+    },
+    {
+      "epoch": 0.18454778469425118,
+      "grad_norm": 0.3992462478484584,
+      "learning_rate": 1.8394160583941608e-05,
+      "loss": 0.2854,
+      "step": 126
+    },
+    {
+      "epoch": 0.18601244965214206,
+      "grad_norm": 0.3604804381477147,
+      "learning_rate": 1.854014598540146e-05,
+      "loss": 0.2981,
+      "step": 127
+    },
+    {
+      "epoch": 0.18747711461003294,
+      "grad_norm": 0.37186861704413154,
+      "learning_rate": 1.8686131386861315e-05,
+      "loss": 0.3258,
+      "step": 128
+    },
+    {
+      "epoch": 0.18894177956792385,
+      "grad_norm": 0.3441534820222307,
+      "learning_rate": 1.883211678832117e-05,
+      "loss": 0.2956,
+      "step": 129
+    },
+    {
+      "epoch": 0.19040644452581473,
+      "grad_norm": 0.4360804547373249,
+      "learning_rate": 1.897810218978102e-05,
+      "loss": 0.3091,
+      "step": 130
+    },
+    {
+      "epoch": 0.19187110948370562,
+      "grad_norm": 0.36928340495254985,
+      "learning_rate": 1.912408759124088e-05,
+      "loss": 0.2982,
+      "step": 131
+    },
+    {
+      "epoch": 0.1933357744415965,
+      "grad_norm": 0.36064269917913877,
+      "learning_rate": 1.9270072992700734e-05,
+      "loss": 0.301,
+      "step": 132
+    },
+    {
+      "epoch": 0.19480043939948738,
+      "grad_norm": 0.3364056638775238,
+      "learning_rate": 1.9416058394160586e-05,
+      "loss": 0.2643,
+      "step": 133
+    },
+    {
+      "epoch": 0.19626510435737826,
+      "grad_norm": 0.385357913235113,
+      "learning_rate": 1.956204379562044e-05,
+      "loss": 0.3151,
+      "step": 134
+    },
+    {
+      "epoch": 0.19772976931526914,
+      "grad_norm": 0.4013408692909421,
+      "learning_rate": 1.9708029197080295e-05,
+      "loss": 0.2937,
+      "step": 135
+    },
+    {
+      "epoch": 0.19919443427316003,
+      "grad_norm": 0.36182554518094284,
+      "learning_rate": 1.9854014598540147e-05,
+      "loss": 0.3009,
+      "step": 136
+    },
+    {
+      "epoch": 0.2006590992310509,
+      "grad_norm": 0.3494640989877029,
+      "learning_rate": 2e-05,
+      "loss": 0.281,
+      "step": 137
+    },
+    {
+      "epoch": 0.2021237641889418,
+      "grad_norm": 0.3725320500681877,
+      "learning_rate": 1.999996722215577e-05,
+      "loss": 0.3091,
+      "step": 138
+    },
+    {
+      "epoch": 0.20358842914683267,
+      "grad_norm": 0.3623692853258256,
+      "learning_rate": 1.9999868888837957e-05,
+      "loss": 0.2911,
+      "step": 139
+    },
+    {
+      "epoch": 0.20505309410472355,
+      "grad_norm": 0.36761059973259275,
+      "learning_rate": 1.999970500069119e-05,
+      "loss": 0.2773,
+      "step": 140
+    },
+    {
+      "epoch": 0.20651775906261444,
+      "grad_norm": 0.34372760731316254,
+      "learning_rate": 1.999947555878985e-05,
+      "loss": 0.2524,
+      "step": 141
+    },
+    {
+      "epoch": 0.20798242402050532,
+      "grad_norm": 0.34586143693014154,
+      "learning_rate": 1.9999180564638056e-05,
+      "loss": 0.3144,
+      "step": 142
+    },
+    {
+      "epoch": 0.2094470889783962,
+      "grad_norm": 0.3715249380161417,
+      "learning_rate": 1.9998820020169668e-05,
+      "loss": 0.2698,
+      "step": 143
+    },
+    {
+      "epoch": 0.21091175393628708,
+      "grad_norm": 0.35728560877179666,
+      "learning_rate": 1.9998393927748257e-05,
+      "loss": 0.2877,
+      "step": 144
+    },
+    {
+      "epoch": 0.21237641889417797,
+      "grad_norm": 0.3457017556464188,
+      "learning_rate": 1.9997902290167104e-05,
+      "loss": 0.2634,
+      "step": 145
+    },
+    {
+      "epoch": 0.21384108385206885,
+      "grad_norm": 0.32677024848818564,
+      "learning_rate": 1.999734511064917e-05,
+      "loss": 0.2962,
+      "step": 146
+    },
+    {
+      "epoch": 0.21530574880995973,
+      "grad_norm": 0.38425582050655327,
+      "learning_rate": 1.9996722392847082e-05,
+      "loss": 0.2769,
+      "step": 147
+    },
+    {
+      "epoch": 0.2167704137678506,
+      "grad_norm": 0.34979074920621583,
+      "learning_rate": 1.9996034140843113e-05,
+      "loss": 0.281,
+      "step": 148
+    },
+    {
+      "epoch": 0.2182350787257415,
+      "grad_norm": 0.32730300075566926,
+      "learning_rate": 1.999528035914915e-05,
+      "loss": 0.3143,
+      "step": 149
+    },
+    {
+      "epoch": 0.21969974368363238,
+      "grad_norm": 0.35690312453916395,
+      "learning_rate": 1.9994461052706652e-05,
+      "loss": 0.2508,
+      "step": 150
+    },
+    {
+      "epoch": 0.22116440864152326,
+      "grad_norm": 0.34613822211772644,
+      "learning_rate": 1.9993576226886644e-05,
+      "loss": 0.2793,
+      "step": 151
+    },
+    {
+      "epoch": 0.22262907359941414,
+      "grad_norm": 0.3922751418648848,
+      "learning_rate": 1.999262588748966e-05,
+      "loss": 0.2875,
+      "step": 152
+    },
+    {
+      "epoch": 0.22409373855730502,
+      "grad_norm": 0.3521535383506413,
+      "learning_rate": 1.9991610040745718e-05,
+      "loss": 0.2673,
+      "step": 153
+    },
+    {
+      "epoch": 0.2255584035151959,
+      "grad_norm": 0.33944461757575506,
+      "learning_rate": 1.9990528693314273e-05,
+      "loss": 0.2553,
+      "step": 154
+    },
+    {
+      "epoch": 0.2270230684730868,
+      "grad_norm": 0.3428799092949669,
+      "learning_rate": 1.9989381852284165e-05,
+      "loss": 0.274,
+      "step": 155
+    },
+    {
+      "epoch": 0.22848773343097767,
+      "grad_norm": 0.3511079662901022,
+      "learning_rate": 1.99881695251736e-05,
+      "loss": 0.2612,
+      "step": 156
+    },
+    {
+      "epoch": 0.22995239838886855,
+      "grad_norm": 0.33067647835038166,
+      "learning_rate": 1.998689171993006e-05,
+      "loss": 0.2653,
+      "step": 157
+    },
+    {
+      "epoch": 0.23141706334675943,
+      "grad_norm": 0.3487260974580051,
+      "learning_rate": 1.9985548444930295e-05,
+      "loss": 0.3049,
+      "step": 158
+    },
+    {
+      "epoch": 0.23288172830465032,
+      "grad_norm": 0.32680975853473054,
+      "learning_rate": 1.9984139708980228e-05,
+      "loss": 0.2641,
+      "step": 159
+    },
+    {
+      "epoch": 0.2343463932625412,
+      "grad_norm": 0.34722059123958804,
+      "learning_rate": 1.9982665521314934e-05,
+      "loss": 0.2713,
+      "step": 160
+    },
+    {
+      "epoch": 0.23581105822043208,
+      "grad_norm": 0.34801179824934036,
+      "learning_rate": 1.9981125891598545e-05,
+      "loss": 0.267,
+      "step": 161
+    },
+    {
+      "epoch": 0.23727572317832296,
+      "grad_norm": 0.3624289342633928,
+      "learning_rate": 1.9979520829924212e-05,
+      "loss": 0.263,
+      "step": 162
+    },
+    {
+      "epoch": 0.23874038813621384,
+      "grad_norm": 0.3299205118170384,
+      "learning_rate": 1.9977850346814026e-05,
+      "loss": 0.2919,
+      "step": 163
+    },
+    {
+      "epoch": 0.24020505309410473,
+      "grad_norm": 0.3595355800684274,
+      "learning_rate": 1.997611445321896e-05,
+      "loss": 0.2934,
+      "step": 164
+    },
+    {
+      "epoch": 0.2416697180519956,
+      "grad_norm": 0.3544383226618447,
+      "learning_rate": 1.9974313160518776e-05,
+      "loss": 0.2505,
+      "step": 165
+    },
+    {
+      "epoch": 0.2431343830098865,
+      "grad_norm": 0.34482857147345436,
+      "learning_rate": 1.9972446480521972e-05,
+      "loss": 0.2858,
+      "step": 166
+    },
+    {
+      "epoch": 0.24459904796777737,
+      "grad_norm": 0.36217560659010756,
+      "learning_rate": 1.9970514425465706e-05,
+      "loss": 0.2478,
+      "step": 167
+    },
+    {
+      "epoch": 0.24606371292566825,
+      "grad_norm": 0.37793259918304806,
+      "learning_rate": 1.996851700801569e-05,
+      "loss": 0.2557,
+      "step": 168
+    },
+    {
+      "epoch": 0.24752837788355914,
+      "grad_norm": 0.34823612797091924,
+      "learning_rate": 1.996645424126613e-05,
+      "loss": 0.2449,
+      "step": 169
+    },
+    {
+      "epoch": 0.24899304284145002,
+      "grad_norm": 0.3456664229365294,
+      "learning_rate": 1.9964326138739645e-05,
+      "loss": 0.2593,
+      "step": 170
+    },
+    {
+      "epoch": 0.2504577077993409,
+      "grad_norm": 0.36365236173385046,
+      "learning_rate": 1.996213271438715e-05,
+      "loss": 0.26,
+      "step": 171
+    },
+    {
+      "epoch": 0.2519223727572318,
+      "grad_norm": 0.3807231884615767,
+      "learning_rate": 1.9959873982587795e-05,
+      "loss": 0.2976,
+      "step": 172
+    },
+    {
+      "epoch": 0.25338703771512266,
+      "grad_norm": 0.373069506679343,
+      "learning_rate": 1.9957549958148844e-05,
+      "loss": 0.2895,
+      "step": 173
+    },
+    {
+      "epoch": 0.25485170267301355,
+      "grad_norm": 0.3531411189874793,
+      "learning_rate": 1.9955160656305606e-05,
+      "loss": 0.2483,
+      "step": 174
+    },
+    {
+      "epoch": 0.25631636763090443,
+      "grad_norm": 0.35346645678091215,
+      "learning_rate": 1.995270609272131e-05,
+      "loss": 0.2649,
+      "step": 175
+    },
+    {
+      "epoch": 0.2577810325887953,
+      "grad_norm": 0.3415190871733395,
+      "learning_rate": 1.995018628348702e-05,
+      "loss": 0.2523,
+      "step": 176
+    },
+    {
+      "epoch": 0.2592456975466862,
+      "grad_norm": 0.3654772088675571,
+      "learning_rate": 1.9947601245121514e-05,
+      "loss": 0.3083,
+      "step": 177
+    },
+    {
+      "epoch": 0.2607103625045771,
+      "grad_norm": 0.3460461457455375,
+      "learning_rate": 1.9944950994571192e-05,
+      "loss": 0.2792,
+      "step": 178
+    },
+    {
+      "epoch": 0.26217502746246796,
+      "grad_norm": 0.3110444104590562,
+      "learning_rate": 1.9942235549209955e-05,
+      "loss": 0.2688,
+      "step": 179
+    },
+    {
+      "epoch": 0.26363969242035884,
+      "grad_norm": 0.3272499534539241,
+      "learning_rate": 1.993945492683909e-05,
+      "loss": 0.2792,
+      "step": 180
+    },
+    {
+      "epoch": 0.2651043573782497,
+      "grad_norm": 0.2878269183852618,
+      "learning_rate": 1.993660914568716e-05,
+      "loss": 0.2424,
+      "step": 181
+    },
+    {
+      "epoch": 0.2665690223361406,
+      "grad_norm": 0.33774279808544694,
+      "learning_rate": 1.9933698224409876e-05,
+      "loss": 0.2496,
+      "step": 182
+    },
+    {
+      "epoch": 0.2680336872940315,
+      "grad_norm": 0.3363925870755136,
+      "learning_rate": 1.993072218208999e-05,
+      "loss": 0.2613,
+      "step": 183
+    },
+    {
+      "epoch": 0.26949835225192237,
+      "grad_norm": 0.3304059597878355,
+      "learning_rate": 1.992768103823714e-05,
+      "loss": 0.2523,
+      "step": 184
+    },
+    {
+      "epoch": 0.27096301720981325,
+      "grad_norm": 0.3556147552320389,
+      "learning_rate": 1.9924574812787766e-05,
+      "loss": 0.2528,
+      "step": 185
+    },
+    {
+      "epoch": 0.27242768216770413,
+      "grad_norm": 0.33358383808877295,
+      "learning_rate": 1.992140352610494e-05,
+      "loss": 0.2371,
+      "step": 186
+    },
+    {
+      "epoch": 0.273892347125595,
+      "grad_norm": 0.32338465251509535,
+      "learning_rate": 1.9918167198978246e-05,
+      "loss": 0.2328,
+      "step": 187
+    },
+    {
+      "epoch": 0.2753570120834859,
+      "grad_norm": 0.34675833301901965,
+      "learning_rate": 1.991486585262365e-05,
+      "loss": 0.2441,
+      "step": 188
+    },
+    {
+      "epoch": 0.2768216770413768,
+      "grad_norm": 0.32681910042540224,
+      "learning_rate": 1.991149950868336e-05,
+      "loss": 0.2244,
+      "step": 189
+    },
+    {
+      "epoch": 0.27828634199926766,
+      "grad_norm": 0.33323736986242397,
+      "learning_rate": 1.9908068189225672e-05,
+      "loss": 0.2375,
+      "step": 190
+    },
+    {
+      "epoch": 0.27975100695715854,
+      "grad_norm": 0.30868253450774236,
+      "learning_rate": 1.9904571916744836e-05,
+      "loss": 0.2354,
+      "step": 191
+    },
+    {
+      "epoch": 0.2812156719150494,
+      "grad_norm": 0.3404026835089613,
+      "learning_rate": 1.990101071416091e-05,
+      "loss": 0.244,
+      "step": 192
+    },
+    {
+      "epoch": 0.2826803368729403,
+      "grad_norm": 0.3250658440023371,
+      "learning_rate": 1.98973846048196e-05,
+      "loss": 0.2436,
+      "step": 193
+    },
+    {
+      "epoch": 0.2841450018308312,
+      "grad_norm": 0.3727507902802708,
+      "learning_rate": 1.9893693612492116e-05,
+      "loss": 0.2605,
+      "step": 194
+    },
+    {
+      "epoch": 0.28560966678872207,
+      "grad_norm": 0.32087796418849673,
+      "learning_rate": 1.9889937761375015e-05,
+      "loss": 0.2473,
+      "step": 195
+    },
+    {
+      "epoch": 0.28707433174661295,
+      "grad_norm": 0.3017096010090274,
+      "learning_rate": 1.9886117076090033e-05,
+      "loss": 0.2333,
+      "step": 196
+    },
+    {
+      "epoch": 0.28853899670450384,
+      "grad_norm": 0.32081985018575454,
+      "learning_rate": 1.9882231581683938e-05,
+      "loss": 0.2401,
+      "step": 197
+    },
+    {
+      "epoch": 0.2900036616623947,
+      "grad_norm": 0.31230329800532286,
+      "learning_rate": 1.9878281303628352e-05,
+      "loss": 0.2362,
+      "step": 198
+    },
+    {
+      "epoch": 0.2914683266202856,
+      "grad_norm": 0.3218942752729413,
+      "learning_rate": 1.9874266267819604e-05,
+      "loss": 0.248,
+      "step": 199
+    },
+    {
+      "epoch": 0.2929329915781765,
+      "grad_norm": 0.34876467385745175,
+      "learning_rate": 1.987018650057853e-05,
+      "loss": 0.2487,
+      "step": 200
+    },
+    {
+      "epoch": 0.29439765653606736,
+      "grad_norm": 0.31745025914965086,
+      "learning_rate": 1.986604202865033e-05,
+      "loss": 0.2394,
+      "step": 201
+    },
+    {
+      "epoch": 0.29586232149395825,
+      "grad_norm": 0.32627016177493307,
+      "learning_rate": 1.986183287920437e-05,
+      "loss": 0.2624,
+      "step": 202
+    },
+    {
+      "epoch": 0.29732698645184913,
+      "grad_norm": 0.32599240801438883,
+      "learning_rate": 1.9857559079834023e-05,
+      "loss": 0.2404,
+      "step": 203
+    },
+    {
+      "epoch": 0.29879165140974,
+      "grad_norm": 0.3110886547328817,
+      "learning_rate": 1.9853220658556474e-05,
+      "loss": 0.246,
+      "step": 204
+    },
+    {
+      "epoch": 0.3002563163676309,
+      "grad_norm": 0.3439412203840107,
+      "learning_rate": 1.984881764381254e-05,
+      "loss": 0.2245,
+      "step": 205
+    },
+    {
+      "epoch": 0.3017209813255218,
+      "grad_norm": 0.32093087763294337,
+      "learning_rate": 1.9844350064466488e-05,
+      "loss": 0.2351,
+      "step": 206
+    },
+    {
+      "epoch": 0.30318564628341266,
+      "grad_norm": 0.3080777182901835,
+      "learning_rate": 1.9839817949805843e-05,
+      "loss": 0.2239,
+      "step": 207
+    },
+    {
+      "epoch": 0.30465031124130354,
+      "grad_norm": 0.33465999975388977,
+      "learning_rate": 1.9835221329541197e-05,
+      "loss": 0.2362,
+      "step": 208
+    },
+    {
+      "epoch": 0.3061149761991944,
+      "grad_norm": 0.2916494897812547,
+      "learning_rate": 1.9830560233806006e-05,
+      "loss": 0.2138,
+      "step": 209
+    },
+    {
+      "epoch": 0.3075796411570853,
+      "grad_norm": 0.31069474576355127,
+      "learning_rate": 1.9825834693156408e-05,
+      "loss": 0.2381,
+      "step": 210
+    },
+    {
+      "epoch": 0.3090443061149762,
+      "grad_norm": 0.30904277970849947,
+      "learning_rate": 1.9821044738571008e-05,
+      "loss": 0.2336,
+      "step": 211
+    },
+    {
+      "epoch": 0.31050897107286707,
+      "grad_norm": 0.30754570168689266,
+      "learning_rate": 1.981619040145068e-05,
+      "loss": 0.2209,
+      "step": 212
+    },
+    {
+      "epoch": 0.31197363603075795,
+      "grad_norm": 0.2976312068720902,
+      "learning_rate": 1.9811271713618372e-05,
+      "loss": 0.2558,
+      "step": 213
+    },
+    {
+      "epoch": 0.31343830098864883,
+      "grad_norm": 0.3039100122516498,
+      "learning_rate": 1.980628870731888e-05,
+      "loss": 0.2265,
+      "step": 214
+    },
+    {
+      "epoch": 0.3149029659465397,
+      "grad_norm": 0.3136540735451798,
+      "learning_rate": 1.9801241415218636e-05,
+      "loss": 0.2551,
+      "step": 215
+    },
+    {
+      "epoch": 0.3163676309044306,
+      "grad_norm": 0.30585301656209735,
+      "learning_rate": 1.979612987040552e-05,
+      "loss": 0.2258,
+      "step": 216
+    },
+    {
+      "epoch": 0.3178322958623215,
+      "grad_norm": 0.3334394299522592,
+      "learning_rate": 1.9790954106388614e-05,
+      "loss": 0.2374,
+      "step": 217
+    },
+    {
+      "epoch": 0.31929696082021236,
+      "grad_norm": 0.3385575692037135,
+      "learning_rate": 1.9785714157097992e-05,
+      "loss": 0.2331,
+      "step": 218
+    },
+    {
+      "epoch": 0.32076162577810324,
+      "grad_norm": 0.3550097472590768,
+      "learning_rate": 1.9780410056884505e-05,
+      "loss": 0.2316,
+      "step": 219
+    },
+    {
+      "epoch": 0.3222262907359941,
+      "grad_norm": 0.3212031876917821,
+      "learning_rate": 1.9775041840519547e-05,
+      "loss": 0.264,
+      "step": 220
+    },
+    {
+      "epoch": 0.323690955693885,
+      "grad_norm": 0.3417161392653498,
+      "learning_rate": 1.976960954319483e-05,
+      "loss": 0.2366,
+      "step": 221
+    },
+    {
+      "epoch": 0.3251556206517759,
+      "grad_norm": 0.31739055359370316,
+      "learning_rate": 1.9764113200522153e-05,
+      "loss": 0.2193,
+      "step": 222
+    },
+    {
+      "epoch": 0.32662028560966677,
+      "grad_norm": 0.3475167845756447,
+      "learning_rate": 1.9758552848533168e-05,
+      "loss": 0.2381,
+      "step": 223
+    },
+    {
+      "epoch": 0.32808495056755765,
+      "grad_norm": 0.3227407213456651,
+      "learning_rate": 1.9752928523679145e-05,
+      "loss": 0.2403,
+      "step": 224
+    },
+    {
+      "epoch": 0.32954961552544854,
+      "grad_norm": 0.32134472119602675,
+      "learning_rate": 1.9747240262830734e-05,
+      "loss": 0.244,
+      "step": 225
+    },
+    {
+      "epoch": 0.3310142804833394,
+      "grad_norm": 0.3060769394485448,
+      "learning_rate": 1.9741488103277722e-05,
+      "loss": 0.2029,
+      "step": 226
+    },
+    {
+      "epoch": 0.3324789454412303,
+      "grad_norm": 0.3422387817778384,
+      "learning_rate": 1.9735672082728785e-05,
+      "loss": 0.2166,
+      "step": 227
+    },
+    {
+      "epoch": 0.3339436103991212,
+      "grad_norm": 0.29907859948451515,
+      "learning_rate": 1.9729792239311243e-05,
+      "loss": 0.223,
+      "step": 228
+    },
+    {
+      "epoch": 0.33540827535701206,
+      "grad_norm": 0.27950206671923955,
+      "learning_rate": 1.972384861157082e-05,
+      "loss": 0.1952,
+      "step": 229
+    },
+    {
+      "epoch": 0.33687294031490295,
+      "grad_norm": 0.3083999590796091,
+      "learning_rate": 1.9717841238471377e-05,
+      "loss": 0.2188,
+      "step": 230
+    },
+    {
+      "epoch": 0.33833760527279383,
+      "grad_norm": 0.295065591964525,
+      "learning_rate": 1.9711770159394654e-05,
+      "loss": 0.2091,
+      "step": 231
+    },
+    {
+      "epoch": 0.3398022702306847,
+      "grad_norm": 0.30797729325093864,
+      "learning_rate": 1.9705635414140035e-05,
+      "loss": 0.2106,
+      "step": 232
+    },
+    {
+      "epoch": 0.3412669351885756,
+      "grad_norm": 0.31169542447907694,
+      "learning_rate": 1.9699437042924266e-05,
+      "loss": 0.2398,
+      "step": 233
+    },
+    {
+      "epoch": 0.3427316001464665,
+      "grad_norm": 0.340358999491338,
+      "learning_rate": 1.969317508638119e-05,
+      "loss": 0.2564,
+      "step": 234
+    },
+    {
+      "epoch": 0.34419626510435736,
+      "grad_norm": 0.3017727587765966,
+      "learning_rate": 1.96868495855615e-05,
+      "loss": 0.2024,
+      "step": 235
+    },
+    {
+      "epoch": 0.34566093006224824,
+      "grad_norm": 0.33465091974213557,
+      "learning_rate": 1.9680460581932448e-05,
+      "loss": 0.2418,
+      "step": 236
+    },
+    {
+      "epoch": 0.3471255950201391,
+      "grad_norm": 0.29191015974637435,
+      "learning_rate": 1.967400811737759e-05,
+      "loss": 0.2018,
+      "step": 237
+    },
+    {
+      "epoch": 0.34859025997803,
+      "grad_norm": 0.28852204171964263,
+      "learning_rate": 1.96674922341965e-05,
+      "loss": 0.2341,
+      "step": 238
+    },
+    {
+      "epoch": 0.3500549249359209,
+      "grad_norm": 0.32713188846613395,
+      "learning_rate": 1.96609129751045e-05,
+      "loss": 0.2304,
+      "step": 239
+    },
+    {
+      "epoch": 0.35151958989381177,
+      "grad_norm": 0.2960166330763219,
+      "learning_rate": 1.9654270383232377e-05,
+      "loss": 0.218,
+      "step": 240
+    },
+    {
+      "epoch": 0.35298425485170265,
+      "grad_norm": 0.3129431801548695,
+      "learning_rate": 1.9647564502126094e-05,
+      "loss": 0.2117,
+      "step": 241
+    },
+    {
+      "epoch": 0.35444891980959353,
+      "grad_norm": 0.294963922960848,
+      "learning_rate": 1.964079537574652e-05,
+      "loss": 0.2203,
+      "step": 242
+    },
+    {
+      "epoch": 0.3559135847674844,
+      "grad_norm": 0.3164361378068273,
+      "learning_rate": 1.963396304846913e-05,
+      "loss": 0.2199,
+      "step": 243
+    },
+    {
+      "epoch": 0.3573782497253753,
+      "grad_norm": 0.3207780914062909,
+      "learning_rate": 1.9627067565083716e-05,
+      "loss": 0.2094,
+      "step": 244
+    },
+    {
+      "epoch": 0.3588429146832662,
+      "grad_norm": 0.31070213874769786,
+      "learning_rate": 1.962010897079409e-05,
+      "loss": 0.1875,
+      "step": 245
+    },
+    {
+      "epoch": 0.36030757964115706,
+      "grad_norm": 0.30122482811704177,
+      "learning_rate": 1.96130873112178e-05,
+      "loss": 0.2377,
+      "step": 246
+    },
+    {
+      "epoch": 0.36177224459904794,
+      "grad_norm": 0.3070889454659401,
+      "learning_rate": 1.9606002632385817e-05,
+      "loss": 0.1948,
+      "step": 247
+    },
+    {
+      "epoch": 0.3632369095569388,
+      "grad_norm": 0.30277497066736986,
+      "learning_rate": 1.959885498074224e-05,
+      "loss": 0.2122,
+      "step": 248
+    },
+    {
+      "epoch": 0.3647015745148297,
+      "grad_norm": 0.33010916657115674,
+      "learning_rate": 1.9591644403143997e-05,
+      "loss": 0.1988,
+      "step": 249
+    },
+    {
+      "epoch": 0.3661662394727206,
+      "grad_norm": 0.3187016737580306,
+      "learning_rate": 1.958437094686052e-05,
+      "loss": 0.1954,
+      "step": 250
+    },
+    {
+      "epoch": 0.36763090443061147,
+      "grad_norm": 0.29492570210704677,
+      "learning_rate": 1.9577034659573452e-05,
+      "loss": 0.1932,
+      "step": 251
+    },
+    {
+      "epoch": 0.36909556938850235,
+      "grad_norm": 0.31163033056152234,
+      "learning_rate": 1.956963558937633e-05,
+      "loss": 0.234,
+      "step": 252
+    },
+    {
+      "epoch": 0.37056023434639324,
+      "grad_norm": 0.29081575214116195,
+      "learning_rate": 1.9562173784774274e-05,
+      "loss": 0.2256,
+      "step": 253
+    },
+    {
+      "epoch": 0.3720248993042841,
+      "grad_norm": 0.2859981212044829,
+      "learning_rate": 1.955464929468365e-05,
+      "loss": 0.1863,
+      "step": 254
+    },
+    {
+      "epoch": 0.373489564262175,
+      "grad_norm": 0.3362313373130634,
+      "learning_rate": 1.9547062168431777e-05,
+      "loss": 0.2191,
+      "step": 255
+    },
+    {
+      "epoch": 0.3749542292200659,
+      "grad_norm": 0.31580606455125804,
+      "learning_rate": 1.9539412455756578e-05,
+      "loss": 0.2219,
+      "step": 256
+    },
+    {
+      "epoch": 0.3764188941779568,
+      "grad_norm": 0.3138360531715862,
+      "learning_rate": 1.9531700206806274e-05,
+      "loss": 0.2216,
+      "step": 257
+    },
+    {
+      "epoch": 0.3778835591358477,
+      "grad_norm": 0.30577065715195095,
+      "learning_rate": 1.952392547213904e-05,
+      "loss": 0.2143,
+      "step": 258
+    },
+    {
+      "epoch": 0.3793482240937386,
+      "grad_norm": 0.3030869019075986,
+      "learning_rate": 1.9516088302722696e-05,
+      "loss": 0.2155,
+      "step": 259
+    },
+    {
+      "epoch": 0.38081288905162947,
+      "grad_norm": 0.33061172813654255,
+      "learning_rate": 1.9508188749934333e-05,
+      "loss": 0.2375,
+      "step": 260
+    },
+    {
+      "epoch": 0.38227755400952035,
+      "grad_norm": 0.31425985710724597,
+      "learning_rate": 1.9500226865560015e-05,
+      "loss": 0.197,
+      "step": 261
+    },
+    {
+      "epoch": 0.38374221896741123,
+      "grad_norm": 0.3084208020631583,
+      "learning_rate": 1.9492202701794432e-05,
+      "loss": 0.2062,
+      "step": 262
+    },
+    {
+      "epoch": 0.3852068839253021,
+      "grad_norm": 0.31591794037568377,
+      "learning_rate": 1.9484116311240534e-05,
+      "loss": 0.2012,
+      "step": 263
+    },
+    {
+      "epoch": 0.386671548883193,
+      "grad_norm": 0.3243347143178351,
+      "learning_rate": 1.9475967746909212e-05,
+      "loss": 0.2157,
+      "step": 264
+    },
+    {
+      "epoch": 0.3881362138410839,
+      "grad_norm": 0.3126285934402634,
+      "learning_rate": 1.946775706221894e-05,
+      "loss": 0.2015,
+      "step": 265
+    },
+    {
+      "epoch": 0.38960087879897476,
+      "grad_norm": 0.29362391198343335,
+      "learning_rate": 1.945948431099543e-05,
+      "loss": 0.2082,
+      "step": 266
+    },
+    {
+      "epoch": 0.39106554375686564,
+      "grad_norm": 0.3306599758257121,
+      "learning_rate": 1.945114954747127e-05,
+      "loss": 0.2217,
+      "step": 267
+    },
+    {
+      "epoch": 0.3925302087147565,
+      "grad_norm": 0.31104138128230796,
+      "learning_rate": 1.9442752826285578e-05,
+      "loss": 0.2347,
+      "step": 268
+    },
+    {
+      "epoch": 0.3939948736726474,
+      "grad_norm": 0.2882519240767416,
+      "learning_rate": 1.9434294202483634e-05,
+      "loss": 0.189,
+      "step": 269
+    },
+    {
+      "epoch": 0.3954595386305383,
+      "grad_norm": 0.2993561393461098,
+      "learning_rate": 1.9425773731516534e-05,
+      "loss": 0.197,
+      "step": 270
+    },
+    {
+      "epoch": 0.39692420358842917,
+      "grad_norm": 0.31696676157406817,
+      "learning_rate": 1.9417191469240806e-05,
+      "loss": 0.1853,
+      "step": 271
+    },
+    {
+      "epoch": 0.39838886854632005,
+      "grad_norm": 0.2976550371764201,
+      "learning_rate": 1.940854747191806e-05,
+      "loss": 0.2062,
+      "step": 272
+    },
+    {
+      "epoch": 0.39985353350421093,
+      "grad_norm": 0.3102511410212597,
+      "learning_rate": 1.9399841796214625e-05,
+      "loss": 0.2118,
+      "step": 273
+    },
+    {
+      "epoch": 0.4013181984621018,
+      "grad_norm": 0.29711750834500367,
+      "learning_rate": 1.9391074499201155e-05,
+      "loss": 0.1736,
+      "step": 274
+    },
+    {
+      "epoch": 0.4027828634199927,
+      "grad_norm": 0.33461146425534555,
+      "learning_rate": 1.938224563835226e-05,
+      "loss": 0.2023,
+      "step": 275
+    },
+    {
+      "epoch": 0.4042475283778836,
+      "grad_norm": 0.2969695409049403,
+      "learning_rate": 1.9373355271546156e-05,
+      "loss": 0.2067,
+      "step": 276
+    },
+    {
+      "epoch": 0.40571219333577446,
+      "grad_norm": 0.32397668137142743,
+      "learning_rate": 1.9364403457064252e-05,
+      "loss": 0.2266,
+      "step": 277
+    },
+    {
+      "epoch": 0.40717685829366534,
+      "grad_norm": 0.35852753225097855,
+      "learning_rate": 1.9355390253590775e-05,
+      "loss": 0.2582,
+      "step": 278
+    },
+    {
+      "epoch": 0.4086415232515562,
+      "grad_norm": 0.2936115798624861,
+      "learning_rate": 1.9346315720212416e-05,
+      "loss": 0.1899,
+      "step": 279
+    },
+    {
+      "epoch": 0.4101061882094471,
+      "grad_norm": 0.3229854576353249,
+      "learning_rate": 1.933717991641789e-05,
+      "loss": 0.2103,
+      "step": 280
+    },
+    {
+      "epoch": 0.411570853167338,
+      "grad_norm": 0.29206997385224825,
+      "learning_rate": 1.9327982902097596e-05,
+      "loss": 0.2157,
+      "step": 281
+    },
+    {
+      "epoch": 0.4130355181252289,
+      "grad_norm": 0.30505940530286574,
+      "learning_rate": 1.931872473754319e-05,
+      "loss": 0.2077,
+      "step": 282
+    },
+    {
+      "epoch": 0.41450018308311976,
+      "grad_norm": 0.2973043577148213,
+      "learning_rate": 1.9309405483447208e-05,
+      "loss": 0.1964,
+      "step": 283
+    },
+    {
+      "epoch": 0.41596484804101064,
+      "grad_norm": 0.29785754593184066,
+      "learning_rate": 1.9300025200902666e-05,
+      "loss": 0.1913,
+      "step": 284
+    },
+    {
+      "epoch": 0.4174295129989015,
+      "grad_norm": 0.33220986093953964,
+      "learning_rate": 1.9290583951402648e-05,
+      "loss": 0.1997,
+      "step": 285
+    },
+    {
+      "epoch": 0.4188941779567924,
+      "grad_norm": 0.3165354043864197,
+      "learning_rate": 1.9281081796839915e-05,
+      "loss": 0.2028,
+      "step": 286
+    },
+    {
+      "epoch": 0.4203588429146833,
+      "grad_norm": 0.30361046171154804,
+      "learning_rate": 1.9271518799506494e-05,
+      "loss": 0.2004,
+      "step": 287
+    },
+    {
+      "epoch": 0.42182350787257417,
+      "grad_norm": 0.3125634361106797,
+      "learning_rate": 1.9261895022093275e-05,
+      "loss": 0.1981,
+      "step": 288
+    },
+    {
+      "epoch": 0.42328817283046505,
+      "grad_norm": 0.30844567580026055,
+      "learning_rate": 1.9252210527689596e-05,
+      "loss": 0.2073,
+      "step": 289
+    },
+    {
+      "epoch": 0.42475283778835593,
+      "grad_norm": 0.3155331823234916,
+      "learning_rate": 1.9242465379782823e-05,
+      "loss": 0.1854,
+      "step": 290
+    },
+    {
+      "epoch": 0.4262175027462468,
+      "grad_norm": 0.34604055990819416,
+      "learning_rate": 1.9232659642257942e-05,
+      "loss": 0.2627,
+      "step": 291
+    },
+    {
+      "epoch": 0.4276821677041377,
+      "grad_norm": 0.31441965292621443,
+      "learning_rate": 1.9222793379397146e-05,
+      "loss": 0.2214,
+      "step": 292
+    },
+    {
+      "epoch": 0.4291468326620286,
+      "grad_norm": 0.3341564227129849,
+      "learning_rate": 1.9212866655879397e-05,
+      "loss": 0.2002,
+      "step": 293
+    },
+    {
+      "epoch": 0.43061149761991946,
+      "grad_norm": 0.314168791046398,
+      "learning_rate": 1.9202879536780013e-05,
+      "loss": 0.199,
+      "step": 294
+    },
+    {
+      "epoch": 0.43207616257781034,
+      "grad_norm": 0.3415225885868485,
+      "learning_rate": 1.919283208757025e-05,
+      "loss": 0.1915,
+      "step": 295
+    },
+    {
+      "epoch": 0.4335408275357012,
+      "grad_norm": 0.34338764313976977,
+      "learning_rate": 1.918272437411684e-05,
+      "loss": 0.2134,
+      "step": 296
+    },
+    {
+      "epoch": 0.4350054924935921,
+      "grad_norm": 0.30091408846543016,
+      "learning_rate": 1.91725564626816e-05,
+      "loss": 0.1956,
+      "step": 297
+    },
+    {
+      "epoch": 0.436470157451483,
+      "grad_norm": 0.30066058971128345,
+      "learning_rate": 1.9162328419920976e-05,
+      "loss": 0.1901,
+      "step": 298
+    },
+    {
+      "epoch": 0.43793482240937387,
+      "grad_norm": 0.30021693344518857,
+      "learning_rate": 1.9152040312885604e-05,
+      "loss": 0.195,
+      "step": 299
+    },
+    {
+      "epoch": 0.43939948736726475,
+      "grad_norm": 0.30589418789890954,
+      "learning_rate": 1.914169220901988e-05,
+      "loss": 0.1904,
+      "step": 300
+    },
+    {
+      "epoch": 0.44086415232515563,
+      "grad_norm": 0.2916943622706973,
+      "learning_rate": 1.9131284176161505e-05,
+      "loss": 0.1862,
+      "step": 301
+    },
+    {
+      "epoch": 0.4423288172830465,
+      "grad_norm": 0.3111182978529323,
+      "learning_rate": 1.9120816282541062e-05,
+      "loss": 0.1903,
+      "step": 302
+    },
+    {
+      "epoch": 0.4437934822409374,
+      "grad_norm": 0.296170352579387,
+      "learning_rate": 1.911028859678155e-05,
+      "loss": 0.1859,
+      "step": 303
+    },
+    {
+      "epoch": 0.4452581471988283,
+      "grad_norm": 0.3144098925860641,
+      "learning_rate": 1.9099701187897927e-05,
+      "loss": 0.2115,
+      "step": 304
+    },
+    {
+      "epoch": 0.44672281215671916,
+      "grad_norm": 0.28420551119731563,
+      "learning_rate": 1.9089054125296692e-05,
+      "loss": 0.1786,
+      "step": 305
+    },
+    {
+      "epoch": 0.44818747711461004,
+      "grad_norm": 0.31553201198823727,
+      "learning_rate": 1.907834747877539e-05,
+      "loss": 0.1828,
+      "step": 306
+    },
+    {
+      "epoch": 0.4496521420725009,
+      "grad_norm": 0.3045234900369304,
+      "learning_rate": 1.906758131852218e-05,
+      "loss": 0.1969,
+      "step": 307
+    },
+    {
+      "epoch": 0.4511168070303918,
+      "grad_norm": 0.3172827302837682,
+      "learning_rate": 1.9056755715115372e-05,
+      "loss": 0.1961,
+      "step": 308
+    },
+    {
+      "epoch": 0.4525814719882827,
+      "grad_norm": 0.31871947429630393,
+      "learning_rate": 1.9045870739522953e-05,
+      "loss": 0.1833,
+      "step": 309
+    },
+    {
+      "epoch": 0.4540461369461736,
+      "grad_norm": 0.29640983142077165,
+      "learning_rate": 1.9034926463102122e-05,
+      "loss": 0.1735,
+      "step": 310
+    },
+    {
+      "epoch": 0.45551080190406446,
+      "grad_norm": 0.2838973203329574,
+      "learning_rate": 1.9023922957598847e-05,
+      "loss": 0.1879,
+      "step": 311
+    },
+    {
+      "epoch": 0.45697546686195534,
+      "grad_norm": 0.31619551941395607,
+      "learning_rate": 1.901286029514736e-05,
+      "loss": 0.1981,
+      "step": 312
+    },
+    {
+      "epoch": 0.4584401318198462,
+      "grad_norm": 0.28916971260480706,
+      "learning_rate": 1.9001738548269707e-05,
+      "loss": 0.1681,
+      "step": 313
+    },
+    {
+      "epoch": 0.4599047967777371,
+      "grad_norm": 0.29652343709617623,
+      "learning_rate": 1.8990557789875265e-05,
+      "loss": 0.1752,
+      "step": 314
+    },
+    {
+      "epoch": 0.461369461735628,
+      "grad_norm": 0.2952557530849967,
+      "learning_rate": 1.8979318093260268e-05,
+      "loss": 0.1731,
+      "step": 315
+    },
+    {
+      "epoch": 0.46283412669351887,
+      "grad_norm": 0.30085307847746645,
+      "learning_rate": 1.8968019532107318e-05,
+      "loss": 0.1975,
+      "step": 316
+    },
+    {
+      "epoch": 0.46429879165140975,
+      "grad_norm": 0.3187121453479262,
+      "learning_rate": 1.8956662180484913e-05,
+      "loss": 0.198,
+      "step": 317
+    },
+    {
+      "epoch": 0.46576345660930063,
+      "grad_norm": 0.30663390384150385,
+      "learning_rate": 1.8945246112846952e-05,
+      "loss": 0.1835,
+      "step": 318
+    },
+    {
+      "epoch": 0.4672281215671915,
+      "grad_norm": 0.30881852574686636,
+      "learning_rate": 1.893377140403225e-05,
+      "loss": 0.1828,
+      "step": 319
+    },
+    {
+      "epoch": 0.4686927865250824,
+      "grad_norm": 0.300908778514595,
+      "learning_rate": 1.892223812926406e-05,
+      "loss": 0.2214,
+      "step": 320
+    },
+    {
+      "epoch": 0.4701574514829733,
+      "grad_norm": 0.2860561871686056,
+      "learning_rate": 1.8910646364149548e-05,
+      "loss": 0.1938,
+      "step": 321
+    },
+    {
+      "epoch": 0.47162211644086416,
+      "grad_norm": 0.2772670520674923,
+      "learning_rate": 1.889899618467933e-05,
+      "loss": 0.1668,
+      "step": 322
+    },
+    {
+      "epoch": 0.47308678139875504,
+      "grad_norm": 0.3062971263411581,
+      "learning_rate": 1.8887287667226964e-05,
+      "loss": 0.199,
+      "step": 323
+    },
+    {
+      "epoch": 0.4745514463566459,
+      "grad_norm": 0.32124027529820043,
+      "learning_rate": 1.887552088854844e-05,
+      "loss": 0.2083,
+      "step": 324
+    },
+    {
+      "epoch": 0.4760161113145368,
+      "grad_norm": 0.295114219753382,
+      "learning_rate": 1.8863695925781685e-05,
+      "loss": 0.1887,
+      "step": 325
+    },
+    {
+      "epoch": 0.4774807762724277,
+      "grad_norm": 0.3194350024464684,
+      "learning_rate": 1.8851812856446062e-05,
+      "loss": 0.2086,
+      "step": 326
+    },
+    {
+      "epoch": 0.47894544123031857,
+      "grad_norm": 0.29449304384305286,
+      "learning_rate": 1.8839871758441842e-05,
+      "loss": 0.1898,
+      "step": 327
+    },
+    {
+      "epoch": 0.48041010618820945,
+      "grad_norm": 0.28550779650922725,
+      "learning_rate": 1.882787271004972e-05,
+      "loss": 0.1751,
+      "step": 328
+    },
+    {
+      "epoch": 0.48187477114610033,
+      "grad_norm": 0.28143493349397003,
+      "learning_rate": 1.8815815789930277e-05,
+      "loss": 0.1853,
+      "step": 329
+    },
+    {
+      "epoch": 0.4833394361039912,
+      "grad_norm": 0.3088998091001307,
+      "learning_rate": 1.8803701077123492e-05,
+      "loss": 0.1901,
+      "step": 330
+    },
+    {
+      "epoch": 0.4848041010618821,
+      "grad_norm": 0.31201486962963987,
+      "learning_rate": 1.8791528651048193e-05,
+      "loss": 0.1956,
+      "step": 331
+    },
+    {
+      "epoch": 0.486268766019773,
+      "grad_norm": 0.30000644267131227,
+      "learning_rate": 1.8779298591501565e-05,
+      "loss": 0.1852,
+      "step": 332
+    },
+    {
+      "epoch": 0.48773343097766386,
+      "grad_norm": 0.279989644276303,
+      "learning_rate": 1.8767010978658597e-05,
+      "loss": 0.1564,
+      "step": 333
+    },
+    {
+      "epoch": 0.48919809593555474,
+      "grad_norm": 0.31244506310168246,
+      "learning_rate": 1.8754665893071583e-05,
+      "loss": 0.2094,
+      "step": 334
+    },
+    {
+      "epoch": 0.4906627608934456,
+      "grad_norm": 0.32128122069078735,
+      "learning_rate": 1.874226341566958e-05,
+      "loss": 0.2013,
+      "step": 335
+    },
+    {
+      "epoch": 0.4921274258513365,
+      "grad_norm": 0.30462090670077413,
+      "learning_rate": 1.872980362775789e-05,
+      "loss": 0.1779,
+      "step": 336
+    },
+    {
+      "epoch": 0.4935920908092274,
+      "grad_norm": 0.30874094182185663,
+      "learning_rate": 1.87172866110175e-05,
+      "loss": 0.1976,
+      "step": 337
+    },
+    {
+      "epoch": 0.4950567557671183,
+      "grad_norm": 0.2978419876068271,
+      "learning_rate": 1.870471244750458e-05,
+      "loss": 0.1718,
+      "step": 338
+    },
+    {
+      "epoch": 0.49652142072500915,
+      "grad_norm": 0.30085229404822755,
+      "learning_rate": 1.8692081219649926e-05,
+      "loss": 0.1729,
+      "step": 339
+    },
+    {
+      "epoch": 0.49798608568290004,
+      "grad_norm": 0.309384694321714,
+      "learning_rate": 1.867939301025842e-05,
+      "loss": 0.1882,
+      "step": 340
+    },
+    {
+      "epoch": 0.4994507506407909,
+      "grad_norm": 0.28838820512419444,
+      "learning_rate": 1.8666647902508493e-05,
+      "loss": 0.1789,
+      "step": 341
+    },
+    {
+      "epoch": 0.5009154155986818,
+      "grad_norm": 0.32424956980873243,
+      "learning_rate": 1.8653845979951577e-05,
+      "loss": 0.1663,
+      "step": 342
+    },
+    {
+      "epoch": 0.5023800805565727,
+      "grad_norm": 0.31499571469715115,
+      "learning_rate": 1.864098732651155e-05,
+      "loss": 0.1991,
+      "step": 343
+    },
+    {
+      "epoch": 0.5038447455144636,
+      "grad_norm": 0.3158502822606953,
+      "learning_rate": 1.8628072026484215e-05,
+      "loss": 0.2042,
+      "step": 344
+    },
+    {
+      "epoch": 0.5053094104723544,
+      "grad_norm": 0.3006662825507403,
+      "learning_rate": 1.8615100164536696e-05,
+      "loss": 0.1799,
+      "step": 345
+    },
+    {
+      "epoch": 0.5067740754302453,
+      "grad_norm": 0.3197735718886773,
+      "learning_rate": 1.8602071825706928e-05,
+      "loss": 0.2037,
+      "step": 346
+    },
+    {
+      "epoch": 0.5082387403881362,
+      "grad_norm": 0.29431082437854705,
+      "learning_rate": 1.858898709540309e-05,
+      "loss": 0.1624,
+      "step": 347
+    },
+    {
+      "epoch": 0.5097034053460271,
+      "grad_norm": 0.3319207275942643,
+      "learning_rate": 1.8575846059403036e-05,
+      "loss": 0.1781,
+      "step": 348
+    },
+    {
+      "epoch": 0.511168070303918,
+      "grad_norm": 0.3548295674736163,
+      "learning_rate": 1.856264880385372e-05,
+      "loss": 0.1777,
+      "step": 349
+    },
+    {
+      "epoch": 0.5126327352618089,
+      "grad_norm": 0.3418284407905513,
+      "learning_rate": 1.8549395415270664e-05,
+      "loss": 0.1658,
+      "step": 350
+    },
+    {
+      "epoch": 0.5140974002196997,
+      "grad_norm": 0.28587381220460634,
+      "learning_rate": 1.8536085980537374e-05,
+      "loss": 0.1521,
+      "step": 351
+    },
+    {
+      "epoch": 0.5155620651775906,
+      "grad_norm": 0.28655840819303974,
+      "learning_rate": 1.8522720586904758e-05,
+      "loss": 0.177,
+      "step": 352
+    },
+    {
+      "epoch": 0.5170267301354815,
+      "grad_norm": 0.3101035018861679,
+      "learning_rate": 1.8509299321990583e-05,
+      "loss": 0.1728,
+      "step": 353
+    },
+    {
+      "epoch": 0.5184913950933724,
+      "grad_norm": 0.309153206811744,
+      "learning_rate": 1.8495822273778867e-05,
+      "loss": 0.1853,
+      "step": 354
+    },
+    {
+      "epoch": 0.5199560600512633,
+      "grad_norm": 0.3072944032463391,
+      "learning_rate": 1.8482289530619332e-05,
+      "loss": 0.1994,
+      "step": 355
+    },
+    {
+      "epoch": 0.5214207250091542,
+      "grad_norm": 0.2882455459923482,
+      "learning_rate": 1.8468701181226803e-05,
+      "loss": 0.1719,
+      "step": 356
+    },
+    {
+      "epoch": 0.522885389967045,
+      "grad_norm": 0.3009687596348506,
+      "learning_rate": 1.8455057314680646e-05,
+      "loss": 0.1762,
+      "step": 357
+    },
+    {
+      "epoch": 0.5243500549249359,
+      "grad_norm": 0.28709132379458197,
+      "learning_rate": 1.8441358020424168e-05,
+      "loss": 0.1872,
+      "step": 358
+    },
+    {
+      "epoch": 0.5258147198828268,
+      "grad_norm": 0.3252839006016281,
+      "learning_rate": 1.8427603388264027e-05,
+      "loss": 0.1658,
+      "step": 359
+    },
+    {
+      "epoch": 0.5272793848407177,
+      "grad_norm": 0.2794496573065983,
+      "learning_rate": 1.8413793508369667e-05,
+      "loss": 0.1536,
+      "step": 360
+    },
+    {
+      "epoch": 0.5287440497986086,
+      "grad_norm": 0.295826215328857,
+      "learning_rate": 1.839992847127271e-05,
+      "loss": 0.1728,
+      "step": 361
+    },
+    {
+      "epoch": 0.5302087147564994,
+      "grad_norm": 0.3057004431965132,
+      "learning_rate": 1.838600836786635e-05,
+      "loss": 0.1737,
+      "step": 362
+    },
+    {
+      "epoch": 0.5316733797143903,
+      "grad_norm": 0.3032546345594625,
+      "learning_rate": 1.8372033289404795e-05,
+      "loss": 0.1804,
+      "step": 363
+    },
+    {
+      "epoch": 0.5331380446722812,
+      "grad_norm": 0.31826939348019884,
+      "learning_rate": 1.835800332750263e-05,
+      "loss": 0.1567,
+      "step": 364
+    },
+    {
+      "epoch": 0.5346027096301721,
+      "grad_norm": 0.3382986306853968,
+      "learning_rate": 1.834391857413423e-05,
+      "loss": 0.1857,
+      "step": 365
+    },
+    {
+      "epoch": 0.536067374588063,
+      "grad_norm": 0.29281271654639596,
+      "learning_rate": 1.8329779121633177e-05,
+      "loss": 0.1755,
+      "step": 366
+    },
+    {
+      "epoch": 0.5375320395459539,
+      "grad_norm": 0.32331356405290573,
+      "learning_rate": 1.8315585062691616e-05,
+      "loss": 0.174,
+      "step": 367
+    },
+    {
+      "epoch": 0.5389967045038447,
+      "grad_norm": 0.3058216598921902,
+      "learning_rate": 1.830133649035968e-05,
+      "loss": 0.1743,
+      "step": 368
+    },
+    {
+      "epoch": 0.5404613694617356,
+      "grad_norm": 0.3067354134944423,
+      "learning_rate": 1.828703349804487e-05,
+      "loss": 0.1715,
+      "step": 369
+    },
+    {
+      "epoch": 0.5419260344196265,
+      "grad_norm": 0.36661193734735914,
+      "learning_rate": 1.8272676179511428e-05,
+      "loss": 0.1738,
+      "step": 370
+    },
+    {
+      "epoch": 0.5433906993775174,
+      "grad_norm": 0.30007784767963736,
+      "learning_rate": 1.8258264628879753e-05,
+      "loss": 0.1552,
+      "step": 371
+    },
+    {
+      "epoch": 0.5448553643354083,
+      "grad_norm": 0.3058776353752914,
+      "learning_rate": 1.8243798940625752e-05,
+      "loss": 0.1656,
+      "step": 372
+    },
+    {
+      "epoch": 0.5463200292932991,
+      "grad_norm": 0.2906329929557275,
+      "learning_rate": 1.8229279209580245e-05,
+      "loss": 0.1724,
+      "step": 373
+    },
+    {
+      "epoch": 0.54778469425119,
+      "grad_norm": 0.2764957936178043,
+      "learning_rate": 1.8214705530928322e-05,
+      "loss": 0.1569,
+      "step": 374
+    },
+    {
+      "epoch": 0.5492493592090809,
+      "grad_norm": 0.31882196921773603,
+      "learning_rate": 1.8200078000208745e-05,
+      "loss": 0.1618,
+      "step": 375
+    },
+    {
+      "epoch": 0.5507140241669718,
+      "grad_norm": 0.3031929350492432,
+      "learning_rate": 1.818539671331329e-05,
+      "loss": 0.1821,
+      "step": 376
+    },
+    {
+      "epoch": 0.5521786891248627,
+      "grad_norm": 0.3285177453767559,
+      "learning_rate": 1.8170661766486147e-05,
+      "loss": 0.1708,
+      "step": 377
+    },
+    {
+      "epoch": 0.5536433540827536,
+      "grad_norm": 0.291362505718777,
+      "learning_rate": 1.815587325632328e-05,
+      "loss": 0.1678,
+      "step": 378
+    },
+    {
+      "epoch": 0.5551080190406444,
+      "grad_norm": 0.3154222441804908,
+      "learning_rate": 1.8141031279771777e-05,
+      "loss": 0.1712,
+      "step": 379
+    },
+    {
+      "epoch": 0.5565726839985353,
+      "grad_norm": 0.2924737445300942,
+      "learning_rate": 1.812613593412924e-05,
+      "loss": 0.1623,
+      "step": 380
+    },
+    {
+      "epoch": 0.5580373489564262,
+      "grad_norm": 0.3327921961518248,
+      "learning_rate": 1.8111187317043136e-05,
+      "loss": 0.168,
+      "step": 381
+    },
+    {
+      "epoch": 0.5595020139143171,
+      "grad_norm": 0.3100694105568342,
+      "learning_rate": 1.8096185526510154e-05,
+      "loss": 0.1457,
+      "step": 382
+    },
+    {
+      "epoch": 0.560966678872208,
+      "grad_norm": 0.31380150144375096,
+      "learning_rate": 1.8081130660875557e-05,
+      "loss": 0.1712,
+      "step": 383
+    },
+    {
+      "epoch": 0.5624313438300989,
+      "grad_norm": 0.29763863880909497,
+      "learning_rate": 1.8066022818832564e-05,
+      "loss": 0.1712,
+      "step": 384
+    },
+    {
+      "epoch": 0.5638960087879897,
+      "grad_norm": 0.31673229457527147,
+      "learning_rate": 1.805086209942166e-05,
+      "loss": 0.1642,
+      "step": 385
+    },
+    {
+      "epoch": 0.5653606737458806,
+      "grad_norm": 0.28416394082082747,
+      "learning_rate": 1.8035648602029997e-05,
+      "loss": 0.1568,
+      "step": 386
+    },
+    {
+      "epoch": 0.5668253387037715,
+      "grad_norm": 0.3118223455455188,
+      "learning_rate": 1.80203824263907e-05,
+      "loss": 0.2036,
+      "step": 387
+    },
+    {
+      "epoch": 0.5682900036616624,
+      "grad_norm": 0.31436408212659767,
+      "learning_rate": 1.8005063672582236e-05,
+      "loss": 0.1761,
+      "step": 388
+    },
+    {
+      "epoch": 0.5697546686195533,
+      "grad_norm": 0.30744846641916707,
+      "learning_rate": 1.7989692441027744e-05,
+      "loss": 0.153,
+      "step": 389
+    },
+    {
+      "epoch": 0.5712193335774441,
+      "grad_norm": 0.29045516723158216,
+      "learning_rate": 1.7974268832494397e-05,
+      "loss": 0.1449,
+      "step": 390
+    },
+    {
+      "epoch": 0.572683998535335,
+      "grad_norm": 0.2752010942760226,
+      "learning_rate": 1.7958792948092726e-05,
+      "loss": 0.181,
+      "step": 391
+    },
+    {
+      "epoch": 0.5741486634932259,
+      "grad_norm": 0.3004642528909644,
+      "learning_rate": 1.7943264889275944e-05,
+      "loss": 0.167,
+      "step": 392
+    },
+    {
+      "epoch": 0.5756133284511168,
+      "grad_norm": 0.28533925990840686,
+      "learning_rate": 1.792768475783932e-05,
+      "loss": 0.1508,
+      "step": 393
+    },
+    {
+      "epoch": 0.5770779934090077,
+      "grad_norm": 0.3074271645217073,
+      "learning_rate": 1.7912052655919478e-05,
+      "loss": 0.1614,
+      "step": 394
+    },
+    {
+      "epoch": 0.5785426583668986,
+      "grad_norm": 0.2860515134927835,
+      "learning_rate": 1.7896368685993738e-05,
+      "loss": 0.1467,
+      "step": 395
+    },
+    {
+      "epoch": 0.5800073233247894,
+      "grad_norm": 0.2826797056983111,
+      "learning_rate": 1.7880632950879438e-05,
+      "loss": 0.1647,
+      "step": 396
+    },
+    {
+      "epoch": 0.5814719882826803,
+      "grad_norm": 0.28758395090779254,
+      "learning_rate": 1.7864845553733276e-05,
+      "loss": 0.1643,
+      "step": 397
+    },
+    {
+      "epoch": 0.5829366532405712,
+      "grad_norm": 0.29461983196821356,
+      "learning_rate": 1.7849006598050626e-05,
+      "loss": 0.1762,
+      "step": 398
+    },
+    {
+      "epoch": 0.5844013181984621,
+      "grad_norm": 0.2777593396108342,
+      "learning_rate": 1.7833116187664846e-05,
+      "loss": 0.1561,
+      "step": 399
+    },
+    {
+      "epoch": 0.585865983156353,
+      "grad_norm": 0.30600321149552173,
+      "learning_rate": 1.781717442674662e-05,
+      "loss": 0.1777,
+      "step": 400
+    },
+    {
+      "epoch": 0.5873306481142438,
+      "grad_norm": 0.2758705111537229,
+      "learning_rate": 1.7801181419803257e-05,
+      "loss": 0.1556,
+      "step": 401
+    },
+    {
+      "epoch": 0.5887953130721347,
+      "grad_norm": 0.2625670019142076,
+      "learning_rate": 1.7785137271678013e-05,
+      "loss": 0.1468,
+      "step": 402
+    },
+    {
+      "epoch": 0.5902599780300256,
+      "grad_norm": 0.29391559042563303,
+      "learning_rate": 1.776904208754941e-05,
+      "loss": 0.1823,
+      "step": 403
+    },
+    {
+      "epoch": 0.5917246429879165,
+      "grad_norm": 0.2824926653515285,
+      "learning_rate": 1.7752895972930538e-05,
+      "loss": 0.1638,
+      "step": 404
+    },
+    {
+      "epoch": 0.5931893079458074,
+      "grad_norm": 0.31289373581775576,
+      "learning_rate": 1.7736699033668353e-05,
+      "loss": 0.1969,
+      "step": 405
+    },
+    {
+      "epoch": 0.5946539729036983,
+      "grad_norm": 0.28541795717364044,
+      "learning_rate": 1.772045137594301e-05,
+      "loss": 0.1445,
+      "step": 406
+    },
+    {
+      "epoch": 0.5961186378615891,
+      "grad_norm": 0.30005077280176484,
+      "learning_rate": 1.770415310626715e-05,
+      "loss": 0.188,
+      "step": 407
+    },
+    {
+      "epoch": 0.59758330281948,
+      "grad_norm": 0.27829833597465525,
+      "learning_rate": 1.7687804331485203e-05,
+      "loss": 0.196,
+      "step": 408
+    },
+    {
+      "epoch": 0.5990479677773709,
+      "grad_norm": 0.28631318068873735,
+      "learning_rate": 1.7671405158772686e-05,
+      "loss": 0.1624,
+      "step": 409
+    },
+    {
+      "epoch": 0.6005126327352618,
+      "grad_norm": 0.2867469013067658,
+      "learning_rate": 1.7654955695635498e-05,
+      "loss": 0.1839,
+      "step": 410
+    },
+    {
+      "epoch": 0.6019772976931527,
+      "grad_norm": 0.2737552258396289,
+      "learning_rate": 1.7638456049909238e-05,
+      "loss": 0.1553,
+      "step": 411
+    },
+    {
+      "epoch": 0.6034419626510436,
+      "grad_norm": 0.3136051529335448,
+      "learning_rate": 1.7621906329758466e-05,
+      "loss": 0.175,
+      "step": 412
+    },
+    {
+      "epoch": 0.6049066276089344,
+      "grad_norm": 0.2773613881722616,
+      "learning_rate": 1.760530664367601e-05,
+      "loss": 0.1696,
+      "step": 413
+    },
+    {
+      "epoch": 0.6063712925668253,
+      "grad_norm": 0.29022770109259144,
+      "learning_rate": 1.758865710048225e-05,
+      "loss": 0.163,
+      "step": 414
+    },
+    {
+      "epoch": 0.6078359575247162,
+      "grad_norm": 0.28813354195965973,
+      "learning_rate": 1.7571957809324422e-05,
+      "loss": 0.1649,
+      "step": 415
+    },
+    {
+      "epoch": 0.6093006224826071,
+      "grad_norm": 0.2985551015333967,
+      "learning_rate": 1.7555208879675875e-05,
+      "loss": 0.1737,
+      "step": 416
+    },
+    {
+      "epoch": 0.610765287440498,
+      "grad_norm": 0.31633838685259236,
+      "learning_rate": 1.7538410421335373e-05,
+      "loss": 0.1606,
+      "step": 417
+    },
+    {
+      "epoch": 0.6122299523983888,
+      "grad_norm": 0.29552853603848234,
+      "learning_rate": 1.752156254442636e-05,
+      "loss": 0.1824,
+      "step": 418
+    },
+    {
+      "epoch": 0.6136946173562797,
+      "grad_norm": 0.31144800745719076,
+      "learning_rate": 1.7504665359396255e-05,
+      "loss": 0.1501,
+      "step": 419
+    },
+    {
+      "epoch": 0.6151592823141706,
+      "grad_norm": 0.30007771034473985,
+      "learning_rate": 1.748771897701572e-05,
+      "loss": 0.1756,
+      "step": 420
+    },
+    {
+      "epoch": 0.6166239472720615,
+      "grad_norm": 0.3153299684246117,
+      "learning_rate": 1.7470723508377935e-05,
+      "loss": 0.1727,
+      "step": 421
+    },
+    {
+      "epoch": 0.6180886122299524,
+      "grad_norm": 0.29874905779465544,
+      "learning_rate": 1.745367906489786e-05,
+      "loss": 0.1583,
+      "step": 422
+    },
+    {
+      "epoch": 0.6195532771878433,
+      "grad_norm": 0.313158860779259,
+      "learning_rate": 1.7436585758311512e-05,
+      "loss": 0.1776,
+      "step": 423
+    },
+    {
+      "epoch": 0.6210179421457341,
+      "grad_norm": 0.29027418449570486,
+      "learning_rate": 1.7419443700675248e-05,
+      "loss": 0.1615,
+      "step": 424
+    },
+    {
+      "epoch": 0.622482607103625,
+      "grad_norm": 0.27091112505454484,
+      "learning_rate": 1.7402253004365007e-05,
+      "loss": 0.1471,
+      "step": 425
+    },
+    {
+      "epoch": 0.6239472720615159,
+      "grad_norm": 0.31079312606497544,
+      "learning_rate": 1.7385013782075575e-05,
+      "loss": 0.1642,
+      "step": 426
+    },
+    {
+      "epoch": 0.6254119370194068,
+      "grad_norm": 0.3191231075261632,
+      "learning_rate": 1.736772614681987e-05,
+      "loss": 0.1799,
+      "step": 427
+    },
+    {
+      "epoch": 0.6268766019772977,
+      "grad_norm": 0.29107955830633764,
+      "learning_rate": 1.7350390211928167e-05,
+      "loss": 0.1364,
+      "step": 428
+    },
+    {
+      "epoch": 0.6283412669351885,
+      "grad_norm": 0.28799608478930533,
+      "learning_rate": 1.7333006091047386e-05,
+      "loss": 0.1446,
+      "step": 429
+    },
+    {
+      "epoch": 0.6298059318930794,
+      "grad_norm": 0.2805703556245509,
+      "learning_rate": 1.7315573898140324e-05,
+      "loss": 0.1578,
+      "step": 430
+    },
+    {
+      "epoch": 0.6312705968509703,
+      "grad_norm": 0.3022531763710714,
+      "learning_rate": 1.7298093747484923e-05,
+      "loss": 0.1454,
+      "step": 431
+    },
+    {
+      "epoch": 0.6327352618088612,
+      "grad_norm": 0.29056765252531186,
+      "learning_rate": 1.7280565753673517e-05,
+      "loss": 0.1409,
+      "step": 432
+    },
+    {
+      "epoch": 0.6341999267667521,
+      "grad_norm": 0.28348330072419187,
+      "learning_rate": 1.7262990031612072e-05,
+      "loss": 0.1632,
+      "step": 433
+    },
+    {
+      "epoch": 0.635664591724643,
+      "grad_norm": 0.27849254820160685,
+      "learning_rate": 1.7245366696519448e-05,
+      "loss": 0.1382,
+      "step": 434
+    },
+    {
+      "epoch": 0.6371292566825338,
+      "grad_norm": 0.2849173212539919,
+      "learning_rate": 1.7227695863926627e-05,
+      "loss": 0.1404,
+      "step": 435
+    },
+    {
+      "epoch": 0.6385939216404247,
+      "grad_norm": 0.29763893191664237,
+      "learning_rate": 1.7209977649675975e-05,
+      "loss": 0.1699,
+      "step": 436
+    },
+    {
+      "epoch": 0.6400585865983156,
+      "grad_norm": 0.28500529876174013,
+      "learning_rate": 1.7192212169920458e-05,
+      "loss": 0.154,
+      "step": 437
+    },
+    {
+      "epoch": 0.6415232515562065,
+      "grad_norm": 0.29341182785229564,
+      "learning_rate": 1.717439954112291e-05,
+      "loss": 0.1629,
+      "step": 438
+    },
+    {
+      "epoch": 0.6429879165140974,
+      "grad_norm": 0.2833467699130664,
+      "learning_rate": 1.7156539880055236e-05,
+      "loss": 0.1479,
+      "step": 439
+    },
+    {
+      "epoch": 0.6444525814719883,
+      "grad_norm": 0.29432333457804116,
+      "learning_rate": 1.7138633303797676e-05,
+      "loss": 0.1447,
+      "step": 440
+    },
+    {
+      "epoch": 0.6459172464298791,
+      "grad_norm": 0.2899857698207261,
+      "learning_rate": 1.712067992973803e-05,
+      "loss": 0.1707,
+      "step": 441
+    },
+    {
+      "epoch": 0.64738191138777,
+      "grad_norm": 0.300112032266628,
+      "learning_rate": 1.710267987557087e-05,
+      "loss": 0.1441,
+      "step": 442
+    },
+    {
+      "epoch": 0.6488465763456609,
+      "grad_norm": 0.28676504299001276,
+      "learning_rate": 1.7084633259296798e-05,
+      "loss": 0.1418,
+      "step": 443
+    },
+    {
+      "epoch": 0.6503112413035518,
+      "grad_norm": 0.27688806374870634,
+      "learning_rate": 1.706654019922164e-05,
+      "loss": 0.1478,
+      "step": 444
+    },
+    {
+      "epoch": 0.6517759062614427,
+      "grad_norm": 0.2853894260456318,
+      "learning_rate": 1.704840081395571e-05,
+      "loss": 0.1524,
+      "step": 445
+    },
+    {
+      "epoch": 0.6532405712193335,
+      "grad_norm": 0.2946623888589574,
+      "learning_rate": 1.703021522241298e-05,
+      "loss": 0.1471,
+      "step": 446
+    },
+    {
+      "epoch": 0.6547052361772244,
+      "grad_norm": 0.2726573492109173,
+      "learning_rate": 1.701198354381036e-05,
+      "loss": 0.1312,
+      "step": 447
+    },
+    {
+      "epoch": 0.6561699011351153,
+      "grad_norm": 0.28410262500836964,
+      "learning_rate": 1.6993705897666873e-05,
+      "loss": 0.1444,
+      "step": 448
+    },
+    {
+      "epoch": 0.6576345660930062,
+      "grad_norm": 0.2833743474032485,
+      "learning_rate": 1.697538240380288e-05,
+      "loss": 0.1596,
+      "step": 449
+    },
+    {
+      "epoch": 0.6590992310508971,
+      "grad_norm": 0.28744248608886525,
+      "learning_rate": 1.695701318233931e-05,
+      "loss": 0.1696,
+      "step": 450
+    },
+    {
+      "epoch": 0.660563896008788,
+      "grad_norm": 0.2922684857852237,
+      "learning_rate": 1.6938598353696864e-05,
+      "loss": 0.1822,
+      "step": 451
+    },
+    {
+      "epoch": 0.6620285609666788,
+      "grad_norm": 0.2824755616354114,
+      "learning_rate": 1.6920138038595214e-05,
+      "loss": 0.16,
+      "step": 452
+    },
+    {
+      "epoch": 0.6634932259245697,
+      "grad_norm": 0.2881779281986939,
+      "learning_rate": 1.6901632358052226e-05,
+      "loss": 0.1306,
+      "step": 453
+    },
+    {
+      "epoch": 0.6649578908824606,
+      "grad_norm": 0.3076676560083059,
+      "learning_rate": 1.6883081433383163e-05,
+      "loss": 0.1693,
+      "step": 454
+    },
+    {
+      "epoch": 0.6664225558403515,
+      "grad_norm": 0.28360460469529464,
+      "learning_rate": 1.6864485386199895e-05,
+      "loss": 0.1347,
+      "step": 455
+    },
+    {
+      "epoch": 0.6678872207982424,
+      "grad_norm": 0.2792514596107041,
+      "learning_rate": 1.6845844338410077e-05,
+      "loss": 0.1584,
+      "step": 456
+    },
+    {
+      "epoch": 0.6693518857561332,
+      "grad_norm": 0.27734844058934455,
+      "learning_rate": 1.6827158412216396e-05,
+      "loss": 0.1489,
+      "step": 457
+    },
+    {
+      "epoch": 0.6708165507140241,
+      "grad_norm": 0.2868505259889296,
+      "learning_rate": 1.6808427730115716e-05,
+      "loss": 0.1556,
+      "step": 458
+    },
+    {
+      "epoch": 0.672281215671915,
+      "grad_norm": 0.27438002988710974,
+      "learning_rate": 1.6789652414898315e-05,
+      "loss": 0.1594,
+      "step": 459
+    },
+    {
+      "epoch": 0.6737458806298059,
+      "grad_norm": 0.26963685143388444,
+      "learning_rate": 1.677083258964707e-05,
+      "loss": 0.1544,
+      "step": 460
+    },
+    {
+      "epoch": 0.6752105455876968,
+      "grad_norm": 0.27489622544391323,
+      "learning_rate": 1.675196837773664e-05,
+      "loss": 0.1779,
+      "step": 461
+    },
+    {
+      "epoch": 0.6766752105455877,
+      "grad_norm": 0.28666936527106884,
+      "learning_rate": 1.673305990283266e-05,
+      "loss": 0.1389,
+      "step": 462
+    },
+    {
+      "epoch": 0.6781398755034785,
+      "grad_norm": 0.2854502743190489,
+      "learning_rate": 1.6714107288890943e-05,
+      "loss": 0.1518,
+      "step": 463
+    },
+    {
+      "epoch": 0.6796045404613694,
+      "grad_norm": 0.28486649873799424,
+      "learning_rate": 1.6695110660156652e-05,
+      "loss": 0.1356,
+      "step": 464
+    },
+    {
+      "epoch": 0.6810692054192603,
+      "grad_norm": 0.3095642301969388,
+      "learning_rate": 1.6676070141163498e-05,
+      "loss": 0.1676,
+      "step": 465
+    },
+    {
+      "epoch": 0.6825338703771512,
+      "grad_norm": 0.3135455510235421,
+      "learning_rate": 1.665698585673291e-05,
+      "loss": 0.1606,
+      "step": 466
+    },
+    {
+      "epoch": 0.6839985353350421,
+      "grad_norm": 0.28266851700362705,
+      "learning_rate": 1.6637857931973233e-05,
+      "loss": 0.1345,
+      "step": 467
+    },
+    {
+      "epoch": 0.685463200292933,
+      "grad_norm": 0.27779394936697266,
+      "learning_rate": 1.6618686492278892e-05,
+      "loss": 0.1682,
+      "step": 468
+    },
+    {
+      "epoch": 0.6869278652508238,
+      "grad_norm": 0.27567920609962066,
+      "learning_rate": 1.6599471663329577e-05,
+      "loss": 0.1334,
+      "step": 469
+    },
+    {
+      "epoch": 0.6883925302087147,
+      "grad_norm": 0.2754295653847063,
+      "learning_rate": 1.6580213571089427e-05,
+      "loss": 0.1573,
+      "step": 470
+    },
+    {
+      "epoch": 0.6898571951666056,
+      "grad_norm": 0.2750050531478153,
+      "learning_rate": 1.656091234180619e-05,
+      "loss": 0.1773,
+      "step": 471
+    },
+    {
+      "epoch": 0.6913218601244965,
+      "grad_norm": 0.27965457420033163,
+      "learning_rate": 1.65415681020104e-05,
+      "loss": 0.1369,
+      "step": 472
+    },
+    {
+      "epoch": 0.6927865250823874,
+      "grad_norm": 0.27090231974844964,
+      "learning_rate": 1.6522180978514556e-05,
+      "loss": 0.1375,
+      "step": 473
+    },
+    {
+      "epoch": 0.6942511900402782,
+      "grad_norm": 0.2824639859361883,
+      "learning_rate": 1.6502751098412282e-05,
+      "loss": 0.1246,
+      "step": 474
+    },
+    {
+      "epoch": 0.6957158549981691,
+      "grad_norm": 0.28655833648300955,
+      "learning_rate": 1.648327858907749e-05,
+      "loss": 0.1648,
+      "step": 475
+    },
+    {
+      "epoch": 0.69718051995606,
+      "grad_norm": 0.27502131490113785,
+      "learning_rate": 1.6463763578163563e-05,
+      "loss": 0.1501,
+      "step": 476
+    },
+    {
+      "epoch": 0.6986451849139509,
+      "grad_norm": 0.26898509925311315,
+      "learning_rate": 1.6444206193602493e-05,
+      "loss": 0.1499,
+      "step": 477
+    },
+    {
+      "epoch": 0.7001098498718418,
+      "grad_norm": 0.3350314192757342,
+      "learning_rate": 1.642460656360406e-05,
+      "loss": 0.1572,
+      "step": 478
+    },
+    {
+      "epoch": 0.7015745148297327,
+      "grad_norm": 0.2773978701354445,
+      "learning_rate": 1.6404964816654993e-05,
+      "loss": 0.1272,
+      "step": 479
+    },
+    {
+      "epoch": 0.7030391797876235,
+      "grad_norm": 0.2901786822045925,
+      "learning_rate": 1.638528108151811e-05,
+      "loss": 0.1892,
+      "step": 480
+    },
+    {
+      "epoch": 0.7045038447455144,
+      "grad_norm": 0.2847317225297176,
+      "learning_rate": 1.63655554872315e-05,
+      "loss": 0.1579,
+      "step": 481
+    },
+    {
+      "epoch": 0.7059685097034053,
+      "grad_norm": 0.2760852156995962,
+      "learning_rate": 1.6345788163107645e-05,
+      "loss": 0.1411,
+      "step": 482
+    },
+    {
+      "epoch": 0.7074331746612962,
+      "grad_norm": 0.2812021526710764,
+      "learning_rate": 1.6325979238732606e-05,
+      "loss": 0.1416,
+      "step": 483
+    },
+    {
+      "epoch": 0.7088978396191871,
+      "grad_norm": 0.27495150899008586,
+      "learning_rate": 1.630612884396515e-05,
+      "loss": 0.1457,
+      "step": 484
+    },
+    {
+      "epoch": 0.710362504577078,
+      "grad_norm": 0.27821852042543493,
+      "learning_rate": 1.62862371089359e-05,
+      "loss": 0.1409,
+      "step": 485
+    },
+    {
+      "epoch": 0.7118271695349688,
+      "grad_norm": 0.26154961742164873,
+      "learning_rate": 1.6266304164046505e-05,
+      "loss": 0.1392,
+      "step": 486
+    },
+    {
+      "epoch": 0.7132918344928597,
+      "grad_norm": 0.3344656299415514,
+      "learning_rate": 1.6246330139968748e-05,
+      "loss": 0.1395,
+      "step": 487
+    },
+    {
+      "epoch": 0.7147564994507506,
+      "grad_norm": 0.2889353353792984,
+      "learning_rate": 1.6226315167643723e-05,
+      "loss": 0.1377,
+      "step": 488
+    },
+    {
+      "epoch": 0.7162211644086415,
+      "grad_norm": 0.2866205322545156,
+      "learning_rate": 1.6206259378280956e-05,
+      "loss": 0.1339,
+      "step": 489
+    },
+    {
+      "epoch": 0.7176858293665324,
+      "grad_norm": 0.2635137582733423,
+      "learning_rate": 1.6186162903357562e-05,
+      "loss": 0.1384,
+      "step": 490
+    },
+    {
+      "epoch": 0.7191504943244232,
+      "grad_norm": 0.264894099077168,
+      "learning_rate": 1.616602587461736e-05,
+      "loss": 0.1415,
+      "step": 491
+    },
+    {
+      "epoch": 0.7206151592823141,
+      "grad_norm": 0.2912590760990707,
+      "learning_rate": 1.6145848424070032e-05,
+      "loss": 0.1377,
+      "step": 492
+    },
+    {
+      "epoch": 0.722079824240205,
+      "grad_norm": 0.28322985703286563,
+      "learning_rate": 1.612563068399024e-05,
+      "loss": 0.1381,
+      "step": 493
+    },
+    {
+      "epoch": 0.7235444891980959,
+      "grad_norm": 0.2761592201460229,
+      "learning_rate": 1.6105372786916776e-05,
+      "loss": 0.1205,
+      "step": 494
+    },
+    {
+      "epoch": 0.7250091541559868,
+      "grad_norm": 0.2593621751501278,
+      "learning_rate": 1.6085074865651672e-05,
+      "loss": 0.1396,
+      "step": 495
+    },
+    {
+      "epoch": 0.7264738191138777,
+      "grad_norm": 0.2698751624804496,
+      "learning_rate": 1.6064737053259355e-05,
+      "loss": 0.1549,
+      "step": 496
+    },
+    {
+      "epoch": 0.7279384840717685,
+      "grad_norm": 0.2784921841300199,
+      "learning_rate": 1.604435948306575e-05,
+      "loss": 0.142,
+      "step": 497
+    },
+    {
+      "epoch": 0.7294031490296594,
+      "grad_norm": 0.282990933381995,
+      "learning_rate": 1.6023942288657423e-05,
+      "loss": 0.1532,
+      "step": 498
+    },
+    {
+      "epoch": 0.7308678139875503,
+      "grad_norm": 0.2810338678318928,
+      "learning_rate": 1.60034856038807e-05,
+      "loss": 0.1379,
+      "step": 499
+    },
+    {
+      "epoch": 0.7323324789454412,
+      "grad_norm": 0.2964732663844032,
+      "learning_rate": 1.5982989562840785e-05,
+      "loss": 0.1324,
+      "step": 500
+    },
+    {
+      "epoch": 0.7337971439033321,
+      "grad_norm": 0.2943783008884028,
+      "learning_rate": 1.596245429990088e-05,
+      "loss": 0.143,
+      "step": 501
+    },
+    {
+      "epoch": 0.7352618088612229,
+      "grad_norm": 0.292096480725846,
+      "learning_rate": 1.5941879949681323e-05,
+      "loss": 0.1466,
+      "step": 502
+    },
+    {
+      "epoch": 0.7367264738191138,
+      "grad_norm": 0.28475675120953237,
+      "learning_rate": 1.5921266647058683e-05,
+      "loss": 0.1573,
+      "step": 503
+    },
+    {
+      "epoch": 0.7381911387770047,
+      "grad_norm": 0.2741769477206498,
+      "learning_rate": 1.5900614527164876e-05,
+      "loss": 0.1485,
+      "step": 504
+    },
+    {
+      "epoch": 0.7396558037348956,
+      "grad_norm": 0.3040355479807483,
+      "learning_rate": 1.5879923725386307e-05,
+      "loss": 0.1642,
+      "step": 505
+    },
+    {
+      "epoch": 0.7411204686927865,
+      "grad_norm": 0.3037335266933942,
+      "learning_rate": 1.5859194377362942e-05,
+      "loss": 0.1497,
+      "step": 506
+    },
+    {
+      "epoch": 0.7425851336506774,
+      "grad_norm": 0.2791249180827828,
+      "learning_rate": 1.5838426618987455e-05,
+      "loss": 0.1521,
+      "step": 507
+    },
+    {
+      "epoch": 0.7440497986085682,
+      "grad_norm": 0.281309610689647,
+      "learning_rate": 1.5817620586404315e-05,
+      "loss": 0.1346,
+      "step": 508
+    },
+    {
+      "epoch": 0.7455144635664591,
+      "grad_norm": 0.28553330164537244,
+      "learning_rate": 1.5796776416008897e-05,
+      "loss": 0.1446,
+      "step": 509
+    },
+    {
+      "epoch": 0.74697912852435,
+      "grad_norm": 0.2844881306714679,
+      "learning_rate": 1.5775894244446603e-05,
+      "loss": 0.1414,
+      "step": 510
+    },
+    {
+      "epoch": 0.7484437934822409,
+      "grad_norm": 0.268988666507586,
+      "learning_rate": 1.575497420861194e-05,
+      "loss": 0.1288,
+      "step": 511
+    },
+    {
+      "epoch": 0.7499084584401318,
+      "grad_norm": 0.27447838457593043,
+      "learning_rate": 1.573401644564764e-05,
+      "loss": 0.1572,
+      "step": 512
+    },
+    {
+      "epoch": 0.7513731233980228,
+      "grad_norm": 0.2973833246231482,
+      "learning_rate": 1.571302109294377e-05,
+      "loss": 0.1406,
+      "step": 513
+    },
+    {
+      "epoch": 0.7528377883559136,
+      "grad_norm": 0.27289540026887815,
+      "learning_rate": 1.569198828813681e-05,
+      "loss": 0.1392,
+      "step": 514
+    },
+    {
+      "epoch": 0.7543024533138045,
+      "grad_norm": 0.2870617946282251,
+      "learning_rate": 1.567091816910875e-05,
+      "loss": 0.1543,
+      "step": 515
+    },
+    {
+      "epoch": 0.7557671182716954,
+      "grad_norm": 0.27283398029853356,
+      "learning_rate": 1.5649810873986214e-05,
+      "loss": 0.1276,
+      "step": 516
+    },
+    {
+      "epoch": 0.7572317832295863,
+      "grad_norm": 0.2637139762799822,
+      "learning_rate": 1.5628666541139523e-05,
+      "loss": 0.1355,
+      "step": 517
+    },
+    {
+      "epoch": 0.7586964481874772,
+      "grad_norm": 0.27854483215025555,
+      "learning_rate": 1.5607485309181813e-05,
+      "loss": 0.1483,
+      "step": 518
+    },
+    {
+      "epoch": 0.760161113145368,
+      "grad_norm": 0.28494692420422013,
+      "learning_rate": 1.55862673169681e-05,
+      "loss": 0.1356,
+      "step": 519
+    },
+    {
+      "epoch": 0.7616257781032589,
+      "grad_norm": 0.2898114191567579,
+      "learning_rate": 1.5565012703594403e-05,
+      "loss": 0.141,
+      "step": 520
+    },
+    {
+      "epoch": 0.7630904430611498,
+      "grad_norm": 0.2822142264481914,
+      "learning_rate": 1.55437216083968e-05,
+      "loss": 0.1311,
+      "step": 521
+    },
+    {
+      "epoch": 0.7645551080190407,
+      "grad_norm": 0.25757582269462975,
+      "learning_rate": 1.552239417095052e-05,
+      "loss": 0.1361,
+      "step": 522
+    },
+    {
+      "epoch": 0.7660197729769316,
+      "grad_norm": 0.2694585038356422,
+      "learning_rate": 1.5501030531069066e-05,
+      "loss": 0.1405,
+      "step": 523
+    },
+    {
+      "epoch": 0.7674844379348225,
+      "grad_norm": 0.27818748796170395,
+      "learning_rate": 1.5479630828803235e-05,
+      "loss": 0.1414,
+      "step": 524
+    },
+    {
+      "epoch": 0.7689491028927133,
+      "grad_norm": 0.2892622203178171,
+      "learning_rate": 1.5458195204440255e-05,
+      "loss": 0.1387,
+      "step": 525
+    },
+    {
+      "epoch": 0.7704137678506042,
+      "grad_norm": 0.2765799688074329,
+      "learning_rate": 1.5436723798502842e-05,
+      "loss": 0.1453,
+      "step": 526
+    },
+    {
+      "epoch": 0.7718784328084951,
+      "grad_norm": 0.27743642550569964,
+      "learning_rate": 1.5415216751748264e-05,
+      "loss": 0.1326,
+      "step": 527
+    },
+    {
+      "epoch": 0.773343097766386,
+      "grad_norm": 0.2602739542105575,
+      "learning_rate": 1.5393674205167453e-05,
+      "loss": 0.1329,
+      "step": 528
+    },
+    {
+      "epoch": 0.7748077627242769,
+      "grad_norm": 0.2673531017304055,
+      "learning_rate": 1.5372096299984064e-05,
+      "loss": 0.1343,
+      "step": 529
+    },
+    {
+      "epoch": 0.7762724276821678,
+      "grad_norm": 0.29845215607643555,
+      "learning_rate": 1.5350483177653528e-05,
+      "loss": 0.1522,
+      "step": 530
+    },
+    {
+      "epoch": 0.7777370926400586,
+      "grad_norm": 0.2725973305117268,
+      "learning_rate": 1.5328834979862158e-05,
+      "loss": 0.13,
+      "step": 531
+    },
+    {
+      "epoch": 0.7792017575979495,
+      "grad_norm": 0.263213840818661,
+      "learning_rate": 1.5307151848526213e-05,
+      "loss": 0.1246,
+      "step": 532
+    },
+    {
+      "epoch": 0.7806664225558404,
+      "grad_norm": 0.2551375025943304,
+      "learning_rate": 1.5285433925790946e-05,
+      "loss": 0.1111,
+      "step": 533
+    },
+    {
+      "epoch": 0.7821310875137313,
+      "grad_norm": 0.2622331657629095,
+      "learning_rate": 1.5263681354029694e-05,
+      "loss": 0.1351,
+      "step": 534
+    },
+    {
+      "epoch": 0.7835957524716222,
+      "grad_norm": 0.3124931282937266,
+      "learning_rate": 1.5241894275842946e-05,
+      "loss": 0.1417,
+      "step": 535
+    },
+    {
+      "epoch": 0.785060417429513,
+      "grad_norm": 0.30036829759264483,
+      "learning_rate": 1.5220072834057387e-05,
+      "loss": 0.1363,
+      "step": 536
+    },
+    {
+      "epoch": 0.7865250823874039,
+      "grad_norm": 0.2596328178928325,
+      "learning_rate": 1.5198217171724982e-05,
+      "loss": 0.1214,
+      "step": 537
+    },
+    {
+      "epoch": 0.7879897473452948,
+      "grad_norm": 0.2805919354886981,
+      "learning_rate": 1.5176327432122028e-05,
+      "loss": 0.1268,
+      "step": 538
+    },
+    {
+      "epoch": 0.7894544123031857,
+      "grad_norm": 0.2771437475611675,
+      "learning_rate": 1.5154403758748228e-05,
+      "loss": 0.1308,
+      "step": 539
+    },
+    {
+      "epoch": 0.7909190772610766,
+      "grad_norm": 0.2731861706982954,
+      "learning_rate": 1.5132446295325722e-05,
+      "loss": 0.128,
+      "step": 540
+    },
+    {
+      "epoch": 0.7923837422189675,
+      "grad_norm": 0.3430389886450956,
+      "learning_rate": 1.511045518579818e-05,
+      "loss": 0.1374,
+      "step": 541
+    },
+    {
+      "epoch": 0.7938484071768583,
+      "grad_norm": 0.2905683813596674,
+      "learning_rate": 1.5088430574329836e-05,
+      "loss": 0.1332,
+      "step": 542
+    },
+    {
+      "epoch": 0.7953130721347492,
+      "grad_norm": 0.28971874507270196,
+      "learning_rate": 1.5066372605304537e-05,
+      "loss": 0.1358,
+      "step": 543
+    },
+    {
+      "epoch": 0.7967777370926401,
+      "grad_norm": 0.2629369085736157,
+      "learning_rate": 1.5044281423324826e-05,
+      "loss": 0.1305,
+      "step": 544
+    },
+    {
+      "epoch": 0.798242402050531,
+      "grad_norm": 0.2624966808667796,
+      "learning_rate": 1.5022157173210969e-05,
+      "loss": 0.1386,
+      "step": 545
+    },
+    {
+      "epoch": 0.7997070670084219,
+      "grad_norm": 0.27495723120059107,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.1261,
+      "step": 546
+    },
+    {
+      "epoch": 0.8011717319663128,
+      "grad_norm": 0.2968841863976358,
+      "learning_rate": 1.4977810048944806e-05,
+      "loss": 0.1438,
+      "step": 547
+    },
+    {
+      "epoch": 0.8026363969242036,
+      "grad_norm": 0.26619846720879786,
+      "learning_rate": 1.495558746551313e-05,
+      "loss": 0.139,
+      "step": 548
+    },
+    {
+      "epoch": 0.8041010618820945,
+      "grad_norm": 0.2733899438656217,
+      "learning_rate": 1.4933332395386652e-05,
+      "loss": 0.1394,
+      "step": 549
+    },
+    {
+      "epoch": 0.8055657268399854,
+      "grad_norm": 0.27528673995767655,
+      "learning_rate": 1.4911044984460015e-05,
+      "loss": 0.1235,
+      "step": 550
+    },
+    {
+      "epoch": 0.8070303917978763,
+      "grad_norm": 0.27120553519443596,
+      "learning_rate": 1.4888725378839877e-05,
+      "loss": 0.1389,
+      "step": 551
+    },
+    {
+      "epoch": 0.8084950567557672,
+      "grad_norm": 0.26970265074149985,
+      "learning_rate": 1.4866373724843945e-05,
+      "loss": 0.1381,
+      "step": 552
+    },
+    {
+      "epoch": 0.809959721713658,
+      "grad_norm": 0.29150434298475913,
+      "learning_rate": 1.484399016900003e-05,
+      "loss": 0.1345,
+      "step": 553
+    },
+    {
+      "epoch": 0.8114243866715489,
+      "grad_norm": 0.29587722131690225,
+      "learning_rate": 1.4821574858045073e-05,
+      "loss": 0.1356,
+      "step": 554
+    },
+    {
+      "epoch": 0.8128890516294398,
+      "grad_norm": 0.26978671672533217,
+      "learning_rate": 1.479912793892419e-05,
+      "loss": 0.1363,
+      "step": 555
+    },
+    {
+      "epoch": 0.8143537165873307,
+      "grad_norm": 0.26428779798336755,
+      "learning_rate": 1.4776649558789698e-05,
+      "loss": 0.1545,
+      "step": 556
+    },
+    {
+      "epoch": 0.8158183815452216,
+      "grad_norm": 0.2715765644749667,
+      "learning_rate": 1.475413986500017e-05,
+      "loss": 0.1437,
+      "step": 557
+    },
+    {
+      "epoch": 0.8172830465031125,
+      "grad_norm": 0.2626982755104111,
+      "learning_rate": 1.4731599005119454e-05,
+      "loss": 0.1277,
+      "step": 558
+    },
+    {
+      "epoch": 0.8187477114610033,
+      "grad_norm": 0.27082844640160836,
+      "learning_rate": 1.470902712691571e-05,
+      "loss": 0.1111,
+      "step": 559
+    },
+    {
+      "epoch": 0.8202123764188942,
+      "grad_norm": 0.2668734287792547,
+      "learning_rate": 1.4686424378360434e-05,
+      "loss": 0.1312,
+      "step": 560
+    },
+    {
+      "epoch": 0.8216770413767851,
+      "grad_norm": 0.2869401318531047,
+      "learning_rate": 1.4663790907627502e-05,
+      "loss": 0.1509,
+      "step": 561
+    },
+    {
+      "epoch": 0.823141706334676,
+      "grad_norm": 0.2773427427864892,
+      "learning_rate": 1.4641126863092194e-05,
+      "loss": 0.1577,
+      "step": 562
+    },
+    {
+      "epoch": 0.8246063712925669,
+      "grad_norm": 0.2717284080506168,
+      "learning_rate": 1.4618432393330211e-05,
+      "loss": 0.1254,
+      "step": 563
+    },
+    {
+      "epoch": 0.8260710362504577,
+      "grad_norm": 0.2919980547507266,
+      "learning_rate": 1.4595707647116713e-05,
+      "loss": 0.158,
+      "step": 564
+    },
+    {
+      "epoch": 0.8275357012083486,
+      "grad_norm": 0.27833252973367867,
+      "learning_rate": 1.4572952773425335e-05,
+      "loss": 0.1252,
+      "step": 565
+    },
+    {
+      "epoch": 0.8290003661662395,
+      "grad_norm": 0.2850694697710808,
+      "learning_rate": 1.455016792142722e-05,
+      "loss": 0.1333,
+      "step": 566
+    },
+    {
+      "epoch": 0.8304650311241304,
+      "grad_norm": 0.26558195129076256,
+      "learning_rate": 1.4527353240490039e-05,
+      "loss": 0.1349,
+      "step": 567
+    },
+    {
+      "epoch": 0.8319296960820213,
+      "grad_norm": 0.2944649491485909,
+      "learning_rate": 1.4504508880176996e-05,
+      "loss": 0.1528,
+      "step": 568
+    },
+    {
+      "epoch": 0.8333943610399122,
+      "grad_norm": 0.2578991107569846,
+      "learning_rate": 1.4481634990245871e-05,
+      "loss": 0.1225,
+      "step": 569
+    },
+    {
+      "epoch": 0.834859025997803,
+      "grad_norm": 0.33398387092007636,
+      "learning_rate": 1.4458731720648024e-05,
+      "loss": 0.1459,
+      "step": 570
+    },
+    {
+      "epoch": 0.8363236909556939,
+      "grad_norm": 0.29274576860822454,
+      "learning_rate": 1.4435799221527417e-05,
+      "loss": 0.1377,
+      "step": 571
+    },
+    {
+      "epoch": 0.8377883559135848,
+      "grad_norm": 0.2865214436536306,
+      "learning_rate": 1.4412837643219625e-05,
+      "loss": 0.1416,
+      "step": 572
+    },
+    {
+      "epoch": 0.8392530208714757,
+      "grad_norm": 0.2803403830745175,
+      "learning_rate": 1.4389847136250858e-05,
+      "loss": 0.1291,
+      "step": 573
+    },
+    {
+      "epoch": 0.8407176858293666,
+      "grad_norm": 0.2692461377236704,
+      "learning_rate": 1.4366827851336964e-05,
+      "loss": 0.1251,
+      "step": 574
+    },
+    {
+      "epoch": 0.8421823507872574,
+      "grad_norm": 0.2626470317555678,
+      "learning_rate": 1.4343779939382451e-05,
+      "loss": 0.1111,
+      "step": 575
+    },
+    {
+      "epoch": 0.8436470157451483,
+      "grad_norm": 0.28937945369153595,
+      "learning_rate": 1.4320703551479494e-05,
+      "loss": 0.1305,
+      "step": 576
+    },
+    {
+      "epoch": 0.8451116807030392,
+      "grad_norm": 0.30541326375044037,
+      "learning_rate": 1.4297598838906938e-05,
+      "loss": 0.1502,
+      "step": 577
+    },
+    {
+      "epoch": 0.8465763456609301,
+      "grad_norm": 0.2796697681217199,
+      "learning_rate": 1.4274465953129326e-05,
+      "loss": 0.1236,
+      "step": 578
+    },
+    {
+      "epoch": 0.848041010618821,
+      "grad_norm": 0.2614920930808128,
+      "learning_rate": 1.4251305045795874e-05,
+      "loss": 0.1126,
+      "step": 579
+    },
+    {
+      "epoch": 0.8495056755767119,
+      "grad_norm": 0.28511419468497884,
+      "learning_rate": 1.422811626873951e-05,
+      "loss": 0.1285,
+      "step": 580
+    },
+    {
+      "epoch": 0.8509703405346027,
+      "grad_norm": 0.2707984473087975,
+      "learning_rate": 1.4204899773975855e-05,
+      "loss": 0.1436,
+      "step": 581
+    },
+    {
+      "epoch": 0.8524350054924936,
+      "grad_norm": 0.281582631209825,
+      "learning_rate": 1.4181655713702242e-05,
+      "loss": 0.1297,
+      "step": 582
+    },
+    {
+      "epoch": 0.8538996704503845,
+      "grad_norm": 0.27592043627026036,
+      "learning_rate": 1.4158384240296707e-05,
+      "loss": 0.1255,
+      "step": 583
+    },
+    {
+      "epoch": 0.8553643354082754,
+      "grad_norm": 0.28772169132389885,
+      "learning_rate": 1.4135085506316997e-05,
+      "loss": 0.1504,
+      "step": 584
+    },
+    {
+      "epoch": 0.8568290003661663,
+      "grad_norm": 0.26245407957464945,
+      "learning_rate": 1.4111759664499562e-05,
+      "loss": 0.1131,
+      "step": 585
+    },
+    {
+      "epoch": 0.8582936653240572,
+      "grad_norm": 0.25257520518661797,
+      "learning_rate": 1.4088406867758573e-05,
+      "loss": 0.1235,
+      "step": 586
+    },
+    {
+      "epoch": 0.859758330281948,
+      "grad_norm": 0.26981349513085906,
+      "learning_rate": 1.4065027269184888e-05,
+      "loss": 0.1429,
+      "step": 587
+    },
+    {
+      "epoch": 0.8612229952398389,
+      "grad_norm": 0.2734637683112845,
+      "learning_rate": 1.404162102204508e-05,
+      "loss": 0.129,
+      "step": 588
+    },
+    {
+      "epoch": 0.8626876601977298,
+      "grad_norm": 0.30246081365181193,
+      "learning_rate": 1.4018188279780412e-05,
+      "loss": 0.1314,
+      "step": 589
+    },
+    {
+      "epoch": 0.8641523251556207,
+      "grad_norm": 0.287487018814624,
+      "learning_rate": 1.3994729196005839e-05,
+      "loss": 0.1259,
+      "step": 590
+    },
+    {
+      "epoch": 0.8656169901135116,
+      "grad_norm": 0.27210880904989876,
+      "learning_rate": 1.3971243924508996e-05,
+      "loss": 0.1178,
+      "step": 591
+    },
+    {
+      "epoch": 0.8670816550714024,
+      "grad_norm": 0.29342064785517297,
+      "learning_rate": 1.3947732619249206e-05,
+      "loss": 0.1341,
+      "step": 592
+    },
+    {
+      "epoch": 0.8685463200292933,
+      "grad_norm": 0.27972993450917555,
+      "learning_rate": 1.3924195434356443e-05,
+      "loss": 0.1313,
+      "step": 593
+    },
+    {
+      "epoch": 0.8700109849871842,
+      "grad_norm": 0.24692549028529914,
+      "learning_rate": 1.3900632524130343e-05,
+      "loss": 0.1233,
+      "step": 594
+    },
+    {
+      "epoch": 0.8714756499450751,
+      "grad_norm": 0.32938798959658994,
+      "learning_rate": 1.3877044043039189e-05,
+      "loss": 0.1561,
+      "step": 595
+    },
+    {
+      "epoch": 0.872940314902966,
+      "grad_norm": 0.27405616789978654,
+      "learning_rate": 1.3853430145718892e-05,
+      "loss": 0.1361,
+      "step": 596
+    },
+    {
+      "epoch": 0.8744049798608569,
+      "grad_norm": 0.2846698580131875,
+      "learning_rate": 1.382979098697198e-05,
+      "loss": 0.1505,
+      "step": 597
+    },
+    {
+      "epoch": 0.8758696448187477,
+      "grad_norm": 0.2886024048200631,
+      "learning_rate": 1.3806126721766586e-05,
+      "loss": 0.1421,
+      "step": 598
+    },
+    {
+      "epoch": 0.8773343097766386,
+      "grad_norm": 0.24843406819575536,
+      "learning_rate": 1.378243750523543e-05,
+      "loss": 0.1166,
+      "step": 599
+    },
+    {
+      "epoch": 0.8787989747345295,
+      "grad_norm": 0.29136988530850255,
+      "learning_rate": 1.3758723492674803e-05,
+      "loss": 0.163,
+      "step": 600
+    },
+    {
+      "epoch": 0.8802636396924204,
+      "grad_norm": 0.27987479624274275,
+      "learning_rate": 1.3734984839543547e-05,
+      "loss": 0.1302,
+      "step": 601
+    },
+    {
+      "epoch": 0.8817283046503113,
+      "grad_norm": 0.2766713098966634,
+      "learning_rate": 1.3711221701462037e-05,
+      "loss": 0.1264,
+      "step": 602
+    },
+    {
+      "epoch": 0.8831929696082021,
+      "grad_norm": 0.281327006141709,
+      "learning_rate": 1.368743423421116e-05,
+      "loss": 0.1433,
+      "step": 603
+    },
+    {
+      "epoch": 0.884657634566093,
+      "grad_norm": 0.2819498194158381,
+      "learning_rate": 1.3663622593731294e-05,
+      "loss": 0.1315,
+      "step": 604
+    },
+    {
+      "epoch": 0.8861222995239839,
+      "grad_norm": 0.2869700943965007,
+      "learning_rate": 1.3639786936121287e-05,
+      "loss": 0.1313,
+      "step": 605
+    },
+    {
+      "epoch": 0.8875869644818748,
+      "grad_norm": 0.28257828430613197,
+      "learning_rate": 1.3615927417637435e-05,
+      "loss": 0.1327,
+      "step": 606
+    },
+    {
+      "epoch": 0.8890516294397657,
+      "grad_norm": 0.2983987956601424,
+      "learning_rate": 1.3592044194692456e-05,
+      "loss": 0.1435,
+      "step": 607
+    },
+    {
+      "epoch": 0.8905162943976566,
+      "grad_norm": 0.2900543426098001,
+      "learning_rate": 1.356813742385446e-05,
+      "loss": 0.1366,
+      "step": 608
+    },
+    {
+      "epoch": 0.8919809593555474,
+      "grad_norm": 0.27572519984679683,
+      "learning_rate": 1.3544207261845928e-05,
+      "loss": 0.1373,
+      "step": 609
+    },
+    {
+      "epoch": 0.8934456243134383,
+      "grad_norm": 0.2721425773906768,
+      "learning_rate": 1.3520253865542687e-05,
+      "loss": 0.1223,
+      "step": 610
+    },
+    {
+      "epoch": 0.8949102892713292,
+      "grad_norm": 0.282049356782574,
+      "learning_rate": 1.3496277391972874e-05,
+      "loss": 0.1192,
+      "step": 611
+    },
+    {
+      "epoch": 0.8963749542292201,
+      "grad_norm": 0.2681081671045421,
+      "learning_rate": 1.3472277998315915e-05,
+      "loss": 0.1279,
+      "step": 612
+    },
+    {
+      "epoch": 0.897839619187111,
+      "grad_norm": 0.276619227889388,
+      "learning_rate": 1.3448255841901481e-05,
+      "loss": 0.1327,
+      "step": 613
+    },
+    {
+      "epoch": 0.8993042841450019,
+      "grad_norm": 0.28855198625509537,
+      "learning_rate": 1.3424211080208478e-05,
+      "loss": 0.1249,
+      "step": 614
+    },
+    {
+      "epoch": 0.9007689491028927,
+      "grad_norm": 0.25545241022812626,
+      "learning_rate": 1.3400143870863997e-05,
+      "loss": 0.1053,
+      "step": 615
+    },
+    {
+      "epoch": 0.9022336140607836,
+      "grad_norm": 0.24939455623227066,
+      "learning_rate": 1.3376054371642282e-05,
+      "loss": 0.1172,
+      "step": 616
+    },
+    {
+      "epoch": 0.9036982790186745,
+      "grad_norm": 0.2890615502578014,
+      "learning_rate": 1.3351942740463705e-05,
+      "loss": 0.1368,
+      "step": 617
+    },
+    {
+      "epoch": 0.9051629439765654,
+      "grad_norm": 0.2566294230704182,
+      "learning_rate": 1.3327809135393728e-05,
+      "loss": 0.1337,
+      "step": 618
+    },
+    {
+      "epoch": 0.9066276089344563,
+      "grad_norm": 0.2902103194088055,
+      "learning_rate": 1.3303653714641853e-05,
+      "loss": 0.1251,
+      "step": 619
+    },
+    {
+      "epoch": 0.9080922738923471,
+      "grad_norm": 0.2666865578103702,
+      "learning_rate": 1.3279476636560608e-05,
+      "loss": 0.1201,
+      "step": 620
+    },
+    {
+      "epoch": 0.909556938850238,
+      "grad_norm": 0.2618945176801082,
+      "learning_rate": 1.3255278059644496e-05,
+      "loss": 0.1145,
+      "step": 621
+    },
+    {
+      "epoch": 0.9110216038081289,
+      "grad_norm": 0.31710586966811777,
+      "learning_rate": 1.323105814252895e-05,
+      "loss": 0.1497,
+      "step": 622
+    },
+    {
+      "epoch": 0.9124862687660198,
+      "grad_norm": 0.29714914241544244,
+      "learning_rate": 1.3206817043989301e-05,
+      "loss": 0.1297,
+      "step": 623
+    },
+    {
+      "epoch": 0.9139509337239107,
+      "grad_norm": 0.26025416086413616,
+      "learning_rate": 1.3182554922939748e-05,
+      "loss": 0.1069,
+      "step": 624
+    },
+    {
+      "epoch": 0.9154155986818016,
+      "grad_norm": 0.2983087250370599,
+      "learning_rate": 1.3158271938432288e-05,
+      "loss": 0.1285,
+      "step": 625
+    },
+    {
+      "epoch": 0.9168802636396924,
+      "grad_norm": 0.27517063473066206,
+      "learning_rate": 1.3133968249655701e-05,
+      "loss": 0.1375,
+      "step": 626
+    },
+    {
+      "epoch": 0.9183449285975833,
+      "grad_norm": 0.2659448809334693,
+      "learning_rate": 1.3109644015934493e-05,
+      "loss": 0.1144,
+      "step": 627
+    },
+    {
+      "epoch": 0.9198095935554742,
+      "grad_norm": 0.31062215771722046,
+      "learning_rate": 1.3085299396727851e-05,
+      "loss": 0.1285,
+      "step": 628
+    },
+    {
+      "epoch": 0.9212742585133651,
+      "grad_norm": 0.2555958267399944,
+      "learning_rate": 1.3060934551628603e-05,
+      "loss": 0.1111,
+      "step": 629
+    },
+    {
+      "epoch": 0.922738923471256,
+      "grad_norm": 0.25825461588582355,
+      "learning_rate": 1.3036549640362169e-05,
+      "loss": 0.1153,
+      "step": 630
+    },
+    {
+      "epoch": 0.9242035884291468,
+      "grad_norm": 0.270581810873964,
+      "learning_rate": 1.301214482278551e-05,
+      "loss": 0.1243,
+      "step": 631
+    },
+    {
+      "epoch": 0.9256682533870377,
+      "grad_norm": 0.2632208259144816,
+      "learning_rate": 1.2987720258886094e-05,
+      "loss": 0.1098,
+      "step": 632
+    },
+    {
+      "epoch": 0.9271329183449286,
+      "grad_norm": 0.2680331255251313,
+      "learning_rate": 1.2963276108780829e-05,
+      "loss": 0.1286,
+      "step": 633
+    },
+    {
+      "epoch": 0.9285975833028195,
+      "grad_norm": 0.293030113117047,
+      "learning_rate": 1.293881253271502e-05,
+      "loss": 0.1559,
+      "step": 634
+    },
+    {
+      "epoch": 0.9300622482607104,
+      "grad_norm": 0.27474296914093777,
+      "learning_rate": 1.2914329691061327e-05,
+      "loss": 0.1144,
+      "step": 635
+    },
+    {
+      "epoch": 0.9315269132186013,
+      "grad_norm": 0.2738691898038605,
+      "learning_rate": 1.2889827744318705e-05,
+      "loss": 0.1159,
+      "step": 636
+    },
+    {
+      "epoch": 0.9329915781764921,
+      "grad_norm": 0.2926627981538906,
+      "learning_rate": 1.286530685311135e-05,
+      "loss": 0.1476,
+      "step": 637
+    },
+    {
+      "epoch": 0.934456243134383,
+      "grad_norm": 0.2542806182987207,
+      "learning_rate": 1.2840767178187657e-05,
+      "loss": 0.1076,
+      "step": 638
+    },
+    {
+      "epoch": 0.9359209080922739,
+      "grad_norm": 0.2752287662506346,
+      "learning_rate": 1.281620888041915e-05,
+      "loss": 0.1369,
+      "step": 639
+    },
+    {
+      "epoch": 0.9373855730501648,
+      "grad_norm": 0.251869065832337,
+      "learning_rate": 1.279163212079944e-05,
+      "loss": 0.116,
+      "step": 640
+    },
+    {
+      "epoch": 0.9388502380080557,
+      "grad_norm": 0.2609971662177529,
+      "learning_rate": 1.2767037060443173e-05,
+      "loss": 0.1058,
+      "step": 641
+    },
+    {
+      "epoch": 0.9403149029659466,
+      "grad_norm": 0.26794778938799796,
+      "learning_rate": 1.2742423860584954e-05,
+      "loss": 0.1212,
+      "step": 642
+    },
+    {
+      "epoch": 0.9417795679238374,
+      "grad_norm": 0.3047657017773294,
+      "learning_rate": 1.271779268257831e-05,
+      "loss": 0.1339,
+      "step": 643
+    },
+    {
+      "epoch": 0.9432442328817283,
+      "grad_norm": 0.2897849310784945,
+      "learning_rate": 1.269314368789463e-05,
+      "loss": 0.1084,
+      "step": 644
+    },
+    {
+      "epoch": 0.9447088978396192,
+      "grad_norm": 0.29786222980952776,
+      "learning_rate": 1.266847703812209e-05,
+      "loss": 0.1106,
+      "step": 645
+    },
+    {
+      "epoch": 0.9461735627975101,
+      "grad_norm": 0.2642203247056678,
+      "learning_rate": 1.2643792894964611e-05,
+      "loss": 0.1053,
+      "step": 646
+    },
+    {
+      "epoch": 0.947638227755401,
+      "grad_norm": 0.28156653824521066,
+      "learning_rate": 1.2619091420240795e-05,
+      "loss": 0.1458,
+      "step": 647
+    },
+    {
+      "epoch": 0.9491028927132918,
+      "grad_norm": 0.2557860588752611,
+      "learning_rate": 1.2594372775882862e-05,
+      "loss": 0.1086,
+      "step": 648
+    },
+    {
+      "epoch": 0.9505675576711827,
+      "grad_norm": 0.26613179341981313,
+      "learning_rate": 1.2569637123935581e-05,
+      "loss": 0.1081,
+      "step": 649
+    },
+    {
+      "epoch": 0.9520322226290736,
+      "grad_norm": 0.27208135899789015,
+      "learning_rate": 1.2544884626555225e-05,
+      "loss": 0.1309,
+      "step": 650
+    },
+    {
+      "epoch": 0.9534968875869645,
+      "grad_norm": 0.2800087048695156,
+      "learning_rate": 1.2520115446008493e-05,
+      "loss": 0.122,
+      "step": 651
+    },
+    {
+      "epoch": 0.9549615525448554,
+      "grad_norm": 0.265808555990337,
+      "learning_rate": 1.2495329744671457e-05,
+      "loss": 0.1143,
+      "step": 652
+    },
+    {
+      "epoch": 0.9564262175027463,
+      "grad_norm": 0.2832651191459517,
+      "learning_rate": 1.2470527685028485e-05,
+      "loss": 0.1119,
+      "step": 653
+    },
+    {
+      "epoch": 0.9578908824606371,
+      "grad_norm": 0.27691932316711165,
+      "learning_rate": 1.2445709429671184e-05,
+      "loss": 0.1196,
+      "step": 654
+    },
+    {
+      "epoch": 0.959355547418528,
+      "grad_norm": 0.24468827419730876,
+      "learning_rate": 1.2420875141297344e-05,
+      "loss": 0.1004,
+      "step": 655
+    },
+    {
+      "epoch": 0.9608202123764189,
+      "grad_norm": 0.28166326250407386,
+      "learning_rate": 1.2396024982709845e-05,
+      "loss": 0.1384,
+      "step": 656
+    },
+    {
+      "epoch": 0.9622848773343098,
+      "grad_norm": 0.2864348243903678,
+      "learning_rate": 1.2371159116815614e-05,
+      "loss": 0.1314,
+      "step": 657
+    },
+    {
+      "epoch": 0.9637495422922007,
+      "grad_norm": 0.29171091081890826,
+      "learning_rate": 1.234627770662455e-05,
+      "loss": 0.1348,
+      "step": 658
+    },
+    {
+      "epoch": 0.9652142072500915,
+      "grad_norm": 0.26640633088569415,
+      "learning_rate": 1.2321380915248446e-05,
+      "loss": 0.1215,
+      "step": 659
+    },
+    {
+      "epoch": 0.9666788722079824,
+      "grad_norm": 0.2719229168980855,
+      "learning_rate": 1.2296468905899937e-05,
+      "loss": 0.1163,
+      "step": 660
+    },
+    {
+      "epoch": 0.9681435371658733,
+      "grad_norm": 0.28951810957646756,
+      "learning_rate": 1.227154184189141e-05,
+      "loss": 0.1321,
+      "step": 661
+    },
+    {
+      "epoch": 0.9696082021237642,
+      "grad_norm": 0.30219688645041665,
+      "learning_rate": 1.2246599886633951e-05,
+      "loss": 0.1295,
+      "step": 662
+    },
+    {
+      "epoch": 0.9710728670816551,
+      "grad_norm": 0.29588778808501925,
+      "learning_rate": 1.222164320363627e-05,
+      "loss": 0.1373,
+      "step": 663
+    },
+    {
+      "epoch": 0.972537532039546,
+      "grad_norm": 0.26451637010025986,
+      "learning_rate": 1.2196671956503611e-05,
+      "loss": 0.1187,
+      "step": 664
+    },
+    {
+      "epoch": 0.9740021969974368,
+      "grad_norm": 0.2670469003783616,
+      "learning_rate": 1.217168630893671e-05,
+      "loss": 0.1223,
+      "step": 665
+    },
+    {
+      "epoch": 0.9754668619553277,
+      "grad_norm": 0.3201350019334305,
+      "learning_rate": 1.2146686424730699e-05,
+      "loss": 0.1238,
+      "step": 666
+    },
+    {
+      "epoch": 0.9769315269132186,
+      "grad_norm": 0.25030723292519363,
+      "learning_rate": 1.212167246777404e-05,
+      "loss": 0.1087,
+      "step": 667
+    },
+    {
+      "epoch": 0.9783961918711095,
+      "grad_norm": 0.30670242040276763,
+      "learning_rate": 1.2096644602047447e-05,
+      "loss": 0.1417,
+      "step": 668
+    },
+    {
+      "epoch": 0.9798608568290004,
+      "grad_norm": 0.3387650718944703,
+      "learning_rate": 1.2071602991622822e-05,
+      "loss": 0.1244,
+      "step": 669
+    },
+    {
+      "epoch": 0.9813255217868913,
+      "grad_norm": 0.31738790684917234,
+      "learning_rate": 1.2046547800662163e-05,
+      "loss": 0.1379,
+      "step": 670
+    },
+    {
+      "epoch": 0.9827901867447821,
+      "grad_norm": 0.253861086544748,
+      "learning_rate": 1.2021479193416502e-05,
+      "loss": 0.0972,
+      "step": 671
+    },
+    {
+      "epoch": 0.984254851702673,
+      "grad_norm": 0.27857395289374787,
+      "learning_rate": 1.1996397334224814e-05,
+      "loss": 0.1453,
+      "step": 672
+    },
+    {
+      "epoch": 0.9857195166605639,
+      "grad_norm": 0.2649436923127358,
+      "learning_rate": 1.1971302387512958e-05,
+      "loss": 0.1338,
+      "step": 673
+    },
+    {
+      "epoch": 0.9871841816184548,
+      "grad_norm": 0.27979461023670804,
+      "learning_rate": 1.1946194517792584e-05,
+      "loss": 0.1245,
+      "step": 674
+    },
+    {
+      "epoch": 0.9886488465763457,
+      "grad_norm": 0.28275198179954225,
+      "learning_rate": 1.1921073889660061e-05,
+      "loss": 0.1316,
+      "step": 675
+    },
+    {
+      "epoch": 0.9901135115342365,
+      "grad_norm": 0.26979156442079305,
+      "learning_rate": 1.1895940667795395e-05,
+      "loss": 0.1219,
+      "step": 676
+    },
+    {
+      "epoch": 0.9915781764921274,
+      "grad_norm": 0.25593923579587835,
+      "learning_rate": 1.1870795016961157e-05,
+      "loss": 0.1048,
+      "step": 677
+    },
+    {
+      "epoch": 0.9930428414500183,
+      "grad_norm": 0.2542007022812753,
+      "learning_rate": 1.1845637102001383e-05,
+      "loss": 0.1147,
+      "step": 678
+    },
+    {
+      "epoch": 0.9945075064079092,
+      "grad_norm": 0.2763537672373692,
+      "learning_rate": 1.1820467087840526e-05,
+      "loss": 0.1228,
+      "step": 679
+    },
+    {
+      "epoch": 0.9959721713658001,
+      "grad_norm": 0.2750343262145036,
+      "learning_rate": 1.1795285139482341e-05,
+      "loss": 0.1294,
+      "step": 680
+    },
+    {
+      "epoch": 0.997436836323691,
+      "grad_norm": 0.2774530874594845,
+      "learning_rate": 1.1770091422008824e-05,
+      "loss": 0.1252,
+      "step": 681
+    },
+    {
+      "epoch": 0.9989015012815818,
+      "grad_norm": 0.2607903166254994,
+      "learning_rate": 1.174488610057913e-05,
+      "loss": 0.119,
+      "step": 682
+    },
+    {
+      "epoch": 1.0003661662394727,
+      "grad_norm": 0.2645760004291393,
+      "learning_rate": 1.1719669340428472e-05,
+      "loss": 0.1128,
+      "step": 683
+    },
+    {
+      "epoch": 1.0018308311973636,
+      "grad_norm": 0.2552221963588916,
+      "learning_rate": 1.1694441306867062e-05,
+      "loss": 0.0897,
+      "step": 684
+    },
+    {
+      "epoch": 1.0032954961552545,
+      "grad_norm": 0.24512146436759236,
+      "learning_rate": 1.1669202165279009e-05,
+      "loss": 0.0966,
+      "step": 685
+    },
+    {
+      "epoch": 1.0047601611131454,
+      "grad_norm": 0.2926151907338783,
+      "learning_rate": 1.164395208112124e-05,
+      "loss": 0.0988,
+      "step": 686
+    },
+    {
+      "epoch": 1.0062248260710362,
+      "grad_norm": 0.30856359947434714,
+      "learning_rate": 1.1618691219922426e-05,
+      "loss": 0.1138,
+      "step": 687
+    },
+    {
+      "epoch": 1.0076894910289271,
+      "grad_norm": 0.26943247650215035,
+      "learning_rate": 1.159341974728188e-05,
+      "loss": 0.1157,
+      "step": 688
+    },
+    {
+      "epoch": 1.009154155986818,
+      "grad_norm": 0.28468116825059964,
+      "learning_rate": 1.1568137828868478e-05,
+      "loss": 0.0881,
+      "step": 689
+    },
+    {
+      "epoch": 1.010618820944709,
+      "grad_norm": 0.2399332416399268,
+      "learning_rate": 1.1542845630419579e-05,
+      "loss": 0.0893,
+      "step": 690
+    },
+    {
+      "epoch": 1.0120834859025998,
+      "grad_norm": 0.3288588984625157,
+      "learning_rate": 1.1517543317739931e-05,
+      "loss": 0.1076,
+      "step": 691
+    },
+    {
+      "epoch": 1.0135481508604907,
+      "grad_norm": 0.26604448567697603,
+      "learning_rate": 1.1492231056700592e-05,
+      "loss": 0.0932,
+      "step": 692
+    },
+    {
+      "epoch": 1.0150128158183815,
+      "grad_norm": 0.24018805379412664,
+      "learning_rate": 1.1466909013237819e-05,
+      "loss": 0.1001,
+      "step": 693
+    },
+    {
+      "epoch": 1.0164774807762724,
+      "grad_norm": 0.2651475188976319,
+      "learning_rate": 1.1441577353352023e-05,
+      "loss": 0.1035,
+      "step": 694
+    },
+    {
+      "epoch": 1.0179421457341633,
+      "grad_norm": 0.31536104086551264,
+      "learning_rate": 1.1416236243106638e-05,
+      "loss": 0.1081,
+      "step": 695
+    },
+    {
+      "epoch": 1.0194068106920542,
+      "grad_norm": 0.2613280520512752,
+      "learning_rate": 1.1390885848627058e-05,
+      "loss": 0.0845,
+      "step": 696
+    },
+    {
+      "epoch": 1.020871475649945,
+      "grad_norm": 0.2837125123903259,
+      "learning_rate": 1.1365526336099542e-05,
+      "loss": 0.1328,
+      "step": 697
+    },
+    {
+      "epoch": 1.022336140607836,
+      "grad_norm": 0.259222134905861,
+      "learning_rate": 1.1340157871770117e-05,
+      "loss": 0.0886,
+      "step": 698
+    },
+    {
+      "epoch": 1.0238008055657268,
+      "grad_norm": 0.2865990464407519,
+      "learning_rate": 1.13147806219435e-05,
+      "loss": 0.1079,
+      "step": 699
+    },
+    {
+      "epoch": 1.0252654705236177,
+      "grad_norm": 0.2794232474113488,
+      "learning_rate": 1.1289394752982e-05,
+      "loss": 0.106,
+      "step": 700
+    },
+    {
+      "epoch": 1.0267301354815086,
+      "grad_norm": 0.2606486568133289,
+      "learning_rate": 1.1264000431304422e-05,
+      "loss": 0.0902,
+      "step": 701
+    },
+    {
+      "epoch": 1.0281948004393995,
+      "grad_norm": 0.27767121394939476,
+      "learning_rate": 1.1238597823385e-05,
+      "loss": 0.0915,
+      "step": 702
+    },
+    {
+      "epoch": 1.0296594653972904,
+      "grad_norm": 0.2865112658979935,
+      "learning_rate": 1.1213187095752271e-05,
+      "loss": 0.0994,
+      "step": 703
+    },
+    {
+      "epoch": 1.0311241303551812,
+      "grad_norm": 0.28382542458351684,
+      "learning_rate": 1.1187768414988015e-05,
+      "loss": 0.0922,
+      "step": 704
+    },
+    {
+      "epoch": 1.0325887953130721,
+      "grad_norm": 0.27487686960098845,
+      "learning_rate": 1.1162341947726139e-05,
+      "loss": 0.1187,
+      "step": 705
+    },
+    {
+      "epoch": 1.034053460270963,
+      "grad_norm": 0.2629377900389775,
+      "learning_rate": 1.1136907860651603e-05,
+      "loss": 0.0894,
+      "step": 706
+    },
+    {
+      "epoch": 1.035518125228854,
+      "grad_norm": 0.2613378798506544,
+      "learning_rate": 1.1111466320499318e-05,
+      "loss": 0.0921,
+      "step": 707
+    },
+    {
+      "epoch": 1.0369827901867448,
+      "grad_norm": 0.255691486058076,
+      "learning_rate": 1.1086017494053046e-05,
+      "loss": 0.0937,
+      "step": 708
+    },
+    {
+      "epoch": 1.0384474551446357,
+      "grad_norm": 0.2595524157003849,
+      "learning_rate": 1.1060561548144321e-05,
+      "loss": 0.0898,
+      "step": 709
+    },
+    {
+      "epoch": 1.0399121201025265,
+      "grad_norm": 0.27529763982095773,
+      "learning_rate": 1.1035098649651355e-05,
+      "loss": 0.0966,
+      "step": 710
+    },
+    {
+      "epoch": 1.0413767850604174,
+      "grad_norm": 0.2518906472665152,
+      "learning_rate": 1.1009628965497927e-05,
+      "loss": 0.0842,
+      "step": 711
+    },
+    {
+      "epoch": 1.0428414500183083,
+      "grad_norm": 0.27904396918210833,
+      "learning_rate": 1.0984152662652307e-05,
+      "loss": 0.1148,
+      "step": 712
+    },
+    {
+      "epoch": 1.0443061149761992,
+      "grad_norm": 0.2651133079255485,
+      "learning_rate": 1.0958669908126151e-05,
+      "loss": 0.087,
+      "step": 713
+    },
+    {
+      "epoch": 1.04577077993409,
+      "grad_norm": 0.2767992985128494,
+      "learning_rate": 1.0933180868973414e-05,
+      "loss": 0.1033,
+      "step": 714
+    },
+    {
+      "epoch": 1.047235444891981,
+      "grad_norm": 0.2527734956067706,
+      "learning_rate": 1.0907685712289244e-05,
+      "loss": 0.0912,
+      "step": 715
+    },
+    {
+      "epoch": 1.0487001098498718,
+      "grad_norm": 0.26205384399328313,
+      "learning_rate": 1.0882184605208895e-05,
+      "loss": 0.089,
+      "step": 716
+    },
+    {
+      "epoch": 1.0501647748077627,
+      "grad_norm": 0.2680315695418657,
+      "learning_rate": 1.0856677714906632e-05,
+      "loss": 0.1001,
+      "step": 717
+    },
+    {
+      "epoch": 1.0516294397656536,
+      "grad_norm": 0.2265048738569077,
+      "learning_rate": 1.083116520859463e-05,
+      "loss": 0.0897,
+      "step": 718
+    },
+    {
+      "epoch": 1.0530941047235445,
+      "grad_norm": 0.2507364042609683,
+      "learning_rate": 1.080564725352188e-05,
+      "loss": 0.0867,
+      "step": 719
+    },
+    {
+      "epoch": 1.0545587696814354,
+      "grad_norm": 0.27387442999591954,
+      "learning_rate": 1.0780124016973095e-05,
+      "loss": 0.0998,
+      "step": 720
+    },
+    {
+      "epoch": 1.0560234346393262,
+      "grad_norm": 0.24270388287807182,
+      "learning_rate": 1.0754595666267609e-05,
+      "loss": 0.0817,
+      "step": 721
+    },
+    {
+      "epoch": 1.0574880995972171,
+      "grad_norm": 0.26311541978879793,
+      "learning_rate": 1.0729062368758278e-05,
+      "loss": 0.0988,
+      "step": 722
+    },
+    {
+      "epoch": 1.058952764555108,
+      "grad_norm": 0.2704618854801208,
+      "learning_rate": 1.0703524291830398e-05,
+      "loss": 0.1082,
+      "step": 723
+    },
+    {
+      "epoch": 1.0604174295129989,
+      "grad_norm": 0.28368113431359887,
+      "learning_rate": 1.067798160290059e-05,
+      "loss": 0.1009,
+      "step": 724
+    },
+    {
+      "epoch": 1.0618820944708898,
+      "grad_norm": 0.26885585300497,
+      "learning_rate": 1.0652434469415705e-05,
+      "loss": 0.0986,
+      "step": 725
+    },
+    {
+      "epoch": 1.0633467594287807,
+      "grad_norm": 0.24916401223915322,
+      "learning_rate": 1.0626883058851737e-05,
+      "loss": 0.0859,
+      "step": 726
+    },
+    {
+      "epoch": 1.0648114243866715,
+      "grad_norm": 0.25631003685203035,
+      "learning_rate": 1.0601327538712723e-05,
+      "loss": 0.0923,
+      "step": 727
+    },
+    {
+      "epoch": 1.0662760893445624,
+      "grad_norm": 0.2552784210398989,
+      "learning_rate": 1.0575768076529627e-05,
+      "loss": 0.0761,
+      "step": 728
+    },
+    {
+      "epoch": 1.0677407543024533,
+      "grad_norm": 0.24165749739319964,
+      "learning_rate": 1.0550204839859265e-05,
+      "loss": 0.0752,
+      "step": 729
+    },
+    {
+      "epoch": 1.0692054192603442,
+      "grad_norm": 0.2675621328921695,
+      "learning_rate": 1.0524637996283195e-05,
+      "loss": 0.1019,
+      "step": 730
+    },
+    {
+      "epoch": 1.070670084218235,
+      "grad_norm": 0.2569008765612613,
+      "learning_rate": 1.0499067713406622e-05,
+      "loss": 0.0964,
+      "step": 731
+    },
+    {
+      "epoch": 1.072134749176126,
+      "grad_norm": 0.24678690292067554,
+      "learning_rate": 1.0473494158857298e-05,
+      "loss": 0.0891,
+      "step": 732
+    },
+    {
+      "epoch": 1.0735994141340168,
+      "grad_norm": 0.2549990776594623,
+      "learning_rate": 1.0447917500284415e-05,
+      "loss": 0.0925,
+      "step": 733
+    },
+    {
+      "epoch": 1.0750640790919077,
+      "grad_norm": 0.28063287681936966,
+      "learning_rate": 1.0422337905357523e-05,
+      "loss": 0.1053,
+      "step": 734
+    },
+    {
+      "epoch": 1.0765287440497986,
+      "grad_norm": 0.27072473670857133,
+      "learning_rate": 1.0396755541765413e-05,
+      "loss": 0.0951,
+      "step": 735
+    },
+    {
+      "epoch": 1.0779934090076895,
+      "grad_norm": 0.2804354889898814,
+      "learning_rate": 1.0371170577215036e-05,
+      "loss": 0.1044,
+      "step": 736
+    },
+    {
+      "epoch": 1.0794580739655804,
+      "grad_norm": 0.2707408971145425,
+      "learning_rate": 1.0345583179430387e-05,
+      "loss": 0.087,
+      "step": 737
+    },
+    {
+      "epoch": 1.0809227389234712,
+      "grad_norm": 0.27265959106298226,
+      "learning_rate": 1.0319993516151412e-05,
+      "loss": 0.0957,
+      "step": 738
+    },
+    {
+      "epoch": 1.0823874038813621,
+      "grad_norm": 0.2664867481183081,
+      "learning_rate": 1.0294401755132912e-05,
+      "loss": 0.0838,
+      "step": 739
+    },
+    {
+      "epoch": 1.083852068839253,
+      "grad_norm": 0.2827928379341626,
+      "learning_rate": 1.0268808064143438e-05,
+      "loss": 0.0981,
+      "step": 740
+    },
+    {
+      "epoch": 1.0853167337971439,
+      "grad_norm": 0.2516210381548982,
+      "learning_rate": 1.0243212610964192e-05,
+      "loss": 0.085,
+      "step": 741
+    },
+    {
+      "epoch": 1.0867813987550348,
+      "grad_norm": 0.2609165490224745,
+      "learning_rate": 1.0217615563387932e-05,
+      "loss": 0.1047,
+      "step": 742
+    },
+    {
+      "epoch": 1.0882460637129256,
+      "grad_norm": 0.2522780231083236,
+      "learning_rate": 1.0192017089217863e-05,
+      "loss": 0.1029,
+      "step": 743
+    },
+    {
+      "epoch": 1.0897107286708165,
+      "grad_norm": 0.2640632238176228,
+      "learning_rate": 1.0166417356266546e-05,
+      "loss": 0.1053,
+      "step": 744
+    },
+    {
+      "epoch": 1.0911753936287074,
+      "grad_norm": 0.266832288851606,
+      "learning_rate": 1.0140816532354793e-05,
+      "loss": 0.1068,
+      "step": 745
+    },
+    {
+      "epoch": 1.0926400585865983,
+      "grad_norm": 0.24267415544535567,
+      "learning_rate": 1.0115214785310567e-05,
+      "loss": 0.0723,
+      "step": 746
+    },
+    {
+      "epoch": 1.0941047235444892,
+      "grad_norm": 0.280429753176534,
+      "learning_rate": 1.0089612282967884e-05,
+      "loss": 0.1071,
+      "step": 747
+    },
+    {
+      "epoch": 1.09556938850238,
+      "grad_norm": 0.26916094567137633,
+      "learning_rate": 1.0064009193165713e-05,
+      "loss": 0.0858,
+      "step": 748
+    },
+    {
+      "epoch": 1.097034053460271,
+      "grad_norm": 0.274415937658407,
+      "learning_rate": 1.0038405683746868e-05,
+      "loss": 0.0869,
+      "step": 749
+    },
+    {
+      "epoch": 1.0984987184181618,
+      "grad_norm": 0.2792545952852677,
+      "learning_rate": 1.0012801922556918e-05,
+      "loss": 0.0893,
+      "step": 750
+    },
+    {
+      "epoch": 1.0999633833760527,
+      "grad_norm": 0.2875272278765476,
+      "learning_rate": 9.987198077443085e-06,
+      "loss": 0.1099,
+      "step": 751
+    },
+    {
+      "epoch": 1.1014280483339436,
+      "grad_norm": 0.260518607411576,
+      "learning_rate": 9.961594316253134e-06,
+      "loss": 0.0828,
+      "step": 752
+    },
+    {
+      "epoch": 1.1028927132918345,
+      "grad_norm": 0.2633760525239851,
+      "learning_rate": 9.93599080683429e-06,
+      "loss": 0.0969,
+      "step": 753
+    },
+    {
+      "epoch": 1.1043573782497254,
+      "grad_norm": 0.25909573953326753,
+      "learning_rate": 9.910387717032115e-06,
+      "loss": 0.0756,
+      "step": 754
+    },
+    {
+      "epoch": 1.1058220432076162,
+      "grad_norm": 0.27253806701721073,
+      "learning_rate": 9.884785214689435e-06,
+      "loss": 0.0856,
+      "step": 755
+    },
+    {
+      "epoch": 1.1072867081655071,
+      "grad_norm": 0.2597508988107494,
+      "learning_rate": 9.859183467645207e-06,
+      "loss": 0.0855,
+      "step": 756
+    },
+    {
+      "epoch": 1.108751373123398,
+      "grad_norm": 0.23868371653633474,
+      "learning_rate": 9.833582643733457e-06,
+      "loss": 0.0826,
+      "step": 757
+    },
+    {
+      "epoch": 1.1102160380812889,
+      "grad_norm": 0.29179266600473996,
+      "learning_rate": 9.807982910782142e-06,
+      "loss": 0.1156,
+      "step": 758
+    },
+    {
+      "epoch": 1.1116807030391798,
+      "grad_norm": 0.2522539204555841,
+      "learning_rate": 9.782384436612072e-06,
+      "loss": 0.0908,
+      "step": 759
+    },
+    {
+      "epoch": 1.1131453679970706,
+      "grad_norm": 0.23700004230402077,
+      "learning_rate": 9.756787389035813e-06,
+      "loss": 0.0832,
+      "step": 760
+    },
+    {
+      "epoch": 1.1146100329549615,
+      "grad_norm": 0.2829087070060208,
+      "learning_rate": 9.731191935856566e-06,
+      "loss": 0.1116,
+      "step": 761
+    },
+    {
+      "epoch": 1.1160746979128524,
+      "grad_norm": 0.26523704277336946,
+      "learning_rate": 9.705598244867093e-06,
+      "loss": 0.0939,
+      "step": 762
+    },
+    {
+      "epoch": 1.1175393628707433,
+      "grad_norm": 0.24086964841423913,
+      "learning_rate": 9.68000648384859e-06,
+      "loss": 0.0873,
+      "step": 763
+    },
+    {
+      "epoch": 1.1190040278286342,
+      "grad_norm": 0.2654475048077318,
+      "learning_rate": 9.654416820569618e-06,
+      "loss": 0.1006,
+      "step": 764
+    },
+    {
+      "epoch": 1.120468692786525,
+      "grad_norm": 0.2601239282279833,
+      "learning_rate": 9.628829422784965e-06,
+      "loss": 0.0773,
+      "step": 765
+    },
+    {
+      "epoch": 1.121933357744416,
+      "grad_norm": 0.2569521050103609,
+      "learning_rate": 9.603244458234589e-06,
+      "loss": 0.0793,
+      "step": 766
+    },
+    {
+      "epoch": 1.1233980227023068,
+      "grad_norm": 0.26593809509671273,
+      "learning_rate": 9.577662094642478e-06,
+      "loss": 0.0801,
+      "step": 767
+    },
+    {
+      "epoch": 1.1248626876601977,
+      "grad_norm": 0.29040791441938224,
+      "learning_rate": 9.552082499715588e-06,
+      "loss": 0.0933,
+      "step": 768
+    },
+    {
+      "epoch": 1.1263273526180886,
+      "grad_norm": 0.2943106175306918,
+      "learning_rate": 9.526505841142702e-06,
+      "loss": 0.0861,
+      "step": 769
+    },
+    {
+      "epoch": 1.1277920175759795,
+      "grad_norm": 0.2601745466903212,
+      "learning_rate": 9.50093228659338e-06,
+      "loss": 0.0956,
+      "step": 770
+    },
+    {
+      "epoch": 1.1292566825338703,
+      "grad_norm": 0.26049784872083914,
+      "learning_rate": 9.475362003716804e-06,
+      "loss": 0.0826,
+      "step": 771
+    },
+    {
+      "epoch": 1.1307213474917612,
+      "grad_norm": 0.2629400244245209,
+      "learning_rate": 9.449795160140737e-06,
+      "loss": 0.0867,
+      "step": 772
+    },
+    {
+      "epoch": 1.1321860124496521,
+      "grad_norm": 0.2621080280544426,
+      "learning_rate": 9.424231923470378e-06,
+      "loss": 0.1065,
+      "step": 773
+    },
+    {
+      "epoch": 1.133650677407543,
+      "grad_norm": 0.26858890551499126,
+      "learning_rate": 9.39867246128728e-06,
+      "loss": 0.0941,
+      "step": 774
+    },
+    {
+      "epoch": 1.1351153423654339,
+      "grad_norm": 0.2576626727357875,
+      "learning_rate": 9.373116941148264e-06,
+      "loss": 0.0964,
+      "step": 775
+    },
+    {
+      "epoch": 1.1365800073233248,
+      "grad_norm": 0.26215360922586484,
+      "learning_rate": 9.347565530584299e-06,
+      "loss": 0.093,
+      "step": 776
+    },
+    {
+      "epoch": 1.1380446722812156,
+      "grad_norm": 0.25775670982713683,
+      "learning_rate": 9.322018397099414e-06,
+      "loss": 0.0921,
+      "step": 777
+    },
+    {
+      "epoch": 1.1395093372391065,
+      "grad_norm": 0.26195192691999436,
+      "learning_rate": 9.296475708169603e-06,
+      "loss": 0.0944,
+      "step": 778
+    },
+    {
+      "epoch": 1.1409740021969974,
+      "grad_norm": 0.24607830549799634,
+      "learning_rate": 9.270937631241723e-06,
+      "loss": 0.0789,
+      "step": 779
+    },
+    {
+      "epoch": 1.1424386671548883,
+      "grad_norm": 0.2504436025347225,
+      "learning_rate": 9.245404333732395e-06,
+      "loss": 0.0844,
+      "step": 780
+    },
+    {
+      "epoch": 1.1439033321127792,
+      "grad_norm": 0.2577203090564361,
+      "learning_rate": 9.219875983026909e-06,
+      "loss": 0.0834,
+      "step": 781
+    },
+    {
+      "epoch": 1.14536799707067,
+      "grad_norm": 0.25652775031593833,
+      "learning_rate": 9.194352746478122e-06,
+      "loss": 0.0806,
+      "step": 782
+    },
+    {
+      "epoch": 1.146832662028561,
+      "grad_norm": 0.27571423516211513,
+      "learning_rate": 9.168834791405374e-06,
+      "loss": 0.1015,
+      "step": 783
+    },
+    {
+      "epoch": 1.1482973269864518,
+      "grad_norm": 0.25833280528124036,
+      "learning_rate": 9.143322285093371e-06,
+      "loss": 0.0879,
+      "step": 784
+    },
+    {
+      "epoch": 1.1497619919443427,
+      "grad_norm": 0.27768929864428427,
+      "learning_rate": 9.117815394791107e-06,
+      "loss": 0.0916,
+      "step": 785
+    },
+    {
+      "epoch": 1.1512266569022336,
+      "grad_norm": 0.27218967256543264,
+      "learning_rate": 9.092314287710757e-06,
+      "loss": 0.0791,
+      "step": 786
+    },
+    {
+      "epoch": 1.1526913218601245,
+      "grad_norm": 0.290779647815006,
+      "learning_rate": 9.066819131026588e-06,
+      "loss": 0.1023,
+      "step": 787
+    },
+    {
+      "epoch": 1.1541559868180153,
+      "grad_norm": 0.27792973768255624,
+      "learning_rate": 9.041330091873852e-06,
+      "loss": 0.0979,
+      "step": 788
+    },
+    {
+      "epoch": 1.1556206517759062,
+      "grad_norm": 0.2373738656510741,
+      "learning_rate": 9.015847337347695e-06,
+      "loss": 0.0806,
+      "step": 789
+    },
+    {
+      "epoch": 1.157085316733797,
+      "grad_norm": 0.25696429168971635,
+      "learning_rate": 8.990371034502078e-06,
+      "loss": 0.0918,
+      "step": 790
+    },
+    {
+      "epoch": 1.158549981691688,
+      "grad_norm": 0.280392566025079,
+      "learning_rate": 8.964901350348648e-06,
+      "loss": 0.0949,
+      "step": 791
+    },
+    {
+      "epoch": 1.1600146466495789,
+      "grad_norm": 0.2535223646372367,
+      "learning_rate": 8.939438451855684e-06,
+      "loss": 0.0816,
+      "step": 792
+    },
+    {
+      "epoch": 1.1614793116074698,
+      "grad_norm": 0.24735896623108233,
+      "learning_rate": 8.913982505946958e-06,
+      "loss": 0.0876,
+      "step": 793
+    },
+    {
+      "epoch": 1.1629439765653606,
+      "grad_norm": 0.2543060334798377,
+      "learning_rate": 8.888533679500688e-06,
+      "loss": 0.0866,
+      "step": 794
+    },
+    {
+      "epoch": 1.1644086415232515,
+      "grad_norm": 0.2648636536906661,
+      "learning_rate": 8.863092139348397e-06,
+      "loss": 0.1009,
+      "step": 795
+    },
+    {
+      "epoch": 1.1658733064811424,
+      "grad_norm": 0.24166274223060005,
+      "learning_rate": 8.837658052273863e-06,
+      "loss": 0.0744,
+      "step": 796
+    },
+    {
+      "epoch": 1.1673379714390333,
+      "grad_norm": 0.2593529998079471,
+      "learning_rate": 8.812231585011987e-06,
+      "loss": 0.1102,
+      "step": 797
+    },
+    {
+      "epoch": 1.1688026363969242,
+      "grad_norm": 0.2547937310328503,
+      "learning_rate": 8.78681290424773e-06,
+      "loss": 0.083,
+      "step": 798
+    },
+    {
+      "epoch": 1.170267301354815,
+      "grad_norm": 0.2678047431199921,
+      "learning_rate": 8.761402176615002e-06,
+      "loss": 0.1004,
+      "step": 799
+    },
+    {
+      "epoch": 1.171731966312706,
+      "grad_norm": 0.2881247911795614,
+      "learning_rate": 8.735999568695581e-06,
+      "loss": 0.0941,
+      "step": 800
+    },
+    {
+      "epoch": 1.1731966312705968,
+      "grad_norm": 0.306665541632391,
+      "learning_rate": 8.710605247018002e-06,
+      "loss": 0.1016,
+      "step": 801
+    },
+    {
+      "epoch": 1.1746612962284877,
+      "grad_norm": 0.26549998092914273,
+      "learning_rate": 8.685219378056503e-06,
+      "loss": 0.1025,
+      "step": 802
+    },
+    {
+      "epoch": 1.1761259611863786,
+      "grad_norm": 0.25884918535779855,
+      "learning_rate": 8.659842128229886e-06,
+      "loss": 0.0796,
+      "step": 803
+    },
+    {
+      "epoch": 1.1775906261442695,
+      "grad_norm": 0.29483213665665864,
+      "learning_rate": 8.634473663900461e-06,
+      "loss": 0.0936,
+      "step": 804
+    },
+    {
+      "epoch": 1.1790552911021603,
+      "grad_norm": 0.2827884978675775,
+      "learning_rate": 8.609114151372947e-06,
+      "loss": 0.0877,
+      "step": 805
+    },
+    {
+      "epoch": 1.1805199560600512,
+      "grad_norm": 0.2929302563222184,
+      "learning_rate": 8.583763756893366e-06,
+      "loss": 0.0955,
+      "step": 806
+    },
+    {
+      "epoch": 1.181984621017942,
+      "grad_norm": 0.2538641882723923,
+      "learning_rate": 8.558422646647984e-06,
+      "loss": 0.0808,
+      "step": 807
+    },
+    {
+      "epoch": 1.183449285975833,
+      "grad_norm": 0.26631517555229367,
+      "learning_rate": 8.533090986762183e-06,
+      "loss": 0.0834,
+      "step": 808
+    },
+    {
+      "epoch": 1.1849139509337239,
+      "grad_norm": 0.2788403763341659,
+      "learning_rate": 8.507768943299415e-06,
+      "loss": 0.0969,
+      "step": 809
+    },
+    {
+      "epoch": 1.1863786158916148,
+      "grad_norm": 0.2781094675108167,
+      "learning_rate": 8.482456682260069e-06,
+      "loss": 0.0963,
+      "step": 810
+    },
+    {
+      "epoch": 1.1878432808495056,
+      "grad_norm": 0.2473923042956088,
+      "learning_rate": 8.457154369580424e-06,
+      "loss": 0.0843,
+      "step": 811
+    },
+    {
+      "epoch": 1.1893079458073965,
+      "grad_norm": 0.23006340181208546,
+      "learning_rate": 8.431862171131524e-06,
+      "loss": 0.0695,
+      "step": 812
+    },
+    {
+      "epoch": 1.1907726107652874,
+      "grad_norm": 0.2556855544794983,
+      "learning_rate": 8.406580252718125e-06,
+      "loss": 0.0773,
+      "step": 813
+    },
+    {
+      "epoch": 1.1922372757231783,
+      "grad_norm": 0.26970894596044315,
+      "learning_rate": 8.381308780077575e-06,
+      "loss": 0.0862,
+      "step": 814
+    },
+    {
+      "epoch": 1.1937019406810692,
+      "grad_norm": 0.25936009464878035,
+      "learning_rate": 8.356047918878762e-06,
+      "loss": 0.0909,
+      "step": 815
+    },
+    {
+      "epoch": 1.19516660563896,
+      "grad_norm": 0.26608540904879535,
+      "learning_rate": 8.330797834720993e-06,
+      "loss": 0.0833,
+      "step": 816
+    },
+    {
+      "epoch": 1.196631270596851,
+      "grad_norm": 0.2591055743281483,
+      "learning_rate": 8.305558693132943e-06,
+      "loss": 0.0798,
+      "step": 817
+    },
+    {
+      "epoch": 1.1980959355547418,
+      "grad_norm": 0.2607615757724153,
+      "learning_rate": 8.280330659571532e-06,
+      "loss": 0.0891,
+      "step": 818
+    },
+    {
+      "epoch": 1.1995606005126327,
+      "grad_norm": 0.27689441116344976,
+      "learning_rate": 8.255113899420873e-06,
+      "loss": 0.1042,
+      "step": 819
+    },
+    {
+      "epoch": 1.2010252654705236,
+      "grad_norm": 0.2634058051717049,
+      "learning_rate": 8.229908577991177e-06,
+      "loss": 0.0794,
+      "step": 820
+    },
+    {
+      "epoch": 1.2024899304284145,
+      "grad_norm": 0.2712603230168793,
+      "learning_rate": 8.204714860517662e-06,
+      "loss": 0.0872,
+      "step": 821
+    },
+    {
+      "epoch": 1.2039545953863053,
+      "grad_norm": 0.26002461903187646,
+      "learning_rate": 8.179532912159477e-06,
+      "loss": 0.0806,
+      "step": 822
+    },
+    {
+      "epoch": 1.2054192603441962,
+      "grad_norm": 0.2778681441466659,
+      "learning_rate": 8.154362897998619e-06,
+      "loss": 0.0954,
+      "step": 823
+    },
+    {
+      "epoch": 1.206883925302087,
+      "grad_norm": 0.2597653658409811,
+      "learning_rate": 8.129204983038847e-06,
+      "loss": 0.0837,
+      "step": 824
+    },
+    {
+      "epoch": 1.208348590259978,
+      "grad_norm": 0.2409553005949464,
+      "learning_rate": 8.104059332204606e-06,
+      "loss": 0.0797,
+      "step": 825
+    },
+    {
+      "epoch": 1.2098132552178689,
+      "grad_norm": 0.26099557745468915,
+      "learning_rate": 8.07892611033994e-06,
+      "loss": 0.0856,
+      "step": 826
+    },
+    {
+      "epoch": 1.2112779201757597,
+      "grad_norm": 0.2670243095090208,
+      "learning_rate": 8.053805482207418e-06,
+      "loss": 0.0981,
+      "step": 827
+    },
+    {
+      "epoch": 1.2127425851336506,
+      "grad_norm": 0.25384500506062874,
+      "learning_rate": 8.028697612487046e-06,
+      "loss": 0.0855,
+      "step": 828
+    },
+    {
+      "epoch": 1.2142072500915415,
+      "grad_norm": 0.27259004018505645,
+      "learning_rate": 8.003602665775189e-06,
+      "loss": 0.0895,
+      "step": 829
+    },
+    {
+      "epoch": 1.2156719150494324,
+      "grad_norm": 0.24539935533491744,
+      "learning_rate": 7.978520806583503e-06,
+      "loss": 0.075,
+      "step": 830
+    },
+    {
+      "epoch": 1.2171365800073233,
+      "grad_norm": 0.28931365047548946,
+      "learning_rate": 7.95345219933784e-06,
+      "loss": 0.0939,
+      "step": 831
+    },
+    {
+      "epoch": 1.2186012449652142,
+      "grad_norm": 0.24781333667809402,
+      "learning_rate": 7.92839700837718e-06,
+      "loss": 0.0847,
+      "step": 832
+    },
+    {
+      "epoch": 1.220065909923105,
+      "grad_norm": 0.2597660565925631,
+      "learning_rate": 7.903355397952557e-06,
+      "loss": 0.0847,
+      "step": 833
+    },
+    {
+      "epoch": 1.221530574880996,
+      "grad_norm": 0.25108362024888203,
+      "learning_rate": 7.878327532225964e-06,
+      "loss": 0.0729,
+      "step": 834
+    },
+    {
+      "epoch": 1.2229952398388868,
+      "grad_norm": 0.2879443475536739,
+      "learning_rate": 7.853313575269306e-06,
+      "loss": 0.0926,
+      "step": 835
+    },
+    {
+      "epoch": 1.2244599047967777,
+      "grad_norm": 0.24985250820254382,
+      "learning_rate": 7.828313691063294e-06,
+      "loss": 0.0729,
+      "step": 836
+    },
+    {
+      "epoch": 1.2259245697546686,
+      "grad_norm": 0.2528963890297066,
+      "learning_rate": 7.803328043496394e-06,
+      "loss": 0.069,
+      "step": 837
+    },
+    {
+      "epoch": 1.2273892347125595,
+      "grad_norm": 0.2470937539451322,
+      "learning_rate": 7.778356796363734e-06,
+      "loss": 0.0829,
+      "step": 838
+    },
+    {
+      "epoch": 1.2288538996704503,
+      "grad_norm": 0.2521378561935959,
+      "learning_rate": 7.753400113366052e-06,
+      "loss": 0.0791,
+      "step": 839
+    },
+    {
+      "epoch": 1.2303185646283412,
+      "grad_norm": 0.2631772936605158,
+      "learning_rate": 7.728458158108592e-06,
+      "loss": 0.0905,
+      "step": 840
+    },
+    {
+      "epoch": 1.231783229586232,
+      "grad_norm": 0.2444159483676253,
+      "learning_rate": 7.703531094100068e-06,
+      "loss": 0.0741,
+      "step": 841
+    },
+    {
+      "epoch": 1.233247894544123,
+      "grad_norm": 0.281794329109641,
+      "learning_rate": 7.678619084751554e-06,
+      "loss": 0.088,
+      "step": 842
+    },
+    {
+      "epoch": 1.2347125595020139,
+      "grad_norm": 0.27262976373807446,
+      "learning_rate": 7.653722293375453e-06,
+      "loss": 0.1,
+      "step": 843
+    },
+    {
+      "epoch": 1.2361772244599047,
+      "grad_norm": 0.25816315241450866,
+      "learning_rate": 7.628840883184385e-06,
+      "loss": 0.0906,
+      "step": 844
+    },
+    {
+      "epoch": 1.2376418894177956,
+      "grad_norm": 0.2537369926820909,
+      "learning_rate": 7.603975017290159e-06,
+      "loss": 0.0777,
+      "step": 845
+    },
+    {
+      "epoch": 1.2391065543756865,
+      "grad_norm": 0.2916357929163886,
+      "learning_rate": 7.579124858702658e-06,
+      "loss": 0.1011,
+      "step": 846
+    },
+    {
+      "epoch": 1.2405712193335774,
+      "grad_norm": 0.27484734516089476,
+      "learning_rate": 7.5542905703288175e-06,
+      "loss": 0.0954,
+      "step": 847
+    },
+    {
+      "epoch": 1.2420358842914683,
+      "grad_norm": 0.2704547563666006,
+      "learning_rate": 7.529472314971522e-06,
+      "loss": 0.0929,
+      "step": 848
+    },
+    {
+      "epoch": 1.2435005492493592,
+      "grad_norm": 0.24275302504305166,
+      "learning_rate": 7.504670255328548e-06,
+      "loss": 0.0684,
+      "step": 849
+    },
+    {
+      "epoch": 1.24496521420725,
+      "grad_norm": 0.2535609358202711,
+      "learning_rate": 7.4798845539915126e-06,
+      "loss": 0.071,
+      "step": 850
+    },
+    {
+      "epoch": 1.246429879165141,
+      "grad_norm": 0.25442354937020867,
+      "learning_rate": 7.455115373444779e-06,
+      "loss": 0.0808,
+      "step": 851
+    },
+    {
+      "epoch": 1.2478945441230318,
+      "grad_norm": 0.25458587191704724,
+      "learning_rate": 7.430362876064424e-06,
+      "loss": 0.0806,
+      "step": 852
+    },
+    {
+      "epoch": 1.2493592090809227,
+      "grad_norm": 0.2636459021351091,
+      "learning_rate": 7.4056272241171425e-06,
+      "loss": 0.0975,
+      "step": 853
+    },
+    {
+      "epoch": 1.2508238740388136,
+      "grad_norm": 0.26588179392313827,
+      "learning_rate": 7.380908579759207e-06,
+      "loss": 0.0908,
+      "step": 854
+    },
+    {
+      "epoch": 1.2522885389967044,
+      "grad_norm": 0.25802097833264015,
+      "learning_rate": 7.356207105035389e-06,
+      "loss": 0.0727,
+      "step": 855
+    },
+    {
+      "epoch": 1.2537532039545953,
+      "grad_norm": 0.2525429271439887,
+      "learning_rate": 7.331522961877914e-06,
+      "loss": 0.0852,
+      "step": 856
+    },
+    {
+      "epoch": 1.2552178689124862,
+      "grad_norm": 0.24515558724346895,
+      "learning_rate": 7.30685631210537e-06,
+      "loss": 0.0716,
+      "step": 857
+    },
+    {
+      "epoch": 1.256682533870377,
+      "grad_norm": 0.25450544202115294,
+      "learning_rate": 7.282207317421691e-06,
+      "loss": 0.0851,
+      "step": 858
+    },
+    {
+      "epoch": 1.258147198828268,
+      "grad_norm": 0.26104307630858015,
+      "learning_rate": 7.2575761394150476e-06,
+      "loss": 0.0824,
+      "step": 859
+    },
+    {
+      "epoch": 1.2596118637861589,
+      "grad_norm": 0.26362892269219795,
+      "learning_rate": 7.232962939556831e-06,
+      "loss": 0.0754,
+      "step": 860
+    },
+    {
+      "epoch": 1.2610765287440497,
+      "grad_norm": 0.26960831313830425,
+      "learning_rate": 7.20836787920056e-06,
+      "loss": 0.088,
+      "step": 861
+    },
+    {
+      "epoch": 1.2625411937019406,
+      "grad_norm": 0.2884650032848352,
+      "learning_rate": 7.183791119580854e-06,
+      "loss": 0.0844,
+      "step": 862
+    },
+    {
+      "epoch": 1.2640058586598315,
+      "grad_norm": 0.2673739678569061,
+      "learning_rate": 7.159232821812348e-06,
+      "loss": 0.0917,
+      "step": 863
+    },
+    {
+      "epoch": 1.2654705236177224,
+      "grad_norm": 0.2647717036142495,
+      "learning_rate": 7.134693146888652e-06,
+      "loss": 0.074,
+      "step": 864
+    },
+    {
+      "epoch": 1.2669351885756133,
+      "grad_norm": 0.24828043520372733,
+      "learning_rate": 7.1101722556813e-06,
+      "loss": 0.0731,
+      "step": 865
+    },
+    {
+      "epoch": 1.2683998535335042,
+      "grad_norm": 0.29021631743576276,
+      "learning_rate": 7.085670308938674e-06,
+      "loss": 0.0943,
+      "step": 866
+    },
+    {
+      "epoch": 1.269864518491395,
+      "grad_norm": 0.25757971293739884,
+      "learning_rate": 7.061187467284985e-06,
+      "loss": 0.0762,
+      "step": 867
+    },
+    {
+      "epoch": 1.271329183449286,
+      "grad_norm": 0.2707115051349976,
+      "learning_rate": 7.0367238912191734e-06,
+      "loss": 0.086,
+      "step": 868
+    },
+    {
+      "epoch": 1.2727938484071768,
+      "grad_norm": 0.2416159349328539,
+      "learning_rate": 7.012279741113909e-06,
+      "loss": 0.0731,
+      "step": 869
+    },
+    {
+      "epoch": 1.2742585133650677,
+      "grad_norm": 0.25001307113728843,
+      "learning_rate": 6.987855177214489e-06,
+      "loss": 0.0793,
+      "step": 870
+    },
+    {
+      "epoch": 1.2757231783229586,
+      "grad_norm": 0.267697774653143,
+      "learning_rate": 6.963450359637835e-06,
+      "loss": 0.0881,
+      "step": 871
+    },
+    {
+      "epoch": 1.2771878432808494,
+      "grad_norm": 0.2656180743033591,
+      "learning_rate": 6.939065448371398e-06,
+      "loss": 0.0805,
+      "step": 872
+    },
+    {
+      "epoch": 1.2786525082387403,
+      "grad_norm": 0.24388348217463868,
+      "learning_rate": 6.914700603272151e-06,
+      "loss": 0.0739,
+      "step": 873
+    },
+    {
+      "epoch": 1.2801171731966312,
+      "grad_norm": 0.27460701489232753,
+      "learning_rate": 6.8903559840655075e-06,
+      "loss": 0.079,
+      "step": 874
+    },
+    {
+      "epoch": 1.281581838154522,
+      "grad_norm": 0.25551151990608106,
+      "learning_rate": 6.866031750344302e-06,
+      "loss": 0.0767,
+      "step": 875
+    },
+    {
+      "epoch": 1.283046503112413,
+      "grad_norm": 0.25705060856070827,
+      "learning_rate": 6.841728061567713e-06,
+      "loss": 0.084,
+      "step": 876
+    },
+    {
+      "epoch": 1.2845111680703039,
+      "grad_norm": 0.2535751605501693,
+      "learning_rate": 6.817445077060256e-06,
+      "loss": 0.0859,
+      "step": 877
+    },
+    {
+      "epoch": 1.2859758330281947,
+      "grad_norm": 0.26041563534508294,
+      "learning_rate": 6.7931829560107e-06,
+      "loss": 0.0841,
+      "step": 878
+    },
+    {
+      "epoch": 1.2874404979860856,
+      "grad_norm": 0.2537135436483843,
+      "learning_rate": 6.768941857471054e-06,
+      "loss": 0.0775,
+      "step": 879
+    },
+    {
+      "epoch": 1.2889051629439765,
+      "grad_norm": 0.2659246611826128,
+      "learning_rate": 6.744721940355508e-06,
+      "loss": 0.0807,
+      "step": 880
+    },
+    {
+      "epoch": 1.2903698279018674,
+      "grad_norm": 0.2540825283987494,
+      "learning_rate": 6.720523363439393e-06,
+      "loss": 0.0896,
+      "step": 881
+    },
+    {
+      "epoch": 1.2918344928597583,
+      "grad_norm": 0.2575386822036422,
+      "learning_rate": 6.69634628535815e-06,
+      "loss": 0.0797,
+      "step": 882
+    },
+    {
+      "epoch": 1.2932991578176491,
+      "grad_norm": 0.2649389460574201,
+      "learning_rate": 6.672190864606276e-06,
+      "loss": 0.0878,
+      "step": 883
+    },
+    {
+      "epoch": 1.29476382277554,
+      "grad_norm": 0.2638636466333762,
+      "learning_rate": 6.648057259536297e-06,
+      "loss": 0.0831,
+      "step": 884
+    },
+    {
+      "epoch": 1.296228487733431,
+      "grad_norm": 0.26545840639974744,
+      "learning_rate": 6.623945628357718e-06,
+      "loss": 0.081,
+      "step": 885
+    },
+    {
+      "epoch": 1.2976931526913218,
+      "grad_norm": 0.27477957230836636,
+      "learning_rate": 6.599856129136006e-06,
+      "loss": 0.0787,
+      "step": 886
+    },
+    {
+      "epoch": 1.2991578176492127,
+      "grad_norm": 0.2584584388897469,
+      "learning_rate": 6.575788919791522e-06,
+      "loss": 0.0888,
+      "step": 887
+    },
+    {
+      "epoch": 1.3006224826071036,
+      "grad_norm": 0.2553038074640935,
+      "learning_rate": 6.551744158098521e-06,
+      "loss": 0.0804,
+      "step": 888
+    },
+    {
+      "epoch": 1.3020871475649944,
+      "grad_norm": 0.2614338959940162,
+      "learning_rate": 6.527722001684087e-06,
+      "loss": 0.0842,
+      "step": 889
+    },
+    {
+      "epoch": 1.3035518125228853,
+      "grad_norm": 0.27266050110549406,
+      "learning_rate": 6.503722608027129e-06,
+      "loss": 0.0953,
+      "step": 890
+    },
+    {
+      "epoch": 1.3050164774807762,
+      "grad_norm": 0.26498428407714436,
+      "learning_rate": 6.479746134457319e-06,
+      "loss": 0.0948,
+      "step": 891
+    },
+    {
+      "epoch": 1.306481142438667,
+      "grad_norm": 0.2425181307556485,
+      "learning_rate": 6.455792738154074e-06,
+      "loss": 0.0771,
+      "step": 892
+    },
+    {
+      "epoch": 1.307945807396558,
+      "grad_norm": 0.3031551149501075,
+      "learning_rate": 6.431862576145546e-06,
+      "loss": 0.1127,
+      "step": 893
+    },
+    {
+      "epoch": 1.3094104723544489,
+      "grad_norm": 0.2605793802598279,
+      "learning_rate": 6.407955805307547e-06,
+      "loss": 0.0838,
+      "step": 894
+    },
+    {
+      "epoch": 1.3108751373123397,
+      "grad_norm": 0.2857025614498841,
+      "learning_rate": 6.3840725823625685e-06,
+      "loss": 0.0958,
+      "step": 895
+    },
+    {
+      "epoch": 1.3123398022702306,
+      "grad_norm": 0.26266676335382666,
+      "learning_rate": 6.3602130638787155e-06,
+      "loss": 0.0919,
+      "step": 896
+    },
+    {
+      "epoch": 1.3138044672281215,
+      "grad_norm": 0.2705157836138465,
+      "learning_rate": 6.336377406268712e-06,
+      "loss": 0.0843,
+      "step": 897
+    },
+    {
+      "epoch": 1.3152691321860124,
+      "grad_norm": 0.2442729301782896,
+      "learning_rate": 6.312565765788843e-06,
+      "loss": 0.0696,
+      "step": 898
+    },
+    {
+      "epoch": 1.3167337971439033,
+      "grad_norm": 0.23540342455420937,
+      "learning_rate": 6.288778298537966e-06,
+      "loss": 0.0736,
+      "step": 899
+    },
+    {
+      "epoch": 1.3181984621017941,
+      "grad_norm": 0.2772712225095351,
+      "learning_rate": 6.2650151604564534e-06,
+      "loss": 0.0938,
+      "step": 900
+    },
+    {
+      "epoch": 1.319663127059685,
+      "grad_norm": 0.25309744480957075,
+      "learning_rate": 6.241276507325198e-06,
+      "loss": 0.0758,
+      "step": 901
+    },
+    {
+      "epoch": 1.321127792017576,
+      "grad_norm": 0.25863367628675915,
+      "learning_rate": 6.217562494764569e-06,
+      "loss": 0.0833,
+      "step": 902
+    },
+    {
+      "epoch": 1.3225924569754668,
+      "grad_norm": 0.26530513213868256,
+      "learning_rate": 6.193873278233417e-06,
+      "loss": 0.0841,
+      "step": 903
+    },
+    {
+      "epoch": 1.3240571219333577,
+      "grad_norm": 0.24979685098439777,
+      "learning_rate": 6.170209013028021e-06,
+      "loss": 0.0876,
+      "step": 904
+    },
+    {
+      "epoch": 1.3255217868912486,
+      "grad_norm": 0.2523647371099479,
+      "learning_rate": 6.146569854281111e-06,
+      "loss": 0.082,
+      "step": 905
+    },
+    {
+      "epoch": 1.3269864518491394,
+      "grad_norm": 0.2666393886883001,
+      "learning_rate": 6.1229559569608144e-06,
+      "loss": 0.0982,
+      "step": 906
+    },
+    {
+      "epoch": 1.3284511168070303,
+      "grad_norm": 0.25240352584777037,
+      "learning_rate": 6.099367475869658e-06,
+      "loss": 0.0815,
+      "step": 907
+    },
+    {
+      "epoch": 1.3299157817649212,
+      "grad_norm": 0.24034010865003308,
+      "learning_rate": 6.075804565643562e-06,
+      "loss": 0.0732,
+      "step": 908
+    },
+    {
+      "epoch": 1.331380446722812,
+      "grad_norm": 0.2601774368496189,
+      "learning_rate": 6.052267380750796e-06,
+      "loss": 0.0804,
+      "step": 909
+    },
+    {
+      "epoch": 1.332845111680703,
+      "grad_norm": 0.2416658354239181,
+      "learning_rate": 6.028756075491007e-06,
+      "loss": 0.0671,
+      "step": 910
+    },
+    {
+      "epoch": 1.3343097766385938,
+      "grad_norm": 0.23420758822988802,
+      "learning_rate": 6.005270803994165e-06,
+      "loss": 0.0694,
+      "step": 911
+    },
+    {
+      "epoch": 1.3357744415964847,
+      "grad_norm": 0.23323493701217127,
+      "learning_rate": 5.981811720219593e-06,
+      "loss": 0.0677,
+      "step": 912
+    },
+    {
+      "epoch": 1.3372391065543756,
+      "grad_norm": 0.26506406252750964,
+      "learning_rate": 5.958378977954922e-06,
+      "loss": 0.0741,
+      "step": 913
+    },
+    {
+      "epoch": 1.3387037715122665,
+      "grad_norm": 0.255574393142094,
+      "learning_rate": 5.934972730815115e-06,
+      "loss": 0.0811,
+      "step": 914
+    },
+    {
+      "epoch": 1.3401684364701574,
+      "grad_norm": 0.2550215436110789,
+      "learning_rate": 5.9115931322414285e-06,
+      "loss": 0.0862,
+      "step": 915
+    },
+    {
+      "epoch": 1.3416331014280483,
+      "grad_norm": 0.25209588662794513,
+      "learning_rate": 5.888240335500439e-06,
+      "loss": 0.0721,
+      "step": 916
+    },
+    {
+      "epoch": 1.3430977663859391,
+      "grad_norm": 0.26335632891089883,
+      "learning_rate": 5.8649144936830045e-06,
+      "loss": 0.0749,
+      "step": 917
+    },
+    {
+      "epoch": 1.34456243134383,
+      "grad_norm": 0.26611291985369345,
+      "learning_rate": 5.841615759703296e-06,
+      "loss": 0.0856,
+      "step": 918
+    },
+    {
+      "epoch": 1.346027096301721,
+      "grad_norm": 0.3180398065874917,
+      "learning_rate": 5.818344286297756e-06,
+      "loss": 0.1111,
+      "step": 919
+    },
+    {
+      "epoch": 1.3474917612596118,
+      "grad_norm": 0.2470882409448809,
+      "learning_rate": 5.795100226024145e-06,
+      "loss": 0.0842,
+      "step": 920
+    },
+    {
+      "epoch": 1.3489564262175027,
+      "grad_norm": 0.24747051474580573,
+      "learning_rate": 5.771883731260492e-06,
+      "loss": 0.0849,
+      "step": 921
+    },
+    {
+      "epoch": 1.3504210911753936,
+      "grad_norm": 0.2689238693234873,
+      "learning_rate": 5.748694954204126e-06,
+      "loss": 0.09,
+      "step": 922
+    },
+    {
+      "epoch": 1.3518857561332844,
+      "grad_norm": 0.24131459800140173,
+      "learning_rate": 5.7255340468706776e-06,
+      "loss": 0.0788,
+      "step": 923
+    },
+    {
+      "epoch": 1.3533504210911753,
+      "grad_norm": 0.23703584143200643,
+      "learning_rate": 5.702401161093061e-06,
+      "loss": 0.07,
+      "step": 924
+    },
+    {
+      "epoch": 1.3548150860490662,
+      "grad_norm": 0.24911481329061927,
+      "learning_rate": 5.679296448520509e-06,
+      "loss": 0.0766,
+      "step": 925
+    },
+    {
+      "epoch": 1.356279751006957,
+      "grad_norm": 0.2493171773305748,
+      "learning_rate": 5.65622006061755e-06,
+      "loss": 0.0743,
+      "step": 926
+    },
+    {
+      "epoch": 1.357744415964848,
+      "grad_norm": 0.25543638955248743,
+      "learning_rate": 5.633172148663039e-06,
+      "loss": 0.0869,
+      "step": 927
+    },
+    {
+      "epoch": 1.3592090809227388,
+      "grad_norm": 0.24551184220415498,
+      "learning_rate": 5.610152863749143e-06,
+      "loss": 0.069,
+      "step": 928
+    },
+    {
+      "epoch": 1.3606737458806297,
+      "grad_norm": 0.26501893240626057,
+      "learning_rate": 5.5871623567803756e-06,
+      "loss": 0.0696,
+      "step": 929
+    },
+    {
+      "epoch": 1.3621384108385206,
+      "grad_norm": 0.27208453202862887,
+      "learning_rate": 5.564200778472583e-06,
+      "loss": 0.0795,
+      "step": 930
+    },
+    {
+      "epoch": 1.3636030757964115,
+      "grad_norm": 0.27329376137848915,
+      "learning_rate": 5.5412682793519765e-06,
+      "loss": 0.0772,
+      "step": 931
+    },
+    {
+      "epoch": 1.3650677407543024,
+      "grad_norm": 0.2631020886292811,
+      "learning_rate": 5.5183650097541295e-06,
+      "loss": 0.0801,
+      "step": 932
+    },
+    {
+      "epoch": 1.3665324057121933,
+      "grad_norm": 0.24801113517701595,
+      "learning_rate": 5.495491119823007e-06,
+      "loss": 0.0722,
+      "step": 933
+    },
+    {
+      "epoch": 1.3679970706700841,
+      "grad_norm": 0.2650739437868264,
+      "learning_rate": 5.472646759509963e-06,
+      "loss": 0.0936,
+      "step": 934
+    },
+    {
+      "epoch": 1.369461735627975,
+      "grad_norm": 0.2605363396379517,
+      "learning_rate": 5.449832078572781e-06,
+      "loss": 0.0878,
+      "step": 935
+    },
+    {
+      "epoch": 1.370926400585866,
+      "grad_norm": 0.2646339088927977,
+      "learning_rate": 5.427047226574671e-06,
+      "loss": 0.0759,
+      "step": 936
+    },
+    {
+      "epoch": 1.3723910655437568,
+      "grad_norm": 0.2601636184080717,
+      "learning_rate": 5.404292352883291e-06,
+      "loss": 0.0944,
+      "step": 937
+    },
+    {
+      "epoch": 1.3738557305016477,
+      "grad_norm": 0.2837985033272464,
+      "learning_rate": 5.3815676066697946e-06,
+      "loss": 0.1013,
+      "step": 938
+    },
+    {
+      "epoch": 1.3753203954595388,
+      "grad_norm": 0.284455217744833,
+      "learning_rate": 5.358873136907808e-06,
+      "loss": 0.0788,
+      "step": 939
+    },
+    {
+      "epoch": 1.3767850604174297,
+      "grad_norm": 0.26876984673679605,
+      "learning_rate": 5.336209092372502e-06,
+      "loss": 0.0902,
+      "step": 940
+    },
+    {
+      "epoch": 1.3782497253753205,
+      "grad_norm": 0.27988318639906634,
+      "learning_rate": 5.313575621639568e-06,
+      "loss": 0.0857,
+      "step": 941
+    },
+    {
+      "epoch": 1.3797143903332114,
+      "grad_norm": 0.29898311525846655,
+      "learning_rate": 5.290972873084296e-06,
+      "loss": 0.0876,
+      "step": 942
+    },
+    {
+      "epoch": 1.3811790552911023,
+      "grad_norm": 0.2392250599528443,
+      "learning_rate": 5.268400994880547e-06,
+      "loss": 0.0688,
+      "step": 943
+    },
+    {
+      "epoch": 1.3826437202489932,
+      "grad_norm": 0.24702025812837353,
+      "learning_rate": 5.245860134999831e-06,
+      "loss": 0.0763,
+      "step": 944
+    },
+    {
+      "epoch": 1.384108385206884,
+      "grad_norm": 0.25177865615717077,
+      "learning_rate": 5.223350441210303e-06,
+      "loss": 0.0804,
+      "step": 945
+    },
+    {
+      "epoch": 1.385573050164775,
+      "grad_norm": 0.27025384321008283,
+      "learning_rate": 5.200872061075814e-06,
+      "loss": 0.0865,
+      "step": 946
+    },
+    {
+      "epoch": 1.3870377151226658,
+      "grad_norm": 0.3196326325850145,
+      "learning_rate": 5.178425141954926e-06,
+      "loss": 0.1001,
+      "step": 947
+    },
+    {
+      "epoch": 1.3885023800805567,
+      "grad_norm": 0.2659528533890264,
+      "learning_rate": 5.15600983099997e-06,
+      "loss": 0.0757,
+      "step": 948
+    },
+    {
+      "epoch": 1.3899670450384476,
+      "grad_norm": 0.25677841458575923,
+      "learning_rate": 5.133626275156055e-06,
+      "loss": 0.0655,
+      "step": 949
+    },
+    {
+      "epoch": 1.3914317099963385,
+      "grad_norm": 0.2847121257417236,
+      "learning_rate": 5.111274621160127e-06,
+      "loss": 0.0866,
+      "step": 950
+    },
+    {
+      "epoch": 1.3928963749542294,
+      "grad_norm": 0.27638736754038373,
+      "learning_rate": 5.088955015539989e-06,
+      "loss": 0.1017,
+      "step": 951
+    },
+    {
+      "epoch": 1.3943610399121202,
+      "grad_norm": 0.24594995873570738,
+      "learning_rate": 5.06666760461335e-06,
+      "loss": 0.0816,
+      "step": 952
+    },
+    {
+      "epoch": 1.3958257048700111,
+      "grad_norm": 0.24663161576513942,
+      "learning_rate": 5.044412534486873e-06,
+      "loss": 0.0721,
+      "step": 953
+    },
+    {
+      "epoch": 1.397290369827902,
+      "grad_norm": 0.24164833719808607,
+      "learning_rate": 5.0221899510551965e-06,
+      "loss": 0.0617,
+      "step": 954
+    },
+    {
+      "epoch": 1.3987550347857929,
+      "grad_norm": 0.2772777632729084,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.0837,
+      "step": 955
+    },
+    {
+      "epoch": 1.4002196997436838,
+      "grad_norm": 0.25527189944389456,
+      "learning_rate": 4.977842826789034e-06,
+      "loss": 0.0812,
+      "step": 956
+    },
+    {
+      "epoch": 1.4016843647015746,
+      "grad_norm": 0.26722398889539783,
+      "learning_rate": 4.955718576675176e-06,
+      "loss": 0.0762,
+      "step": 957
+    },
+    {
+      "epoch": 1.4031490296594655,
+      "grad_norm": 0.2575445402991656,
+      "learning_rate": 4.933627394695464e-06,
+      "loss": 0.0761,
+      "step": 958
+    },
+    {
+      "epoch": 1.4046136946173564,
+      "grad_norm": 0.26548423508114666,
+      "learning_rate": 4.911569425670168e-06,
+      "loss": 0.0797,
+      "step": 959
+    },
+    {
+      "epoch": 1.4060783595752473,
+      "grad_norm": 0.23644665405104945,
+      "learning_rate": 4.88954481420182e-06,
+      "loss": 0.073,
+      "step": 960
+    },
+    {
+      "epoch": 1.4075430245331382,
+      "grad_norm": 0.25187795846556027,
+      "learning_rate": 4.867553704674279e-06,
+      "loss": 0.0872,
+      "step": 961
+    },
+    {
+      "epoch": 1.409007689491029,
+      "grad_norm": 0.2698865546073546,
+      "learning_rate": 4.845596241251773e-06,
+      "loss": 0.0895,
+      "step": 962
+    },
+    {
+      "epoch": 1.41047235444892,
+      "grad_norm": 0.24104521653041425,
+      "learning_rate": 4.823672567877973e-06,
+      "loss": 0.0758,
+      "step": 963
+    },
+    {
+      "epoch": 1.4119370194068108,
+      "grad_norm": 0.2442657706715064,
+      "learning_rate": 4.801782828275019e-06,
+      "loss": 0.0702,
+      "step": 964
+    },
+    {
+      "epoch": 1.4134016843647017,
+      "grad_norm": 0.2592721496347663,
+      "learning_rate": 4.779927165942616e-06,
+      "loss": 0.0702,
+      "step": 965
+    },
+    {
+      "epoch": 1.4148663493225926,
+      "grad_norm": 0.249769758106348,
+      "learning_rate": 4.758105724157058e-06,
+      "loss": 0.0717,
+      "step": 966
+    },
+    {
+      "epoch": 1.4163310142804835,
+      "grad_norm": 0.30631601628088784,
+      "learning_rate": 4.7363186459703055e-06,
+      "loss": 0.0855,
+      "step": 967
+    },
+    {
+      "epoch": 1.4177956792383744,
+      "grad_norm": 0.2580744961558955,
+      "learning_rate": 4.714566074209058e-06,
+      "loss": 0.0815,
+      "step": 968
+    },
+    {
+      "epoch": 1.4192603441962652,
+      "grad_norm": 0.27175417696412074,
+      "learning_rate": 4.692848151473789e-06,
+      "loss": 0.0884,
+      "step": 969
+    },
+    {
+      "epoch": 1.4207250091541561,
+      "grad_norm": 0.26107195316664245,
+      "learning_rate": 4.671165020137844e-06,
+      "loss": 0.0914,
+      "step": 970
+    },
+    {
+      "epoch": 1.422189674112047,
+      "grad_norm": 0.2659897433575913,
+      "learning_rate": 4.649516822346474e-06,
+      "loss": 0.0841,
+      "step": 971
+    },
+    {
+      "epoch": 1.4236543390699379,
+      "grad_norm": 0.24447135185611163,
+      "learning_rate": 4.62790370001594e-06,
+      "loss": 0.0761,
+      "step": 972
+    },
+    {
+      "epoch": 1.4251190040278288,
+      "grad_norm": 0.27972979502307804,
+      "learning_rate": 4.606325794832545e-06,
+      "loss": 0.0771,
+      "step": 973
+    },
+    {
+      "epoch": 1.4265836689857196,
+      "grad_norm": 0.2736878498962296,
+      "learning_rate": 4.584783248251738e-06,
+      "loss": 0.0842,
+      "step": 974
+    },
+    {
+      "epoch": 1.4280483339436105,
+      "grad_norm": 0.2837151699593587,
+      "learning_rate": 4.563276201497161e-06,
+      "loss": 0.0929,
+      "step": 975
+    },
+    {
+      "epoch": 1.4295129989015014,
+      "grad_norm": 0.2517370953490891,
+      "learning_rate": 4.5418047955597465e-06,
+      "loss": 0.0702,
+      "step": 976
+    },
+    {
+      "epoch": 1.4309776638593923,
+      "grad_norm": 0.28908876741636197,
+      "learning_rate": 4.520369171196766e-06,
+      "loss": 0.0928,
+      "step": 977
+    },
+    {
+      "epoch": 1.4324423288172832,
+      "grad_norm": 0.3619682932427526,
+      "learning_rate": 4.4989694689309394e-06,
+      "loss": 0.086,
+      "step": 978
+    },
+    {
+      "epoch": 1.433906993775174,
+      "grad_norm": 0.28663475644140607,
+      "learning_rate": 4.477605829049479e-06,
+      "loss": 0.0811,
+      "step": 979
+    },
+    {
+      "epoch": 1.435371658733065,
+      "grad_norm": 0.2698934408611363,
+      "learning_rate": 4.456278391603207e-06,
+      "loss": 0.0811,
+      "step": 980
+    },
+    {
+      "epoch": 1.4368363236909558,
+      "grad_norm": 0.2663708566419197,
+      "learning_rate": 4.434987296405602e-06,
+      "loss": 0.106,
+      "step": 981
+    },
+    {
+      "epoch": 1.4383009886488467,
+      "grad_norm": 0.24655631319842156,
+      "learning_rate": 4.413732683031901e-06,
+      "loss": 0.078,
+      "step": 982
+    },
+    {
+      "epoch": 1.4397656536067376,
+      "grad_norm": 0.24077565145998617,
+      "learning_rate": 4.392514690818194e-06,
+      "loss": 0.0695,
+      "step": 983
+    },
+    {
+      "epoch": 1.4412303185646285,
+      "grad_norm": 0.2717847069075042,
+      "learning_rate": 4.37133345886048e-06,
+      "loss": 0.0902,
+      "step": 984
+    },
+    {
+      "epoch": 1.4426949835225193,
+      "grad_norm": 0.2318112119726766,
+      "learning_rate": 4.350189126013793e-06,
+      "loss": 0.0641,
+      "step": 985
+    },
+    {
+      "epoch": 1.4441596484804102,
+      "grad_norm": 0.26503499304444195,
+      "learning_rate": 4.329081830891253e-06,
+      "loss": 0.0967,
+      "step": 986
+    },
+    {
+      "epoch": 1.4456243134383011,
+      "grad_norm": 0.2691482856135504,
+      "learning_rate": 4.308011711863196e-06,
+      "loss": 0.0911,
+      "step": 987
+    },
+    {
+      "epoch": 1.447088978396192,
+      "grad_norm": 0.27532174581408686,
+      "learning_rate": 4.28697890705623e-06,
+      "loss": 0.0885,
+      "step": 988
+    },
+    {
+      "epoch": 1.4485536433540829,
+      "grad_norm": 0.27243518360307484,
+      "learning_rate": 4.265983554352361e-06,
+      "loss": 0.0815,
+      "step": 989
+    },
+    {
+      "epoch": 1.4500183083119738,
+      "grad_norm": 0.25834681516095914,
+      "learning_rate": 4.245025791388063e-06,
+      "loss": 0.0736,
+      "step": 990
+    },
+    {
+      "epoch": 1.4514829732698646,
+      "grad_norm": 0.2850395862995923,
+      "learning_rate": 4.224105755553402e-06,
+      "loss": 0.0851,
+      "step": 991
+    },
+    {
+      "epoch": 1.4529476382277555,
+      "grad_norm": 0.3092877439958859,
+      "learning_rate": 4.203223583991103e-06,
+      "loss": 0.1023,
+      "step": 992
+    },
+    {
+      "epoch": 1.4544123031856464,
+      "grad_norm": 0.26774401365459694,
+      "learning_rate": 4.18237941359569e-06,
+      "loss": 0.0829,
+      "step": 993
+    },
+    {
+      "epoch": 1.4558769681435373,
+      "grad_norm": 0.23800118678712484,
+      "learning_rate": 4.161573381012547e-06,
+      "loss": 0.0676,
+      "step": 994
+    },
+    {
+      "epoch": 1.4573416331014282,
+      "grad_norm": 0.2666429232022083,
+      "learning_rate": 4.140805622637062e-06,
+      "loss": 0.0859,
+      "step": 995
+    },
+    {
+      "epoch": 1.458806298059319,
+      "grad_norm": 0.24644352364865663,
+      "learning_rate": 4.1200762746137e-06,
+      "loss": 0.0761,
+      "step": 996
+    },
+    {
+      "epoch": 1.46027096301721,
+      "grad_norm": 0.2860182896303882,
+      "learning_rate": 4.099385472835128e-06,
+      "loss": 0.1061,
+      "step": 997
+    },
+    {
+      "epoch": 1.4617356279751008,
+      "grad_norm": 0.24601895494668852,
+      "learning_rate": 4.078733352941322e-06,
+      "loss": 0.0889,
+      "step": 998
+    },
+    {
+      "epoch": 1.4632002929329917,
+      "grad_norm": 0.2766785303442865,
+      "learning_rate": 4.05812005031868e-06,
+      "loss": 0.0904,
+      "step": 999
+    },
+    {
+      "epoch": 1.4646649578908826,
+      "grad_norm": 0.2652781699314485,
+      "learning_rate": 4.0375457000991216e-06,
+      "loss": 0.0816,
+      "step": 1000
+    },
+    {
+      "epoch": 1.4661296228487735,
+      "grad_norm": 0.25088853578582626,
+      "learning_rate": 4.01701043715922e-06,
+      "loss": 0.0762,
+      "step": 1001
+    },
+    {
+      "epoch": 1.4675942878066643,
+      "grad_norm": 0.2818469333221271,
+      "learning_rate": 3.996514396119301e-06,
+      "loss": 0.0774,
+      "step": 1002
+    },
+    {
+      "epoch": 1.4690589527645552,
+      "grad_norm": 0.26036857004758346,
+      "learning_rate": 3.976057711342578e-06,
+      "loss": 0.0795,
+      "step": 1003
+    },
+    {
+      "epoch": 1.470523617722446,
+      "grad_norm": 0.27442281645232014,
+      "learning_rate": 3.95564051693425e-06,
+      "loss": 0.07,
+      "step": 1004
+    },
+    {
+      "epoch": 1.471988282680337,
+      "grad_norm": 0.24267887450363582,
+      "learning_rate": 3.935262946740648e-06,
+      "loss": 0.0756,
+      "step": 1005
+    },
+    {
+      "epoch": 1.4734529476382279,
+      "grad_norm": 0.2511350381009027,
+      "learning_rate": 3.914925134348328e-06,
+      "loss": 0.069,
+      "step": 1006
+    },
+    {
+      "epoch": 1.4749176125961188,
+      "grad_norm": 0.2342240326378387,
+      "learning_rate": 3.894627213083227e-06,
+      "loss": 0.0688,
+      "step": 1007
+    },
+    {
+      "epoch": 1.4763822775540096,
+      "grad_norm": 0.23553774309732706,
+      "learning_rate": 3.874369316009759e-06,
+      "loss": 0.0744,
+      "step": 1008
+    },
+    {
+      "epoch": 1.4778469425119005,
+      "grad_norm": 0.23926474659226465,
+      "learning_rate": 3.85415157592997e-06,
+      "loss": 0.0734,
+      "step": 1009
+    },
+    {
+      "epoch": 1.4793116074697914,
+      "grad_norm": 0.268715175607369,
+      "learning_rate": 3.833974125382639e-06,
+      "loss": 0.0723,
+      "step": 1010
+    },
+    {
+      "epoch": 1.4807762724276823,
+      "grad_norm": 0.2662279770589081,
+      "learning_rate": 3.8138370966424386e-06,
+      "loss": 0.0717,
+      "step": 1011
+    },
+    {
+      "epoch": 1.4822409373855732,
+      "grad_norm": 0.2737509205481138,
+      "learning_rate": 3.793740621719042e-06,
+      "loss": 0.0814,
+      "step": 1012
+    },
+    {
+      "epoch": 1.483705602343464,
+      "grad_norm": 0.2639004103415663,
+      "learning_rate": 3.7736848323562803e-06,
+      "loss": 0.091,
+      "step": 1013
+    },
+    {
+      "epoch": 1.485170267301355,
+      "grad_norm": 0.24888682745157728,
+      "learning_rate": 3.753669860031254e-06,
+      "loss": 0.0791,
+      "step": 1014
+    },
+    {
+      "epoch": 1.4866349322592458,
+      "grad_norm": 0.2768808484904024,
+      "learning_rate": 3.7336958359534992e-06,
+      "loss": 0.1019,
+      "step": 1015
+    },
+    {
+      "epoch": 1.4880995972171367,
+      "grad_norm": 0.24635968007971593,
+      "learning_rate": 3.7137628910640997e-06,
+      "loss": 0.0656,
+      "step": 1016
+    },
+    {
+      "epoch": 1.4895642621750276,
+      "grad_norm": 0.2769850878713628,
+      "learning_rate": 3.693871156034854e-06,
+      "loss": 0.0824,
+      "step": 1017
+    },
+    {
+      "epoch": 1.4910289271329185,
+      "grad_norm": 0.2618910230484893,
+      "learning_rate": 3.674020761267394e-06,
+      "loss": 0.0749,
+      "step": 1018
+    },
+    {
+      "epoch": 1.4924935920908093,
+      "grad_norm": 0.25597642357105027,
+      "learning_rate": 3.6542118368923562e-06,
+      "loss": 0.0698,
+      "step": 1019
+    },
+    {
+      "epoch": 1.4939582570487002,
+      "grad_norm": 0.2717269361679958,
+      "learning_rate": 3.634444512768501e-06,
+      "loss": 0.083,
+      "step": 1020
+    },
+    {
+      "epoch": 1.495422922006591,
+      "grad_norm": 0.2992590236586871,
+      "learning_rate": 3.61471891848189e-06,
+      "loss": 0.0886,
+      "step": 1021
+    },
+    {
+      "epoch": 1.496887586964482,
+      "grad_norm": 0.2455530601341986,
+      "learning_rate": 3.595035183345007e-06,
+      "loss": 0.0666,
+      "step": 1022
+    },
+    {
+      "epoch": 1.4983522519223729,
+      "grad_norm": 0.2811055696814055,
+      "learning_rate": 3.575393436395941e-06,
+      "loss": 0.0837,
+      "step": 1023
+    },
+    {
+      "epoch": 1.4998169168802638,
+      "grad_norm": 0.25019620839932355,
+      "learning_rate": 3.5557938063975105e-06,
+      "loss": 0.0755,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5012815818381546,
+      "grad_norm": 0.24368791106013798,
+      "learning_rate": 3.5362364218364387e-06,
+      "loss": 0.0731,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5027462467960455,
+      "grad_norm": 0.26573828045849746,
+      "learning_rate": 3.5167214109225113e-06,
+      "loss": 0.0791,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5042109117539364,
+      "grad_norm": 0.2550833556385589,
+      "learning_rate": 3.497248901587721e-06,
+      "loss": 0.08,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5056755767118273,
+      "grad_norm": 0.2639518823077813,
+      "learning_rate": 3.477819021485448e-06,
+      "loss": 0.0782,
+      "step": 1028
+    },
+    {
+      "epoch": 1.5071402416697182,
+      "grad_norm": 0.2646189162558523,
+      "learning_rate": 3.4584318979896028e-06,
+      "loss": 0.0878,
+      "step": 1029
+    },
+    {
+      "epoch": 1.508604906627609,
+      "grad_norm": 0.2641718505669378,
+      "learning_rate": 3.439087658193816e-06,
+      "loss": 0.0936,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5100695715855,
+      "grad_norm": 0.24594190472472738,
+      "learning_rate": 3.4197864289105763e-06,
+      "loss": 0.0778,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5115342365433908,
+      "grad_norm": 0.28815729095628295,
+      "learning_rate": 3.4005283366704268e-06,
+      "loss": 0.0761,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5129989015012817,
+      "grad_norm": 0.2528110175883712,
+      "learning_rate": 3.381313507721111e-06,
+      "loss": 0.0747,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5144635664591726,
+      "grad_norm": 0.2533306883004603,
+      "learning_rate": 3.36214206802677e-06,
+      "loss": 0.0691,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5159282314170635,
+      "grad_norm": 0.25930598769235536,
+      "learning_rate": 3.343014143267089e-06,
+      "loss": 0.0692,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5173928963749543,
+      "grad_norm": 0.25924131927849636,
+      "learning_rate": 3.3239298588365045e-06,
+      "loss": 0.0775,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5188575613328452,
+      "grad_norm": 0.2491634194032525,
+      "learning_rate": 3.304889339843347e-06,
+      "loss": 0.0723,
+      "step": 1037
+    },
+    {
+      "epoch": 1.520322226290736,
+      "grad_norm": 0.26529983131571094,
+      "learning_rate": 3.285892711109059e-06,
+      "loss": 0.0914,
+      "step": 1038
+    },
+    {
+      "epoch": 1.521786891248627,
+      "grad_norm": 0.22521005677274164,
+      "learning_rate": 3.2669400971673425e-06,
+      "loss": 0.0616,
+      "step": 1039
+    },
+    {
+      "epoch": 1.5232515562065179,
+      "grad_norm": 0.24331418307789873,
+      "learning_rate": 3.2480316222633614e-06,
+      "loss": 0.0652,
+      "step": 1040
+    },
+    {
+      "epoch": 1.5247162211644087,
+      "grad_norm": 0.2488037942280336,
+      "learning_rate": 3.229167410352931e-06,
+      "loss": 0.0715,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5261808861222996,
+      "grad_norm": 0.26897121918679934,
+      "learning_rate": 3.210347585101684e-06,
+      "loss": 0.0793,
+      "step": 1042
+    },
+    {
+      "epoch": 1.5276455510801905,
+      "grad_norm": 0.25348059670947104,
+      "learning_rate": 3.1915722698842877e-06,
+      "loss": 0.078,
+      "step": 1043
+    },
+    {
+      "epoch": 1.5291102160380814,
+      "grad_norm": 0.233264660760853,
+      "learning_rate": 3.1728415877836072e-06,
+      "loss": 0.0761,
+      "step": 1044
+    },
+    {
+      "epoch": 1.5305748809959723,
+      "grad_norm": 0.26375032434001777,
+      "learning_rate": 3.154155661589924e-06,
+      "loss": 0.0819,
+      "step": 1045
+    },
+    {
+      "epoch": 1.5320395459538632,
+      "grad_norm": 0.26419620695615686,
+      "learning_rate": 3.135514613800108e-06,
+      "loss": 0.0765,
+      "step": 1046
+    },
+    {
+      "epoch": 1.533504210911754,
+      "grad_norm": 0.23976923188526442,
+      "learning_rate": 3.1169185666168377e-06,
+      "loss": 0.0692,
+      "step": 1047
+    },
+    {
+      "epoch": 1.534968875869645,
+      "grad_norm": 0.2493779944554121,
+      "learning_rate": 3.098367641947775e-06,
+      "loss": 0.0845,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5364335408275358,
+      "grad_norm": 0.29326865343508624,
+      "learning_rate": 3.0798619614047885e-06,
+      "loss": 0.0998,
+      "step": 1049
+    },
+    {
+      "epoch": 1.5378982057854267,
+      "grad_norm": 0.2789752607489951,
+      "learning_rate": 3.061401646303136e-06,
+      "loss": 0.0886,
+      "step": 1050
+    },
+    {
+      "epoch": 1.5393628707433176,
+      "grad_norm": 0.2737662371091596,
+      "learning_rate": 3.04298681766069e-06,
+      "loss": 0.0684,
+      "step": 1051
+    },
+    {
+      "epoch": 1.5408275357012085,
+      "grad_norm": 0.2637024271799521,
+      "learning_rate": 3.024617596197121e-06,
+      "loss": 0.0859,
+      "step": 1052
+    },
+    {
+      "epoch": 1.5422922006590993,
+      "grad_norm": 0.23552670812452456,
+      "learning_rate": 3.0062941023331316e-06,
+      "loss": 0.0639,
+      "step": 1053
+    },
+    {
+      "epoch": 1.5437568656169902,
+      "grad_norm": 0.2442542376702699,
+      "learning_rate": 2.988016456189644e-06,
+      "loss": 0.0721,
+      "step": 1054
+    },
+    {
+      "epoch": 1.545221530574881,
+      "grad_norm": 0.2681121581501204,
+      "learning_rate": 2.9697847775870227e-06,
+      "loss": 0.0781,
+      "step": 1055
+    },
+    {
+      "epoch": 1.546686195532772,
+      "grad_norm": 0.3009772111270036,
+      "learning_rate": 2.9515991860442973e-06,
+      "loss": 0.0887,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5481508604906629,
+      "grad_norm": 0.2438680927400026,
+      "learning_rate": 2.933459800778361e-06,
+      "loss": 0.0708,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5496155254485537,
+      "grad_norm": 0.24192854414225795,
+      "learning_rate": 2.9153667407032073e-06,
+      "loss": 0.0651,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5510801904064446,
+      "grad_norm": 0.25560089457881163,
+      "learning_rate": 2.8973201244291305e-06,
+      "loss": 0.0748,
+      "step": 1059
+    },
+    {
+      "epoch": 1.5525448553643355,
+      "grad_norm": 0.24635545740478473,
+      "learning_rate": 2.879320070261974e-06,
+      "loss": 0.0638,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5540095203222264,
+      "grad_norm": 0.3187176598509699,
+      "learning_rate": 2.861366696202326e-06,
+      "loss": 0.1081,
+      "step": 1061
+    },
+    {
+      "epoch": 1.5554741852801173,
+      "grad_norm": 0.25629735568872725,
+      "learning_rate": 2.8434601199447698e-06,
+      "loss": 0.0769,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5569388502380082,
+      "grad_norm": 0.24990001099936748,
+      "learning_rate": 2.825600458877095e-06,
+      "loss": 0.0898,
+      "step": 1063
+    },
+    {
+      "epoch": 1.558403515195899,
+      "grad_norm": 0.250013809003241,
+      "learning_rate": 2.8077878300795446e-06,
+      "loss": 0.0746,
+      "step": 1064
+    },
+    {
+      "epoch": 1.55986818015379,
+      "grad_norm": 0.25934456236335596,
+      "learning_rate": 2.7900223503240265e-06,
+      "loss": 0.0958,
+      "step": 1065
+    },
+    {
+      "epoch": 1.5613328451116808,
+      "grad_norm": 0.24098659840069045,
+      "learning_rate": 2.7723041360733737e-06,
+      "loss": 0.0695,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5627975100695717,
+      "grad_norm": 0.26301653549643544,
+      "learning_rate": 2.7546333034805528e-06,
+      "loss": 0.0741,
+      "step": 1067
+    },
+    {
+      "epoch": 1.5642621750274626,
+      "grad_norm": 0.2391640316419889,
+      "learning_rate": 2.737009968387929e-06,
+      "loss": 0.0605,
+      "step": 1068
+    },
+    {
+      "epoch": 1.5657268399853534,
+      "grad_norm": 0.25060691825740045,
+      "learning_rate": 2.719434246326487e-06,
+      "loss": 0.0643,
+      "step": 1069
+    },
+    {
+      "epoch": 1.5671915049432443,
+      "grad_norm": 0.274283165127303,
+      "learning_rate": 2.7019062525150783e-06,
+      "loss": 0.0874,
+      "step": 1070
+    },
+    {
+      "epoch": 1.5686561699011352,
+      "grad_norm": 0.2978642107033507,
+      "learning_rate": 2.6844261018596806e-06,
+      "loss": 0.0823,
+      "step": 1071
+    },
+    {
+      "epoch": 1.570120834859026,
+      "grad_norm": 0.2690424055706032,
+      "learning_rate": 2.6669939089526177e-06,
+      "loss": 0.0856,
+      "step": 1072
+    },
+    {
+      "epoch": 1.571585499816917,
+      "grad_norm": 0.28386560171341124,
+      "learning_rate": 2.6496097880718364e-06,
+      "loss": 0.079,
+      "step": 1073
+    },
+    {
+      "epoch": 1.5730501647748079,
+      "grad_norm": 0.2898252725065379,
+      "learning_rate": 2.632273853180132e-06,
+      "loss": 0.0801,
+      "step": 1074
+    },
+    {
+      "epoch": 1.5745148297326987,
+      "grad_norm": 0.2510043284519159,
+      "learning_rate": 2.6149862179244257e-06,
+      "loss": 0.0692,
+      "step": 1075
+    },
+    {
+      "epoch": 1.5759794946905896,
+      "grad_norm": 0.2723382092456178,
+      "learning_rate": 2.5977469956349956e-06,
+      "loss": 0.0956,
+      "step": 1076
+    },
+    {
+      "epoch": 1.5774441596484805,
+      "grad_norm": 0.26220486838825635,
+      "learning_rate": 2.5805562993247536e-06,
+      "loss": 0.0705,
+      "step": 1077
+    },
+    {
+      "epoch": 1.5789088246063714,
+      "grad_norm": 0.2645372898294826,
+      "learning_rate": 2.563414241688489e-06,
+      "loss": 0.0691,
+      "step": 1078
+    },
+    {
+      "epoch": 1.5803734895642623,
+      "grad_norm": 0.25156658959890105,
+      "learning_rate": 2.5463209351021457e-06,
+      "loss": 0.0796,
+      "step": 1079
+    },
+    {
+      "epoch": 1.5818381545221532,
+      "grad_norm": 0.25334520784266373,
+      "learning_rate": 2.529276491622067e-06,
+      "loss": 0.0737,
+      "step": 1080
+    },
+    {
+      "epoch": 1.583302819480044,
+      "grad_norm": 0.23790674931444958,
+      "learning_rate": 2.5122810229842807e-06,
+      "loss": 0.0629,
+      "step": 1081
+    },
+    {
+      "epoch": 1.584767484437935,
+      "grad_norm": 0.2303091761767703,
+      "learning_rate": 2.495334640603746e-06,
+      "loss": 0.0767,
+      "step": 1082
+    },
+    {
+      "epoch": 1.5862321493958258,
+      "grad_norm": 0.2510575661559066,
+      "learning_rate": 2.4784374555736445e-06,
+      "loss": 0.0649,
+      "step": 1083
+    },
+    {
+      "epoch": 1.5876968143537167,
+      "grad_norm": 0.26143738599858646,
+      "learning_rate": 2.4615895786646337e-06,
+      "loss": 0.0831,
+      "step": 1084
+    },
+    {
+      "epoch": 1.5891614793116076,
+      "grad_norm": 0.2567282259841764,
+      "learning_rate": 2.444791120324127e-06,
+      "loss": 0.0754,
+      "step": 1085
+    },
+    {
+      "epoch": 1.5906261442694984,
+      "grad_norm": 0.2730764216157527,
+      "learning_rate": 2.4280421906755814e-06,
+      "loss": 0.0803,
+      "step": 1086
+    },
+    {
+      "epoch": 1.5920908092273893,
+      "grad_norm": 0.25081318067201624,
+      "learning_rate": 2.4113428995177522e-06,
+      "loss": 0.0778,
+      "step": 1087
+    },
+    {
+      "epoch": 1.5935554741852802,
+      "grad_norm": 0.24769756992277367,
+      "learning_rate": 2.394693356323997e-06,
+      "loss": 0.0775,
+      "step": 1088
+    },
+    {
+      "epoch": 1.595020139143171,
+      "grad_norm": 0.2546886460741762,
+      "learning_rate": 2.378093670241538e-06,
+      "loss": 0.0688,
+      "step": 1089
+    },
+    {
+      "epoch": 1.596484804101062,
+      "grad_norm": 0.2563027758513368,
+      "learning_rate": 2.3615439500907657e-06,
+      "loss": 0.0617,
+      "step": 1090
+    },
+    {
+      "epoch": 1.5979494690589529,
+      "grad_norm": 0.25885408016228334,
+      "learning_rate": 2.3450443043645035e-06,
+      "loss": 0.0697,
+      "step": 1091
+    },
+    {
+      "epoch": 1.5994141340168437,
+      "grad_norm": 0.26381952528868263,
+      "learning_rate": 2.3285948412273198e-06,
+      "loss": 0.0732,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6008787989747346,
+      "grad_norm": 0.22722995241678426,
+      "learning_rate": 2.3121956685147995e-06,
+      "loss": 0.0571,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6023434639326255,
+      "grad_norm": 0.2828387588961395,
+      "learning_rate": 2.2958468937328528e-06,
+      "loss": 0.0749,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6038081288905164,
+      "grad_norm": 0.2558084885710104,
+      "learning_rate": 2.279548624056992e-06,
+      "loss": 0.0733,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6052727938484073,
+      "grad_norm": 0.27462713945439543,
+      "learning_rate": 2.263300966331652e-06,
+      "loss": 0.0934,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6067374588062981,
+      "grad_norm": 0.2904102780660436,
+      "learning_rate": 2.247104027069467e-06,
+      "loss": 0.0901,
+      "step": 1097
+    },
+    {
+      "epoch": 1.608202123764189,
+      "grad_norm": 0.2771702213465073,
+      "learning_rate": 2.230957912450592e-06,
+      "loss": 0.08,
+      "step": 1098
+    },
+    {
+      "epoch": 1.60966678872208,
+      "grad_norm": 0.2528112186314324,
+      "learning_rate": 2.214862728321987e-06,
+      "loss": 0.0755,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6111314536799708,
+      "grad_norm": 0.23629870958401772,
+      "learning_rate": 2.1988185801967464e-06,
+      "loss": 0.0687,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6125961186378617,
+      "grad_norm": 0.24410925927191227,
+      "learning_rate": 2.182825573253382e-06,
+      "loss": 0.0727,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6140607835957526,
+      "grad_norm": 0.2561472124227788,
+      "learning_rate": 2.1668838123351566e-06,
+      "loss": 0.0909,
+      "step": 1102
+    },
+    {
+      "epoch": 1.6155254485536434,
+      "grad_norm": 0.2635995106202067,
+      "learning_rate": 2.150993401949376e-06,
+      "loss": 0.076,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6169901135115343,
+      "grad_norm": 0.24617657014500907,
+      "learning_rate": 2.135154446266726e-06,
+      "loss": 0.068,
+      "step": 1104
+    },
+    {
+      "epoch": 1.6184547784694252,
+      "grad_norm": 0.24053480589468268,
+      "learning_rate": 2.119367049120565e-06,
+      "loss": 0.0635,
+      "step": 1105
+    },
+    {
+      "epoch": 1.619919443427316,
+      "grad_norm": 0.24204380335717454,
+      "learning_rate": 2.103631314006267e-06,
+      "loss": 0.0678,
+      "step": 1106
+    },
+    {
+      "epoch": 1.621384108385207,
+      "grad_norm": 0.22690779315880516,
+      "learning_rate": 2.087947344080522e-06,
+      "loss": 0.057,
+      "step": 1107
+    },
+    {
+      "epoch": 1.6228487733430979,
+      "grad_norm": 0.28433243723669543,
+      "learning_rate": 2.0723152421606805e-06,
+      "loss": 0.0951,
+      "step": 1108
+    },
+    {
+      "epoch": 1.6243134383009887,
+      "grad_norm": 0.2599968195648375,
+      "learning_rate": 2.0567351107240563e-06,
+      "loss": 0.0695,
+      "step": 1109
+    },
+    {
+      "epoch": 1.6257781032588796,
+      "grad_norm": 0.26139586892133587,
+      "learning_rate": 2.041207051907279e-06,
+      "loss": 0.078,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6272427682167705,
+      "grad_norm": 0.2885911977842118,
+      "learning_rate": 2.0257311675056025e-06,
+      "loss": 0.0893,
+      "step": 1111
+    },
+    {
+      "epoch": 1.6287074331746614,
+      "grad_norm": 0.23726597176087105,
+      "learning_rate": 2.0103075589722576e-06,
+      "loss": 0.0614,
+      "step": 1112
+    },
+    {
+      "epoch": 1.6301720981325523,
+      "grad_norm": 0.26578035267907674,
+      "learning_rate": 1.9949363274177667e-06,
+      "loss": 0.0921,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6316367630904431,
+      "grad_norm": 0.2749652618636836,
+      "learning_rate": 1.9796175736093027e-06,
+      "loss": 0.0835,
+      "step": 1114
+    },
+    {
+      "epoch": 1.633101428048334,
+      "grad_norm": 0.24658373064000075,
+      "learning_rate": 1.9643513979700035e-06,
+      "loss": 0.0777,
+      "step": 1115
+    },
+    {
+      "epoch": 1.634566093006225,
+      "grad_norm": 0.22932017498786317,
+      "learning_rate": 1.9491379005783405e-06,
+      "loss": 0.0723,
+      "step": 1116
+    },
+    {
+      "epoch": 1.6360307579641158,
+      "grad_norm": 0.2545440028300663,
+      "learning_rate": 1.933977181167439e-06,
+      "loss": 0.076,
+      "step": 1117
+    },
+    {
+      "epoch": 1.6374954229220067,
+      "grad_norm": 0.2552570068416127,
+      "learning_rate": 1.9188693391244438e-06,
+      "loss": 0.0754,
+      "step": 1118
+    },
+    {
+      "epoch": 1.6389600878798976,
+      "grad_norm": 0.2478371358109087,
+      "learning_rate": 1.9038144734898478e-06,
+      "loss": 0.0729,
+      "step": 1119
+    },
+    {
+      "epoch": 1.6404247528377884,
+      "grad_norm": 0.2291241183631655,
+      "learning_rate": 1.8888126829568642e-06,
+      "loss": 0.0634,
+      "step": 1120
+    },
+    {
+      "epoch": 1.6418894177956793,
+      "grad_norm": 0.24193265605324504,
+      "learning_rate": 1.8738640658707585e-06,
+      "loss": 0.0671,
+      "step": 1121
+    },
+    {
+      "epoch": 1.6433540827535702,
+      "grad_norm": 0.2576873945419135,
+      "learning_rate": 1.8589687202282247e-06,
+      "loss": 0.0723,
+      "step": 1122
+    },
+    {
+      "epoch": 1.644818747711461,
+      "grad_norm": 0.24676250477181633,
+      "learning_rate": 1.844126743676722e-06,
+      "loss": 0.0748,
+      "step": 1123
+    },
+    {
+      "epoch": 1.646283412669352,
+      "grad_norm": 0.263090468630097,
+      "learning_rate": 1.8293382335138533e-06,
+      "loss": 0.0752,
+      "step": 1124
+    },
+    {
+      "epoch": 1.6477480776272428,
+      "grad_norm": 0.2540498525445381,
+      "learning_rate": 1.8146032866867114e-06,
+      "loss": 0.0708,
+      "step": 1125
+    },
+    {
+      "epoch": 1.6492127425851337,
+      "grad_norm": 0.2690977824033387,
+      "learning_rate": 1.7999219997912575e-06,
+      "loss": 0.0718,
+      "step": 1126
+    },
+    {
+      "epoch": 1.6506774075430246,
+      "grad_norm": 0.2565788082278319,
+      "learning_rate": 1.7852944690716766e-06,
+      "loss": 0.0834,
+      "step": 1127
+    },
+    {
+      "epoch": 1.6521420725009155,
+      "grad_norm": 0.250344072055428,
+      "learning_rate": 1.7707207904197566e-06,
+      "loss": 0.0648,
+      "step": 1128
+    },
+    {
+      "epoch": 1.6536067374588064,
+      "grad_norm": 0.2531519839830438,
+      "learning_rate": 1.7562010593742496e-06,
+      "loss": 0.0761,
+      "step": 1129
+    },
+    {
+      "epoch": 1.6550714024166973,
+      "grad_norm": 0.2671783705252202,
+      "learning_rate": 1.7417353711202478e-06,
+      "loss": 0.0806,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6565360673745881,
+      "grad_norm": 0.27608727196744226,
+      "learning_rate": 1.7273238204885734e-06,
+      "loss": 0.0776,
+      "step": 1131
+    },
+    {
+      "epoch": 1.658000732332479,
+      "grad_norm": 0.21272809051359043,
+      "learning_rate": 1.7129665019551333e-06,
+      "loss": 0.0543,
+      "step": 1132
+    },
+    {
+      "epoch": 1.65946539729037,
+      "grad_norm": 0.2981100085240527,
+      "learning_rate": 1.6986635096403213e-06,
+      "loss": 0.0917,
+      "step": 1133
+    },
+    {
+      "epoch": 1.6609300622482608,
+      "grad_norm": 0.2283996510993873,
+      "learning_rate": 1.6844149373083852e-06,
+      "loss": 0.0689,
+      "step": 1134
+    },
+    {
+      "epoch": 1.6623947272061517,
+      "grad_norm": 0.2932656999891174,
+      "learning_rate": 1.670220878366826e-06,
+      "loss": 0.0913,
+      "step": 1135
+    },
+    {
+      "epoch": 1.6638593921640425,
+      "grad_norm": 0.24998740552484003,
+      "learning_rate": 1.6560814258657687e-06,
+      "loss": 0.0842,
+      "step": 1136
+    },
+    {
+      "epoch": 1.6653240571219334,
+      "grad_norm": 0.24855056763964203,
+      "learning_rate": 1.6419966724973734e-06,
+      "loss": 0.0734,
+      "step": 1137
+    },
+    {
+      "epoch": 1.6667887220798243,
+      "grad_norm": 0.24369157654216977,
+      "learning_rate": 1.6279667105952057e-06,
+      "loss": 0.0911,
+      "step": 1138
+    },
+    {
+      "epoch": 1.6682533870377152,
+      "grad_norm": 0.26041234841753214,
+      "learning_rate": 1.6139916321336513e-06,
+      "loss": 0.0704,
+      "step": 1139
+    },
+    {
+      "epoch": 1.669718051995606,
+      "grad_norm": 0.2654682154981918,
+      "learning_rate": 1.6000715287272938e-06,
+      "loss": 0.0762,
+      "step": 1140
+    },
+    {
+      "epoch": 1.671182716953497,
+      "grad_norm": 0.2379683222765605,
+      "learning_rate": 1.5862064916303343e-06,
+      "loss": 0.0734,
+      "step": 1141
+    },
+    {
+      "epoch": 1.6726473819113878,
+      "grad_norm": 0.26506248533884014,
+      "learning_rate": 1.5723966117359745e-06,
+      "loss": 0.0673,
+      "step": 1142
+    },
+    {
+      "epoch": 1.6741120468692787,
+      "grad_norm": 0.2380435720292988,
+      "learning_rate": 1.5586419795758356e-06,
+      "loss": 0.0726,
+      "step": 1143
+    },
+    {
+      "epoch": 1.6755767118271696,
+      "grad_norm": 0.23717677968711492,
+      "learning_rate": 1.5449426853193549e-06,
+      "loss": 0.0796,
+      "step": 1144
+    },
+    {
+      "epoch": 1.6770413767850605,
+      "grad_norm": 0.27958169485281625,
+      "learning_rate": 1.5312988187731969e-06,
+      "loss": 0.0855,
+      "step": 1145
+    },
+    {
+      "epoch": 1.6785060417429514,
+      "grad_norm": 0.2568879441056742,
+      "learning_rate": 1.5177104693806721e-06,
+      "loss": 0.0712,
+      "step": 1146
+    },
+    {
+      "epoch": 1.6799707067008423,
+      "grad_norm": 0.22152990710129233,
+      "learning_rate": 1.5041777262211355e-06,
+      "loss": 0.0584,
+      "step": 1147
+    },
+    {
+      "epoch": 1.6814353716587331,
+      "grad_norm": 0.2652525501551631,
+      "learning_rate": 1.4907006780094212e-06,
+      "loss": 0.0697,
+      "step": 1148
+    },
+    {
+      "epoch": 1.682900036616624,
+      "grad_norm": 0.23706028774243557,
+      "learning_rate": 1.4772794130952416e-06,
+      "loss": 0.0556,
+      "step": 1149
+    },
+    {
+      "epoch": 1.684364701574515,
+      "grad_norm": 0.232103427190539,
+      "learning_rate": 1.4639140194626289e-06,
+      "loss": 0.0591,
+      "step": 1150
+    },
+    {
+      "epoch": 1.6858293665324058,
+      "grad_norm": 0.24774373170203481,
+      "learning_rate": 1.450604584729336e-06,
+      "loss": 0.061,
+      "step": 1151
+    },
+    {
+      "epoch": 1.6872940314902967,
+      "grad_norm": 0.24757349813419924,
+      "learning_rate": 1.4373511961462828e-06,
+      "loss": 0.0859,
+      "step": 1152
+    },
+    {
+      "epoch": 1.6887586964481875,
+      "grad_norm": 0.239803460293539,
+      "learning_rate": 1.4241539405969662e-06,
+      "loss": 0.0694,
+      "step": 1153
+    },
+    {
+      "epoch": 1.6902233614060784,
+      "grad_norm": 0.24607947040534622,
+      "learning_rate": 1.411012904596909e-06,
+      "loss": 0.0726,
+      "step": 1154
+    },
+    {
+      "epoch": 1.6916880263639693,
+      "grad_norm": 0.24298503192914125,
+      "learning_rate": 1.3979281742930706e-06,
+      "loss": 0.071,
+      "step": 1155
+    },
+    {
+      "epoch": 1.6931526913218602,
+      "grad_norm": 0.25982515681290586,
+      "learning_rate": 1.3848998354633082e-06,
+      "loss": 0.0663,
+      "step": 1156
+    },
+    {
+      "epoch": 1.694617356279751,
+      "grad_norm": 0.2824274775862737,
+      "learning_rate": 1.3719279735157875e-06,
+      "loss": 0.0815,
+      "step": 1157
+    },
+    {
+      "epoch": 1.696082021237642,
+      "grad_norm": 0.25503279831214637,
+      "learning_rate": 1.359012673488449e-06,
+      "loss": 0.084,
+      "step": 1158
+    },
+    {
+      "epoch": 1.6975466861955328,
+      "grad_norm": 0.2572621551136547,
+      "learning_rate": 1.346154020048428e-06,
+      "loss": 0.0647,
+      "step": 1159
+    },
+    {
+      "epoch": 1.6990113511534237,
+      "grad_norm": 0.26401952613949853,
+      "learning_rate": 1.3333520974915093e-06,
+      "loss": 0.0611,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7004760161113146,
+      "grad_norm": 0.27287208188302337,
+      "learning_rate": 1.320606989741583e-06,
+      "loss": 0.0887,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7019406810692055,
+      "grad_norm": 0.25110794107433626,
+      "learning_rate": 1.307918780350077e-06,
+      "loss": 0.0726,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7034053460270964,
+      "grad_norm": 0.26427102366363486,
+      "learning_rate": 1.2952875524954233e-06,
+      "loss": 0.0744,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7048700109849872,
+      "grad_norm": 0.2577483404898282,
+      "learning_rate": 1.2827133889825039e-06,
+      "loss": 0.0799,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7063346759428781,
+      "grad_norm": 0.2379395103991794,
+      "learning_rate": 1.2701963722421162e-06,
+      "loss": 0.0668,
+      "step": 1165
+    },
+    {
+      "epoch": 1.707799340900769,
+      "grad_norm": 0.24586889905164777,
+      "learning_rate": 1.2577365843304212e-06,
+      "loss": 0.0888,
+      "step": 1166
+    },
+    {
+      "epoch": 1.70926400585866,
+      "grad_norm": 0.23659079747711304,
+      "learning_rate": 1.245334106928422e-06,
+      "loss": 0.06,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7107286708165508,
+      "grad_norm": 0.23685869500201995,
+      "learning_rate": 1.2329890213414063e-06,
+      "loss": 0.0639,
+      "step": 1168
+    },
+    {
+      "epoch": 1.7121933357744417,
+      "grad_norm": 0.2375875883124259,
+      "learning_rate": 1.220701408498438e-06,
+      "loss": 0.0669,
+      "step": 1169
+    },
+    {
+      "epoch": 1.7136580007323325,
+      "grad_norm": 0.26520434490127226,
+      "learning_rate": 1.2084713489518063e-06,
+      "loss": 0.0737,
+      "step": 1170
+    },
+    {
+      "epoch": 1.7151226656902234,
+      "grad_norm": 0.2629329649647797,
+      "learning_rate": 1.19629892287651e-06,
+      "loss": 0.0942,
+      "step": 1171
+    },
+    {
+      "epoch": 1.7165873306481143,
+      "grad_norm": 0.23516934988436172,
+      "learning_rate": 1.1841842100697253e-06,
+      "loss": 0.0551,
+      "step": 1172
+    },
+    {
+      "epoch": 1.7180519956060052,
+      "grad_norm": 0.27697429411981783,
+      "learning_rate": 1.1721272899502856e-06,
+      "loss": 0.0762,
+      "step": 1173
+    },
+    {
+      "epoch": 1.719516660563896,
+      "grad_norm": 0.25426135353111506,
+      "learning_rate": 1.1601282415581627e-06,
+      "loss": 0.0664,
+      "step": 1174
+    },
+    {
+      "epoch": 1.720981325521787,
+      "grad_norm": 0.24626607935465855,
+      "learning_rate": 1.1481871435539415e-06,
+      "loss": 0.0619,
+      "step": 1175
+    },
+    {
+      "epoch": 1.7224459904796778,
+      "grad_norm": 0.25729901031239943,
+      "learning_rate": 1.1363040742183162e-06,
+      "loss": 0.073,
+      "step": 1176
+    },
+    {
+      "epoch": 1.7239106554375687,
+      "grad_norm": 0.2567425743663647,
+      "learning_rate": 1.1244791114515608e-06,
+      "loss": 0.0578,
+      "step": 1177
+    },
+    {
+      "epoch": 1.7253753203954596,
+      "grad_norm": 0.2624890331507839,
+      "learning_rate": 1.1127123327730383e-06,
+      "loss": 0.0695,
+      "step": 1178
+    },
+    {
+      "epoch": 1.7268399853533505,
+      "grad_norm": 0.25775906362459766,
+      "learning_rate": 1.1010038153206703e-06,
+      "loss": 0.0671,
+      "step": 1179
+    },
+    {
+      "epoch": 1.7283046503112414,
+      "grad_norm": 0.2357579800372649,
+      "learning_rate": 1.0893536358504553e-06,
+      "loss": 0.0626,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7297693152691322,
+      "grad_norm": 0.27772530636880377,
+      "learning_rate": 1.0777618707359427e-06,
+      "loss": 0.0768,
+      "step": 1181
+    },
+    {
+      "epoch": 1.7312339802270231,
+      "grad_norm": 0.2383877132505535,
+      "learning_rate": 1.0662285959677499e-06,
+      "loss": 0.0667,
+      "step": 1182
+    },
+    {
+      "epoch": 1.732698645184914,
+      "grad_norm": 0.23742698011306232,
+      "learning_rate": 1.0547538871530482e-06,
+      "loss": 0.0599,
+      "step": 1183
+    },
+    {
+      "epoch": 1.734163310142805,
+      "grad_norm": 0.2587588714854726,
+      "learning_rate": 1.0433378195150889e-06,
+      "loss": 0.1056,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7356279751006958,
+      "grad_norm": 0.2623510431514894,
+      "learning_rate": 1.0319804678926825e-06,
+      "loss": 0.0724,
+      "step": 1185
+    },
+    {
+      "epoch": 1.7370926400585867,
+      "grad_norm": 0.23000218942084621,
+      "learning_rate": 1.0206819067397345e-06,
+      "loss": 0.0631,
+      "step": 1186
+    },
+    {
+      "epoch": 1.7385573050164775,
+      "grad_norm": 0.23524204735507692,
+      "learning_rate": 1.009442210124737e-06,
+      "loss": 0.0642,
+      "step": 1187
+    },
+    {
+      "epoch": 1.7400219699743684,
+      "grad_norm": 0.26310108373605007,
+      "learning_rate": 9.982614517302958e-07,
+      "loss": 0.0732,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7414866349322593,
+      "grad_norm": 0.2822237389406422,
+      "learning_rate": 9.871397048526431e-07,
+      "loss": 0.09,
+      "step": 1189
+    },
+    {
+      "epoch": 1.7429512998901502,
+      "grad_norm": 0.27062484433619666,
+      "learning_rate": 9.760770424011557e-07,
+      "loss": 0.0742,
+      "step": 1190
+    },
+    {
+      "epoch": 1.744415964848041,
+      "grad_norm": 0.2942597554218074,
+      "learning_rate": 9.650735368978793e-07,
+      "loss": 0.0893,
+      "step": 1191
+    },
+    {
+      "epoch": 1.745880629805932,
+      "grad_norm": 0.2595099599479516,
+      "learning_rate": 9.541292604770502e-07,
+      "loss": 0.0801,
+      "step": 1192
+    },
+    {
+      "epoch": 1.7473452947638228,
+      "grad_norm": 0.25996832033844564,
+      "learning_rate": 9.432442848846291e-07,
+      "loss": 0.0802,
+      "step": 1193
+    },
+    {
+      "epoch": 1.7488099597217137,
+      "grad_norm": 0.25099023362939665,
+      "learning_rate": 9.324186814778202e-07,
+      "loss": 0.0759,
+      "step": 1194
+    },
+    {
+      "epoch": 1.7502746246796046,
+      "grad_norm": 0.261784151531945,
+      "learning_rate": 9.216525212246131e-07,
+      "loss": 0.068,
+      "step": 1195
+    },
+    {
+      "epoch": 1.7517392896374955,
+      "grad_norm": 0.2676041434167344,
+      "learning_rate": 9.109458747033106e-07,
+      "loss": 0.0758,
+      "step": 1196
+    },
+    {
+      "epoch": 1.7532039545953864,
+      "grad_norm": 0.27324413990185265,
+      "learning_rate": 9.00298812102075e-07,
+      "loss": 0.0781,
+      "step": 1197
+    },
+    {
+      "epoch": 1.7546686195532772,
+      "grad_norm": 0.23917286810904692,
+      "learning_rate": 8.897114032184539e-07,
+      "loss": 0.0604,
+      "step": 1198
+    },
+    {
+      "epoch": 1.7561332845111681,
+      "grad_norm": 0.2632087773397572,
+      "learning_rate": 8.791837174589401e-07,
+      "loss": 0.0835,
+      "step": 1199
+    },
+    {
+      "epoch": 1.757597949469059,
+      "grad_norm": 0.2745654342512245,
+      "learning_rate": 8.687158238384963e-07,
+      "loss": 0.0688,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7590626144269499,
+      "grad_norm": 0.255009900836695,
+      "learning_rate": 8.583077909801252e-07,
+      "loss": 0.0729,
+      "step": 1201
+    },
+    {
+      "epoch": 1.7605272793848408,
+      "grad_norm": 0.24279405877132665,
+      "learning_rate": 8.47959687114398e-07,
+      "loss": 0.064,
+      "step": 1202
+    },
+    {
+      "epoch": 1.7619919443427317,
+      "grad_norm": 0.2671484719871084,
+      "learning_rate": 8.37671580079027e-07,
+      "loss": 0.0689,
+      "step": 1203
+    },
+    {
+      "epoch": 1.7634566093006225,
+      "grad_norm": 0.22793656038154106,
+      "learning_rate": 8.274435373184009e-07,
+      "loss": 0.0609,
+      "step": 1204
+    },
+    {
+      "epoch": 1.7649212742585134,
+      "grad_norm": 0.25888704046474315,
+      "learning_rate": 8.172756258831638e-07,
+      "loss": 0.0663,
+      "step": 1205
+    },
+    {
+      "epoch": 1.7663859392164043,
+      "grad_norm": 0.2694033818200205,
+      "learning_rate": 8.071679124297537e-07,
+      "loss": 0.077,
+      "step": 1206
+    },
+    {
+      "epoch": 1.7678506041742952,
+      "grad_norm": 0.25795707216950153,
+      "learning_rate": 7.971204632199869e-07,
+      "loss": 0.0738,
+      "step": 1207
+    },
+    {
+      "epoch": 1.769315269132186,
+      "grad_norm": 0.25028412383513904,
+      "learning_rate": 7.871333441206053e-07,
+      "loss": 0.0677,
+      "step": 1208
+    },
+    {
+      "epoch": 1.770779934090077,
+      "grad_norm": 0.23004892259257617,
+      "learning_rate": 7.772066206028572e-07,
+      "loss": 0.0612,
+      "step": 1209
+    },
+    {
+      "epoch": 1.7722445990479678,
+      "grad_norm": 0.25383482467849333,
+      "learning_rate": 7.673403577420591e-07,
+      "loss": 0.0685,
+      "step": 1210
+    },
+    {
+      "epoch": 1.7737092640058587,
+      "grad_norm": 0.2519042969852701,
+      "learning_rate": 7.575346202171819e-07,
+      "loss": 0.0647,
+      "step": 1211
+    },
+    {
+      "epoch": 1.7751739289637496,
+      "grad_norm": 0.26600143625920347,
+      "learning_rate": 7.477894723104073e-07,
+      "loss": 0.0807,
+      "step": 1212
+    },
+    {
+      "epoch": 1.7766385939216405,
+      "grad_norm": 0.24525033390296022,
+      "learning_rate": 7.381049779067273e-07,
+      "loss": 0.0616,
+      "step": 1213
+    },
+    {
+      "epoch": 1.7781032588795314,
+      "grad_norm": 0.24140459393603353,
+      "learning_rate": 7.284812004935083e-07,
+      "loss": 0.0648,
+      "step": 1214
+    },
+    {
+      "epoch": 1.7795679238374222,
+      "grad_norm": 0.27804404244360953,
+      "learning_rate": 7.189182031600906e-07,
+      "loss": 0.0886,
+      "step": 1215
+    },
+    {
+      "epoch": 1.7810325887953131,
+      "grad_norm": 0.2619152806722488,
+      "learning_rate": 7.094160485973567e-07,
+      "loss": 0.0826,
+      "step": 1216
+    },
+    {
+      "epoch": 1.782497253753204,
+      "grad_norm": 0.2204653484419197,
+      "learning_rate": 6.999747990973382e-07,
+      "loss": 0.0555,
+      "step": 1217
+    },
+    {
+      "epoch": 1.7839619187110949,
+      "grad_norm": 0.26650987169853924,
+      "learning_rate": 6.905945165527928e-07,
+      "loss": 0.0699,
+      "step": 1218
+    },
+    {
+      "epoch": 1.7854265836689858,
+      "grad_norm": 0.2451107677206505,
+      "learning_rate": 6.812752624568131e-07,
+      "loss": 0.0739,
+      "step": 1219
+    },
+    {
+      "epoch": 1.7868912486268766,
+      "grad_norm": 0.24387464427507713,
+      "learning_rate": 6.720170979024065e-07,
+      "loss": 0.066,
+      "step": 1220
+    },
+    {
+      "epoch": 1.7883559135847675,
+      "grad_norm": 0.24790390060583067,
+      "learning_rate": 6.628200835821119e-07,
+      "loss": 0.0568,
+      "step": 1221
+    },
+    {
+      "epoch": 1.7898205785426584,
+      "grad_norm": 0.26645550894398845,
+      "learning_rate": 6.536842797875876e-07,
+      "loss": 0.0864,
+      "step": 1222
+    },
+    {
+      "epoch": 1.7912852435005493,
+      "grad_norm": 0.2719628682812989,
+      "learning_rate": 6.446097464092249e-07,
+      "loss": 0.0822,
+      "step": 1223
+    },
+    {
+      "epoch": 1.7927499084584402,
+      "grad_norm": 0.25379643552394104,
+      "learning_rate": 6.355965429357513e-07,
+      "loss": 0.0737,
+      "step": 1224
+    },
+    {
+      "epoch": 1.794214573416331,
+      "grad_norm": 0.22752210074212567,
+      "learning_rate": 6.266447284538446e-07,
+      "loss": 0.0541,
+      "step": 1225
+    },
+    {
+      "epoch": 1.795679238374222,
+      "grad_norm": 0.2665704509120899,
+      "learning_rate": 6.177543616477377e-07,
+      "loss": 0.0759,
+      "step": 1226
+    },
+    {
+      "epoch": 1.7971439033321128,
+      "grad_norm": 0.2500751657498049,
+      "learning_rate": 6.08925500798847e-07,
+      "loss": 0.0631,
+      "step": 1227
+    },
+    {
+      "epoch": 1.7986085682900037,
+      "grad_norm": 0.259144080680451,
+      "learning_rate": 6.001582037853726e-07,
+      "loss": 0.0777,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8000732332478946,
+      "grad_norm": 0.2897130385154202,
+      "learning_rate": 5.914525280819383e-07,
+      "loss": 0.0802,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8015378982057855,
+      "grad_norm": 0.2553061949427675,
+      "learning_rate": 5.828085307591969e-07,
+      "loss": 0.0692,
+      "step": 1230
+    },
+    {
+      "epoch": 1.8030025631636764,
+      "grad_norm": 0.2572539562396507,
+      "learning_rate": 5.742262684834698e-07,
+      "loss": 0.0833,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8044672281215672,
+      "grad_norm": 0.24772988825308792,
+      "learning_rate": 5.657057975163682e-07,
+      "loss": 0.0673,
+      "step": 1232
+    },
+    {
+      "epoch": 1.8059318930794581,
+      "grad_norm": 0.2717711132164481,
+      "learning_rate": 5.572471737144247e-07,
+      "loss": 0.0778,
+      "step": 1233
+    },
+    {
+      "epoch": 1.807396558037349,
+      "grad_norm": 0.2539596096042241,
+      "learning_rate": 5.488504525287319e-07,
+      "loss": 0.0727,
+      "step": 1234
+    },
+    {
+      "epoch": 1.8088612229952399,
+      "grad_norm": 0.2345694037301944,
+      "learning_rate": 5.405156890045704e-07,
+      "loss": 0.0616,
+      "step": 1235
+    },
+    {
+      "epoch": 1.8103258879531308,
+      "grad_norm": 0.2609506722100195,
+      "learning_rate": 5.322429377810612e-07,
+      "loss": 0.0853,
+      "step": 1236
+    },
+    {
+      "epoch": 1.8117905529110216,
+      "grad_norm": 0.25093904412233514,
+      "learning_rate": 5.240322530907893e-07,
+      "loss": 0.0823,
+      "step": 1237
+    },
+    {
+      "epoch": 1.8132552178689125,
+      "grad_norm": 0.2544084467508812,
+      "learning_rate": 5.158836887594687e-07,
+      "loss": 0.063,
+      "step": 1238
+    },
+    {
+      "epoch": 1.8147198828268034,
+      "grad_norm": 0.26650992359632614,
+      "learning_rate": 5.07797298205569e-07,
+      "loss": 0.0789,
+      "step": 1239
+    },
+    {
+      "epoch": 1.8161845477846943,
+      "grad_norm": 0.23630880168422222,
+      "learning_rate": 4.997731344399837e-07,
+      "loss": 0.0606,
+      "step": 1240
+    },
+    {
+      "epoch": 1.8176492127425852,
+      "grad_norm": 0.24553703591038675,
+      "learning_rate": 4.91811250065668e-07,
+      "loss": 0.0628,
+      "step": 1241
+    },
+    {
+      "epoch": 1.819113877700476,
+      "grad_norm": 0.2265227129455893,
+      "learning_rate": 4.839116972773061e-07,
+      "loss": 0.0583,
+      "step": 1242
+    },
+    {
+      "epoch": 1.820578542658367,
+      "grad_norm": 0.25312671344061183,
+      "learning_rate": 4.7607452786095686e-07,
+      "loss": 0.0608,
+      "step": 1243
+    },
+    {
+      "epoch": 1.8220432076162578,
+      "grad_norm": 0.23959774438699266,
+      "learning_rate": 4.682997931937283e-07,
+      "loss": 0.0628,
+      "step": 1244
+    },
+    {
+      "epoch": 1.8235078725741487,
+      "grad_norm": 0.31949707497166124,
+      "learning_rate": 4.605875442434238e-07,
+      "loss": 0.0964,
+      "step": 1245
+    },
+    {
+      "epoch": 1.8249725375320396,
+      "grad_norm": 0.2500895954628304,
+      "learning_rate": 4.5293783156822533e-07,
+      "loss": 0.0679,
+      "step": 1246
+    },
+    {
+      "epoch": 1.8264372024899305,
+      "grad_norm": 0.24693595115617872,
+      "learning_rate": 4.4535070531635195e-07,
+      "loss": 0.0722,
+      "step": 1247
+    },
+    {
+      "epoch": 1.8279018674478213,
+      "grad_norm": 0.25545938072704727,
+      "learning_rate": 4.378262152257273e-07,
+      "loss": 0.0683,
+      "step": 1248
+    },
+    {
+      "epoch": 1.8293665324057122,
+      "grad_norm": 0.25876832202455974,
+      "learning_rate": 4.303644106236704e-07,
+      "loss": 0.0662,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8308311973636031,
+      "grad_norm": 0.26720989490367125,
+      "learning_rate": 4.2296534042654993e-07,
+      "loss": 0.0682,
+      "step": 1250
+    },
+    {
+      "epoch": 1.832295862321494,
+      "grad_norm": 0.26258372219745035,
+      "learning_rate": 4.1562905313948354e-07,
+      "loss": 0.0666,
+      "step": 1251
+    },
+    {
+      "epoch": 1.8337605272793849,
+      "grad_norm": 0.26711183739660005,
+      "learning_rate": 4.083555968560049e-07,
+      "loss": 0.0801,
+      "step": 1252
+    },
+    {
+      "epoch": 1.8352251922372758,
+      "grad_norm": 0.25571699526490943,
+      "learning_rate": 4.0114501925775927e-07,
+      "loss": 0.0765,
+      "step": 1253
+    },
+    {
+      "epoch": 1.8366898571951666,
+      "grad_norm": 0.25954869259656904,
+      "learning_rate": 3.9399736761418395e-07,
+      "loss": 0.0876,
+      "step": 1254
+    },
+    {
+      "epoch": 1.8381545221530575,
+      "grad_norm": 0.25678722587829267,
+      "learning_rate": 3.8691268878220165e-07,
+      "loss": 0.069,
+      "step": 1255
+    },
+    {
+      "epoch": 1.8396191871109484,
+      "grad_norm": 0.2589134232675844,
+      "learning_rate": 3.7989102920591103e-07,
+      "loss": 0.069,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8410838520688393,
+      "grad_norm": 0.2450077649829702,
+      "learning_rate": 3.729324349162866e-07,
+      "loss": 0.0684,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8425485170267302,
+      "grad_norm": 0.2853806694841532,
+      "learning_rate": 3.660369515308715e-07,
+      "loss": 0.0634,
+      "step": 1258
+    },
+    {
+      "epoch": 1.844013181984621,
+      "grad_norm": 0.2621846237984564,
+      "learning_rate": 3.592046242534819e-07,
+      "loss": 0.0714,
+      "step": 1259
+    },
+    {
+      "epoch": 1.845477846942512,
+      "grad_norm": 0.23779005991008018,
+      "learning_rate": 3.524354978739075e-07,
+      "loss": 0.0583,
+      "step": 1260
+    },
+    {
+      "epoch": 1.8469425119004028,
+      "grad_norm": 0.25039681136747866,
+      "learning_rate": 3.4572961676762715e-07,
+      "loss": 0.0738,
+      "step": 1261
+    },
+    {
+      "epoch": 1.8484071768582937,
+      "grad_norm": 0.26098923170285054,
+      "learning_rate": 3.390870248955025e-07,
+      "loss": 0.0807,
+      "step": 1262
+    },
+    {
+      "epoch": 1.8498718418161846,
+      "grad_norm": 0.26327776179903706,
+      "learning_rate": 3.3250776580350143e-07,
+      "loss": 0.0866,
+      "step": 1263
+    },
+    {
+      "epoch": 1.8513365067740755,
+      "grad_norm": 0.23667264720127792,
+      "learning_rate": 3.259918826224118e-07,
+      "loss": 0.0649,
+      "step": 1264
+    },
+    {
+      "epoch": 1.8528011717319663,
+      "grad_norm": 0.23761172772297415,
+      "learning_rate": 3.1953941806755265e-07,
+      "loss": 0.0845,
+      "step": 1265
+    },
+    {
+      "epoch": 1.8542658366898572,
+      "grad_norm": 0.24024069607921544,
+      "learning_rate": 3.131504144385023e-07,
+      "loss": 0.0703,
+      "step": 1266
+    },
+    {
+      "epoch": 1.855730501647748,
+      "grad_norm": 0.2420868970268621,
+      "learning_rate": 3.0682491361881064e-07,
+      "loss": 0.0748,
+      "step": 1267
+    },
+    {
+      "epoch": 1.857195166605639,
+      "grad_norm": 0.2577949839616261,
+      "learning_rate": 3.0056295707573736e-07,
+      "loss": 0.0665,
+      "step": 1268
+    },
+    {
+      "epoch": 1.8586598315635299,
+      "grad_norm": 0.2685695682554958,
+      "learning_rate": 2.943645858599653e-07,
+      "loss": 0.0792,
+      "step": 1269
+    },
+    {
+      "epoch": 1.8601244965214208,
+      "grad_norm": 0.25460210036235115,
+      "learning_rate": 2.8822984060534854e-07,
+      "loss": 0.0738,
+      "step": 1270
+    },
+    {
+      "epoch": 1.8615891614793116,
+      "grad_norm": 0.23461840979365672,
+      "learning_rate": 2.82158761528627e-07,
+      "loss": 0.0754,
+      "step": 1271
+    },
+    {
+      "epoch": 1.8630538264372025,
+      "grad_norm": 0.22557458232043867,
+      "learning_rate": 2.761513884291822e-07,
+      "loss": 0.0612,
+      "step": 1272
+    },
+    {
+      "epoch": 1.8645184913950934,
+      "grad_norm": 0.2470496535205524,
+      "learning_rate": 2.7020776068875876e-07,
+      "loss": 0.0711,
+      "step": 1273
+    },
+    {
+      "epoch": 1.8659831563529843,
+      "grad_norm": 0.2329397076753076,
+      "learning_rate": 2.6432791727121984e-07,
+      "loss": 0.0704,
+      "step": 1274
+    },
+    {
+      "epoch": 1.8674478213108752,
+      "grad_norm": 0.25477498790129094,
+      "learning_rate": 2.5851189672228103e-07,
+      "loss": 0.0773,
+      "step": 1275
+    },
+    {
+      "epoch": 1.868912486268766,
+      "grad_norm": 0.23446595512868088,
+      "learning_rate": 2.5275973716926804e-07,
+      "loss": 0.0709,
+      "step": 1276
+    },
+    {
+      "epoch": 1.870377151226657,
+      "grad_norm": 0.2605236921574388,
+      "learning_rate": 2.4707147632085815e-07,
+      "loss": 0.0758,
+      "step": 1277
+    },
+    {
+      "epoch": 1.8718418161845478,
+      "grad_norm": 0.24531947955587904,
+      "learning_rate": 2.414471514668348e-07,
+      "loss": 0.0778,
+      "step": 1278
+    },
+    {
+      "epoch": 1.8733064811424387,
+      "grad_norm": 0.2625806435047024,
+      "learning_rate": 2.358867994778502e-07,
+      "loss": 0.0742,
+      "step": 1279
+    },
+    {
+      "epoch": 1.8747711461003296,
+      "grad_norm": 0.26172740430959623,
+      "learning_rate": 2.3039045680517292e-07,
+      "loss": 0.071,
+      "step": 1280
+    },
+    {
+      "epoch": 1.8762358110582205,
+      "grad_norm": 0.25812348656180795,
+      "learning_rate": 2.249581594804562e-07,
+      "loss": 0.0702,
+      "step": 1281
+    },
+    {
+      "epoch": 1.8777004760161113,
+      "grad_norm": 0.22412143150895958,
+      "learning_rate": 2.1958994311549797e-07,
+      "loss": 0.0633,
+      "step": 1282
+    },
+    {
+      "epoch": 1.8791651409740022,
+      "grad_norm": 0.25437645779594675,
+      "learning_rate": 2.1428584290201116e-07,
+      "loss": 0.0697,
+      "step": 1283
+    },
+    {
+      "epoch": 1.880629805931893,
+      "grad_norm": 0.2728173599791171,
+      "learning_rate": 2.0904589361138927e-07,
+      "loss": 0.0813,
+      "step": 1284
+    },
+    {
+      "epoch": 1.882094470889784,
+      "grad_norm": 0.24620604884353783,
+      "learning_rate": 2.0387012959448227e-07,
+      "loss": 0.0812,
+      "step": 1285
+    },
+    {
+      "epoch": 1.8835591358476749,
+      "grad_norm": 0.2246638728164293,
+      "learning_rate": 1.9875858478136557e-07,
+      "loss": 0.0591,
+      "step": 1286
+    },
+    {
+      "epoch": 1.8850238008055658,
+      "grad_norm": 0.22425127272955642,
+      "learning_rate": 1.9371129268112466e-07,
+      "loss": 0.0615,
+      "step": 1287
+    },
+    {
+      "epoch": 1.8864884657634566,
+      "grad_norm": 0.47369574027243294,
+      "learning_rate": 1.8872828638162866e-07,
+      "loss": 0.0678,
+      "step": 1288
+    },
+    {
+      "epoch": 1.8879531307213475,
+      "grad_norm": 0.24744976308618485,
+      "learning_rate": 1.8380959854932045e-07,
+      "loss": 0.0745,
+      "step": 1289
+    },
+    {
+      "epoch": 1.8894177956792384,
+      "grad_norm": 0.2619097810479823,
+      "learning_rate": 1.7895526142899466e-07,
+      "loss": 0.0705,
+      "step": 1290
+    },
+    {
+      "epoch": 1.8908824606371293,
+      "grad_norm": 0.23871395237352025,
+      "learning_rate": 1.7416530684359444e-07,
+      "loss": 0.0551,
+      "step": 1291
+    },
+    {
+      "epoch": 1.8923471255950202,
+      "grad_norm": 0.24628533422104648,
+      "learning_rate": 1.6943976619399615e-07,
+      "loss": 0.0677,
+      "step": 1292
+    },
+    {
+      "epoch": 1.893811790552911,
+      "grad_norm": 0.23856368422966648,
+      "learning_rate": 1.6477867045880613e-07,
+      "loss": 0.0815,
+      "step": 1293
+    },
+    {
+      "epoch": 1.895276455510802,
+      "grad_norm": 0.2555025186995139,
+      "learning_rate": 1.6018205019415866e-07,
+      "loss": 0.0708,
+      "step": 1294
+    },
+    {
+      "epoch": 1.8967411204686928,
+      "grad_norm": 0.2528265155563436,
+      "learning_rate": 1.5564993553351394e-07,
+      "loss": 0.0738,
+      "step": 1295
+    },
+    {
+      "epoch": 1.8982057854265837,
+      "grad_norm": 0.26162054123344747,
+      "learning_rate": 1.511823561874637e-07,
+      "loss": 0.07,
+      "step": 1296
+    },
+    {
+      "epoch": 1.8996704503844746,
+      "grad_norm": 0.24902092413650095,
+      "learning_rate": 1.4677934144352923e-07,
+      "loss": 0.0668,
+      "step": 1297
+    },
+    {
+      "epoch": 1.9011351153423655,
+      "grad_norm": 0.2635425216052296,
+      "learning_rate": 1.4244092016597933e-07,
+      "loss": 0.0797,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9025997803002563,
+      "grad_norm": 0.24273285605827147,
+      "learning_rate": 1.3816712079563034e-07,
+      "loss": 0.0738,
+      "step": 1299
+    },
+    {
+      "epoch": 1.9040644452581472,
+      "grad_norm": 0.23589111930612008,
+      "learning_rate": 1.3395797134967192e-07,
+      "loss": 0.0633,
+      "step": 1300
+    },
+    {
+      "epoch": 1.905529110216038,
+      "grad_norm": 0.2886312181672797,
+      "learning_rate": 1.2981349942146947e-07,
+      "loss": 0.0801,
+      "step": 1301
+    },
+    {
+      "epoch": 1.906993775173929,
+      "grad_norm": 0.2504691073271145,
+      "learning_rate": 1.257337321803964e-07,
+      "loss": 0.0719,
+      "step": 1302
+    },
+    {
+      "epoch": 1.9084584401318199,
+      "grad_norm": 0.2455439126340956,
+      "learning_rate": 1.2171869637164769e-07,
+      "loss": 0.0694,
+      "step": 1303
+    },
+    {
+      "epoch": 1.9099231050897107,
+      "grad_norm": 0.2443358735504617,
+      "learning_rate": 1.1776841831606544e-07,
+      "loss": 0.0629,
+      "step": 1304
+    },
+    {
+      "epoch": 1.9113877700476016,
+      "grad_norm": 0.2375928891798295,
+      "learning_rate": 1.1388292390997035e-07,
+      "loss": 0.0636,
+      "step": 1305
+    },
+    {
+      "epoch": 1.9128524350054925,
+      "grad_norm": 0.2430807124576779,
+      "learning_rate": 1.1006223862498944e-07,
+      "loss": 0.0654,
+      "step": 1306
+    },
+    {
+      "epoch": 1.9143170999633834,
+      "grad_norm": 0.2582360186234949,
+      "learning_rate": 1.0630638750788625e-07,
+      "loss": 0.0772,
+      "step": 1307
+    },
+    {
+      "epoch": 1.9157817649212743,
+      "grad_norm": 0.22936545545360856,
+      "learning_rate": 1.026153951804032e-07,
+      "loss": 0.0555,
+      "step": 1308
+    },
+    {
+      "epoch": 1.9172464298791652,
+      "grad_norm": 0.24362803135706435,
+      "learning_rate": 9.898928583909284e-08,
+      "loss": 0.0643,
+      "step": 1309
+    },
+    {
+      "epoch": 1.918711094837056,
+      "grad_norm": 0.23451079860288676,
+      "learning_rate": 9.542808325516573e-08,
+      "loss": 0.0562,
+      "step": 1310
+    },
+    {
+      "epoch": 1.920175759794947,
+      "grad_norm": 0.23945075253871337,
+      "learning_rate": 9.193181077433055e-08,
+      "loss": 0.07,
+      "step": 1311
+    },
+    {
+      "epoch": 1.9216404247528378,
+      "grad_norm": 0.2586122862421169,
+      "learning_rate": 8.850049131664206e-08,
+      "loss": 0.0793,
+      "step": 1312
+    },
+    {
+      "epoch": 1.9231050897107287,
+      "grad_norm": 0.25189426121665187,
+      "learning_rate": 8.513414737635006e-08,
+      "loss": 0.0665,
+      "step": 1313
+    },
+    {
+      "epoch": 1.9245697546686196,
+      "grad_norm": 0.2331012828544825,
+      "learning_rate": 8.183280102175617e-08,
+      "loss": 0.0626,
+      "step": 1314
+    },
+    {
+      "epoch": 1.9260344196265105,
+      "grad_norm": 0.24925536553110555,
+      "learning_rate": 7.859647389506176e-08,
+      "loss": 0.0688,
+      "step": 1315
+    },
+    {
+      "epoch": 1.9274990845844013,
+      "grad_norm": 0.25467958607825186,
+      "learning_rate": 7.542518721223469e-08,
+      "loss": 0.0894,
+      "step": 1316
+    },
+    {
+      "epoch": 1.9289637495422922,
+      "grad_norm": 0.25285918104772465,
+      "learning_rate": 7.231896176285946e-08,
+      "loss": 0.0773,
+      "step": 1317
+    },
+    {
+      "epoch": 1.930428414500183,
+      "grad_norm": 0.23450711664911186,
+      "learning_rate": 6.927781791001398e-08,
+      "loss": 0.0639,
+      "step": 1318
+    },
+    {
+      "epoch": 1.931893079458074,
+      "grad_norm": 0.23097872990832516,
+      "learning_rate": 6.630177559012518e-08,
+      "loss": 0.068,
+      "step": 1319
+    },
+    {
+      "epoch": 1.9333577444159649,
+      "grad_norm": 0.2379637592228772,
+      "learning_rate": 6.339085431284253e-08,
+      "loss": 0.0632,
+      "step": 1320
+    },
+    {
+      "epoch": 1.9348224093738557,
+      "grad_norm": 0.24051632358031436,
+      "learning_rate": 6.054507316091141e-08,
+      "loss": 0.0677,
+      "step": 1321
+    },
+    {
+      "epoch": 1.9362870743317466,
+      "grad_norm": 0.28404424390350286,
+      "learning_rate": 5.7764450790046554e-08,
+      "loss": 0.0783,
+      "step": 1322
+    },
+    {
+      "epoch": 1.9377517392896375,
+      "grad_norm": 0.24006965033366337,
+      "learning_rate": 5.5049005428808865e-08,
+      "loss": 0.0696,
+      "step": 1323
+    },
+    {
+      "epoch": 1.9392164042475284,
+      "grad_norm": 0.2496157646264579,
+      "learning_rate": 5.239875487848878e-08,
+      "loss": 0.0635,
+      "step": 1324
+    },
+    {
+      "epoch": 1.9406810692054193,
+      "grad_norm": 0.26403747456776216,
+      "learning_rate": 4.981371651298306e-08,
+      "loss": 0.0652,
+      "step": 1325
+    },
+    {
+      "epoch": 1.9421457341633102,
+      "grad_norm": 0.25920069106169014,
+      "learning_rate": 4.729390727869154e-08,
+      "loss": 0.0799,
+      "step": 1326
+    },
+    {
+      "epoch": 1.943610399121201,
+      "grad_norm": 0.26170514428715264,
+      "learning_rate": 4.483934369439613e-08,
+      "loss": 0.0755,
+      "step": 1327
+    },
+    {
+      "epoch": 1.945075064079092,
+      "grad_norm": 0.24542483964564413,
+      "learning_rate": 4.245004185115753e-08,
+      "loss": 0.0615,
+      "step": 1328
+    },
+    {
+      "epoch": 1.9465397290369828,
+      "grad_norm": 0.24837578672571922,
+      "learning_rate": 4.0126017412207565e-08,
+      "loss": 0.071,
+      "step": 1329
+    },
+    {
+      "epoch": 1.9480043939948737,
+      "grad_norm": 0.23824898115769536,
+      "learning_rate": 3.786728561285036e-08,
+      "loss": 0.0594,
+      "step": 1330
+    },
+    {
+      "epoch": 1.9494690589527646,
+      "grad_norm": 0.25242932208373725,
+      "learning_rate": 3.567386126035577e-08,
+      "loss": 0.0807,
+      "step": 1331
+    },
+    {
+      "epoch": 1.9509337239106554,
+      "grad_norm": 0.26364658800502216,
+      "learning_rate": 3.354575873386945e-08,
+      "loss": 0.0701,
+      "step": 1332
+    },
+    {
+      "epoch": 1.9523983888685463,
+      "grad_norm": 0.22064989501595925,
+      "learning_rate": 3.1482991984312926e-08,
+      "loss": 0.0601,
+      "step": 1333
+    },
+    {
+      "epoch": 1.9538630538264372,
+      "grad_norm": 0.25926143810431157,
+      "learning_rate": 2.948557453429701e-08,
+      "loss": 0.0671,
+      "step": 1334
+    },
+    {
+      "epoch": 1.955327718784328,
+      "grad_norm": 0.23624107516760934,
+      "learning_rate": 2.7553519478028535e-08,
+      "loss": 0.0665,
+      "step": 1335
+    },
+    {
+      "epoch": 1.956792383742219,
+      "grad_norm": 0.2705488604754587,
+      "learning_rate": 2.5686839481227077e-08,
+      "loss": 0.0773,
+      "step": 1336
+    },
+    {
+      "epoch": 1.9582570487001099,
+      "grad_norm": 0.2473961524857793,
+      "learning_rate": 2.3885546781042824e-08,
+      "loss": 0.073,
+      "step": 1337
+    },
+    {
+      "epoch": 1.9597217136580007,
+      "grad_norm": 0.2441090901112598,
+      "learning_rate": 2.2149653185973285e-08,
+      "loss": 0.0769,
+      "step": 1338
+    },
+    {
+      "epoch": 1.9611863786158916,
+      "grad_norm": 0.27046514605231436,
+      "learning_rate": 2.0479170075788924e-08,
+      "loss": 0.0716,
+      "step": 1339
+    },
+    {
+      "epoch": 1.9626510435737825,
+      "grad_norm": 0.22983632777466814,
+      "learning_rate": 1.8874108401456538e-08,
+      "loss": 0.0575,
+      "step": 1340
+    },
+    {
+      "epoch": 1.9641157085316734,
+      "grad_norm": 0.24070147293086863,
+      "learning_rate": 1.7334478685068212e-08,
+      "loss": 0.1044,
+      "step": 1341
+    },
+    {
+      "epoch": 1.9655803734895643,
+      "grad_norm": 0.24856888752891837,
+      "learning_rate": 1.586029101977249e-08,
+      "loss": 0.0585,
+      "step": 1342
+    },
+    {
+      "epoch": 1.9670450384474552,
+      "grad_norm": 0.25551924386712754,
+      "learning_rate": 1.4451555069708856e-08,
+      "loss": 0.0823,
+      "step": 1343
+    },
+    {
+      "epoch": 1.968509703405346,
+      "grad_norm": 0.24403209467826953,
+      "learning_rate": 1.3108280069941137e-08,
+      "loss": 0.0723,
+      "step": 1344
+    },
+    {
+      "epoch": 1.969974368363237,
+      "grad_norm": 0.25587914011988994,
+      "learning_rate": 1.1830474826404204e-08,
+      "loss": 0.0775,
+      "step": 1345
+    },
+    {
+      "epoch": 1.9714390333211278,
+      "grad_norm": 0.2481616027718323,
+      "learning_rate": 1.0618147715835137e-08,
+      "loss": 0.073,
+      "step": 1346
+    },
+    {
+      "epoch": 1.9729036982790187,
+      "grad_norm": 0.2503297758869869,
+      "learning_rate": 9.471306685728821e-09,
+      "loss": 0.0841,
+      "step": 1347
+    },
+    {
+      "epoch": 1.9743683632369096,
+      "grad_norm": 0.25325687071902725,
+      "learning_rate": 8.389959254281322e-09,
+      "loss": 0.0674,
+      "step": 1348
+    },
+    {
+      "epoch": 1.9758330281948004,
+      "grad_norm": 0.2413650911828191,
+      "learning_rate": 7.3741125103399254e-09,
+      "loss": 0.0737,
+      "step": 1349
+    },
+    {
+      "epoch": 1.9772976931526913,
+      "grad_norm": 0.23526516349129203,
+      "learning_rate": 6.423773113357623e-09,
+      "loss": 0.0631,
+      "step": 1350
+    },
+    {
+      "epoch": 1.9787623581105822,
+      "grad_norm": 0.25457045216657825,
+      "learning_rate": 5.538947293349806e-09,
+      "loss": 0.0765,
+      "step": 1351
+    },
+    {
+      "epoch": 1.980227023068473,
+      "grad_norm": 0.23325185099514448,
+      "learning_rate": 4.719640850852081e-09,
+      "loss": 0.0612,
+      "step": 1352
+    },
+    {
+      "epoch": 1.981691688026364,
+      "grad_norm": 0.26527379982321414,
+      "learning_rate": 3.965859156885854e-09,
+      "loss": 0.0775,
+      "step": 1353
+    },
+    {
+      "epoch": 1.9831563529842549,
+      "grad_norm": 0.24201253323675176,
+      "learning_rate": 3.2776071529183608e-09,
+      "loss": 0.0685,
+      "step": 1354
+    },
+    {
+      "epoch": 1.9846210179421457,
+      "grad_norm": 0.23684069181793793,
+      "learning_rate": 2.65488935083158e-09,
+      "loss": 0.0686,
+      "step": 1355
+    },
+    {
+      "epoch": 1.9860856829000366,
+      "grad_norm": 0.25231448957519526,
+      "learning_rate": 2.0977098328978098e-09,
+      "loss": 0.0726,
+      "step": 1356
+    },
+    {
+      "epoch": 1.9875503478579275,
+      "grad_norm": 0.2774834947544314,
+      "learning_rate": 1.6060722517430293e-09,
+      "loss": 0.0787,
+      "step": 1357
+    },
+    {
+      "epoch": 1.9890150128158184,
+      "grad_norm": 0.2503245321475594,
+      "learning_rate": 1.1799798303335775e-09,
+      "loss": 0.0683,
+      "step": 1358
+    },
+    {
+      "epoch": 1.9904796777737093,
+      "grad_norm": 0.2423085553528186,
+      "learning_rate": 8.194353619450646e-10,
+      "loss": 0.0731,
+      "step": 1359
+    },
+    {
+      "epoch": 1.9919443427316001,
+      "grad_norm": 0.24083594462983027,
+      "learning_rate": 5.244412101534924e-10,
+      "loss": 0.066,
+      "step": 1360
+    },
+    {
+      "epoch": 1.993409007689491,
+      "grad_norm": 0.23578471416137511,
+      "learning_rate": 2.949993088130487e-10,
+      "loss": 0.0585,
+      "step": 1361
+    },
+    {
+      "epoch": 1.994873672647382,
+      "grad_norm": 0.24433039086610767,
+      "learning_rate": 1.3111116204500562e-10,
+      "loss": 0.0602,
+      "step": 1362
+    },
+    {
+      "epoch": 1.9963383376052728,
+      "grad_norm": 0.2719863457397181,
+      "learning_rate": 3.277784423105779e-11,
+      "loss": 0.0863,
+      "step": 1363
+    },
+    {
+      "epoch": 1.9978030025631637,
+      "grad_norm": 0.261652070533884,
+      "learning_rate": 0.0,
+      "loss": 0.0707,
+      "step": 1364
+    },
+    {
+      "epoch": 1.9978030025631637,
+      "step": 1364,
+      "total_flos": 612552856354816.0,
+      "train_loss": 0.16333763612996885,
+      "train_runtime": 7219.9243,
+      "train_samples_per_second": 24.207,
+      "train_steps_per_second": 0.189
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1364,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "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": 612552856354816.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}