diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,41988 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 5990,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000333889816360601,
+      "grad_norm": 16.150521804678494,
+      "learning_rate": 1.669449081803005e-08,
+      "loss": 1.3354,
+      "step": 1
+    },
+    {
+      "epoch": 0.000667779632721202,
+      "grad_norm": 16.480776582942426,
+      "learning_rate": 3.33889816360601e-08,
+      "loss": 1.3657,
+      "step": 2
+    },
+    {
+      "epoch": 0.001001669449081803,
+      "grad_norm": 15.716580738384542,
+      "learning_rate": 5.008347245409015e-08,
+      "loss": 1.3094,
+      "step": 3
+    },
+    {
+      "epoch": 0.001335559265442404,
+      "grad_norm": 16.270254812256358,
+      "learning_rate": 6.67779632721202e-08,
+      "loss": 1.3514,
+      "step": 4
+    },
+    {
+      "epoch": 0.001669449081803005,
+      "grad_norm": 16.234075458363257,
+      "learning_rate": 8.347245409015026e-08,
+      "loss": 1.3446,
+      "step": 5
+    },
+    {
+      "epoch": 0.002003338898163606,
+      "grad_norm": 16.26563246528343,
+      "learning_rate": 1.001669449081803e-07,
+      "loss": 1.3578,
+      "step": 6
+    },
+    {
+      "epoch": 0.002337228714524207,
+      "grad_norm": 16.220160683811375,
+      "learning_rate": 1.1686143572621036e-07,
+      "loss": 1.3477,
+      "step": 7
+    },
+    {
+      "epoch": 0.002671118530884808,
+      "grad_norm": 15.655363417957004,
+      "learning_rate": 1.335559265442404e-07,
+      "loss": 1.3367,
+      "step": 8
+    },
+    {
+      "epoch": 0.003005008347245409,
+      "grad_norm": 16.405724753949496,
+      "learning_rate": 1.5025041736227048e-07,
+      "loss": 1.3748,
+      "step": 9
+    },
+    {
+      "epoch": 0.00333889816360601,
+      "grad_norm": 15.483113157235739,
+      "learning_rate": 1.6694490818030052e-07,
+      "loss": 1.3074,
+      "step": 10
+    },
+    {
+      "epoch": 0.003672787979966611,
+      "grad_norm": 15.931594602801674,
+      "learning_rate": 1.8363939899833056e-07,
+      "loss": 1.356,
+      "step": 11
+    },
+    {
+      "epoch": 0.004006677796327212,
+      "grad_norm": 15.942731996958083,
+      "learning_rate": 2.003338898163606e-07,
+      "loss": 1.3406,
+      "step": 12
+    },
+    {
+      "epoch": 0.004340567612687813,
+      "grad_norm": 14.989736329528052,
+      "learning_rate": 2.1702838063439068e-07,
+      "loss": 1.3207,
+      "step": 13
+    },
+    {
+      "epoch": 0.004674457429048414,
+      "grad_norm": 15.905345728716101,
+      "learning_rate": 2.3372287145242072e-07,
+      "loss": 1.341,
+      "step": 14
+    },
+    {
+      "epoch": 0.005008347245409015,
+      "grad_norm": 15.522977851368124,
+      "learning_rate": 2.504173622704508e-07,
+      "loss": 1.3497,
+      "step": 15
+    },
+    {
+      "epoch": 0.005342237061769616,
+      "grad_norm": 15.553746827358466,
+      "learning_rate": 2.671118530884808e-07,
+      "loss": 1.3354,
+      "step": 16
+    },
+    {
+      "epoch": 0.005676126878130217,
+      "grad_norm": 13.730132846014154,
+      "learning_rate": 2.838063439065109e-07,
+      "loss": 1.2667,
+      "step": 17
+    },
+    {
+      "epoch": 0.006010016694490818,
+      "grad_norm": 13.562543818651182,
+      "learning_rate": 3.0050083472454095e-07,
+      "loss": 1.2947,
+      "step": 18
+    },
+    {
+      "epoch": 0.006343906510851419,
+      "grad_norm": 12.787825381547702,
+      "learning_rate": 3.1719532554257097e-07,
+      "loss": 1.2597,
+      "step": 19
+    },
+    {
+      "epoch": 0.00667779632721202,
+      "grad_norm": 12.417513886566905,
+      "learning_rate": 3.3388981636060104e-07,
+      "loss": 1.2407,
+      "step": 20
+    },
+    {
+      "epoch": 0.007011686143572621,
+      "grad_norm": 13.006725198021567,
+      "learning_rate": 3.5058430717863106e-07,
+      "loss": 1.2696,
+      "step": 21
+    },
+    {
+      "epoch": 0.007345575959933222,
+      "grad_norm": 11.744236599890026,
+      "learning_rate": 3.672787979966611e-07,
+      "loss": 1.201,
+      "step": 22
+    },
+    {
+      "epoch": 0.007679465776293823,
+      "grad_norm": 9.440177250177987,
+      "learning_rate": 3.839732888146912e-07,
+      "loss": 1.1329,
+      "step": 23
+    },
+    {
+      "epoch": 0.008013355592654424,
+      "grad_norm": 8.35803088937029,
+      "learning_rate": 4.006677796327212e-07,
+      "loss": 1.1092,
+      "step": 24
+    },
+    {
+      "epoch": 0.008347245409015025,
+      "grad_norm": 8.026898610383288,
+      "learning_rate": 4.173622704507513e-07,
+      "loss": 1.0758,
+      "step": 25
+    },
+    {
+      "epoch": 0.008681135225375626,
+      "grad_norm": 7.946952118013097,
+      "learning_rate": 4.3405676126878136e-07,
+      "loss": 1.0799,
+      "step": 26
+    },
+    {
+      "epoch": 0.009015025041736227,
+      "grad_norm": 7.764749373233746,
+      "learning_rate": 4.5075125208681137e-07,
+      "loss": 1.0592,
+      "step": 27
+    },
+    {
+      "epoch": 0.009348914858096828,
+      "grad_norm": 7.790157830770967,
+      "learning_rate": 4.6744574290484144e-07,
+      "loss": 1.0845,
+      "step": 28
+    },
+    {
+      "epoch": 0.009682804674457429,
+      "grad_norm": 7.692446138013041,
+      "learning_rate": 4.841402337228715e-07,
+      "loss": 1.0634,
+      "step": 29
+    },
+    {
+      "epoch": 0.01001669449081803,
+      "grad_norm": 7.612079199807567,
+      "learning_rate": 5.008347245409016e-07,
+      "loss": 1.0704,
+      "step": 30
+    },
+    {
+      "epoch": 0.010350584307178631,
+      "grad_norm": 6.661736337443945,
+      "learning_rate": 5.175292153589316e-07,
+      "loss": 0.9406,
+      "step": 31
+    },
+    {
+      "epoch": 0.010684474123539232,
+      "grad_norm": 6.594460627437106,
+      "learning_rate": 5.342237061769616e-07,
+      "loss": 0.8449,
+      "step": 32
+    },
+    {
+      "epoch": 0.011018363939899833,
+      "grad_norm": 6.970636919130252,
+      "learning_rate": 5.509181969949917e-07,
+      "loss": 0.7998,
+      "step": 33
+    },
+    {
+      "epoch": 0.011352253756260434,
+      "grad_norm": 7.079168718626818,
+      "learning_rate": 5.676126878130218e-07,
+      "loss": 0.7769,
+      "step": 34
+    },
+    {
+      "epoch": 0.011686143572621035,
+      "grad_norm": 6.903195101391057,
+      "learning_rate": 5.843071786310518e-07,
+      "loss": 0.7737,
+      "step": 35
+    },
+    {
+      "epoch": 0.012020033388981636,
+      "grad_norm": 6.255272544757875,
+      "learning_rate": 6.010016694490819e-07,
+      "loss": 0.7794,
+      "step": 36
+    },
+    {
+      "epoch": 0.012353923205342237,
+      "grad_norm": 5.761484860354292,
+      "learning_rate": 6.176961602671119e-07,
+      "loss": 0.7378,
+      "step": 37
+    },
+    {
+      "epoch": 0.012687813021702838,
+      "grad_norm": 5.700177549906487,
+      "learning_rate": 6.343906510851419e-07,
+      "loss": 0.7445,
+      "step": 38
+    },
+    {
+      "epoch": 0.01302170283806344,
+      "grad_norm": 5.526567271633307,
+      "learning_rate": 6.510851419031721e-07,
+      "loss": 0.7212,
+      "step": 39
+    },
+    {
+      "epoch": 0.01335559265442404,
+      "grad_norm": 5.15276598450368,
+      "learning_rate": 6.677796327212021e-07,
+      "loss": 0.6822,
+      "step": 40
+    },
+    {
+      "epoch": 0.013689482470784642,
+      "grad_norm": 4.465379343131255,
+      "learning_rate": 6.844741235392322e-07,
+      "loss": 0.6476,
+      "step": 41
+    },
+    {
+      "epoch": 0.014023372287145243,
+      "grad_norm": 5.595032803709159,
+      "learning_rate": 7.011686143572621e-07,
+      "loss": 0.5747,
+      "step": 42
+    },
+    {
+      "epoch": 0.014357262103505844,
+      "grad_norm": 3.435997319680881,
+      "learning_rate": 7.178631051752922e-07,
+      "loss": 0.5691,
+      "step": 43
+    },
+    {
+      "epoch": 0.014691151919866445,
+      "grad_norm": 2.6663039330702762,
+      "learning_rate": 7.345575959933223e-07,
+      "loss": 0.5609,
+      "step": 44
+    },
+    {
+      "epoch": 0.015025041736227046,
+      "grad_norm": 2.481030612137373,
+      "learning_rate": 7.512520868113524e-07,
+      "loss": 0.5204,
+      "step": 45
+    },
+    {
+      "epoch": 0.015358931552587647,
+      "grad_norm": 3.001255148213942,
+      "learning_rate": 7.679465776293824e-07,
+      "loss": 0.571,
+      "step": 46
+    },
+    {
+      "epoch": 0.015692821368948246,
+      "grad_norm": 1.970464081874921,
+      "learning_rate": 7.846410684474125e-07,
+      "loss": 0.5314,
+      "step": 47
+    },
+    {
+      "epoch": 0.016026711185308847,
+      "grad_norm": 1.6994853099676646,
+      "learning_rate": 8.013355592654424e-07,
+      "loss": 0.5315,
+      "step": 48
+    },
+    {
+      "epoch": 0.016360601001669448,
+      "grad_norm": 1.6165195049439356,
+      "learning_rate": 8.180300500834726e-07,
+      "loss": 0.5075,
+      "step": 49
+    },
+    {
+      "epoch": 0.01669449081803005,
+      "grad_norm": 1.760111867718028,
+      "learning_rate": 8.347245409015026e-07,
+      "loss": 0.5083,
+      "step": 50
+    },
+    {
+      "epoch": 0.01702838063439065,
+      "grad_norm": 1.362833283488048,
+      "learning_rate": 8.514190317195327e-07,
+      "loss": 0.501,
+      "step": 51
+    },
+    {
+      "epoch": 0.01736227045075125,
+      "grad_norm": 1.4939469223750803,
+      "learning_rate": 8.681135225375627e-07,
+      "loss": 0.4789,
+      "step": 52
+    },
+    {
+      "epoch": 0.017696160267111852,
+      "grad_norm": 1.4560475631465044,
+      "learning_rate": 8.848080133555926e-07,
+      "loss": 0.4755,
+      "step": 53
+    },
+    {
+      "epoch": 0.018030050083472453,
+      "grad_norm": 1.306980561936265,
+      "learning_rate": 9.015025041736227e-07,
+      "loss": 0.4774,
+      "step": 54
+    },
+    {
+      "epoch": 0.018363939899833055,
+      "grad_norm": 1.3753857214774345,
+      "learning_rate": 9.181969949916528e-07,
+      "loss": 0.4823,
+      "step": 55
+    },
+    {
+      "epoch": 0.018697829716193656,
+      "grad_norm": 1.3195766727728997,
+      "learning_rate": 9.348914858096829e-07,
+      "loss": 0.4842,
+      "step": 56
+    },
+    {
+      "epoch": 0.019031719532554257,
+      "grad_norm": 1.1821270910650958,
+      "learning_rate": 9.515859766277129e-07,
+      "loss": 0.465,
+      "step": 57
+    },
+    {
+      "epoch": 0.019365609348914858,
+      "grad_norm": 1.234702802226954,
+      "learning_rate": 9.68280467445743e-07,
+      "loss": 0.4626,
+      "step": 58
+    },
+    {
+      "epoch": 0.01969949916527546,
+      "grad_norm": 1.2893678465596252,
+      "learning_rate": 9.84974958263773e-07,
+      "loss": 0.4489,
+      "step": 59
+    },
+    {
+      "epoch": 0.02003338898163606,
+      "grad_norm": 1.292830420964843,
+      "learning_rate": 1.0016694490818032e-06,
+      "loss": 0.447,
+      "step": 60
+    },
+    {
+      "epoch": 0.02036727879799666,
+      "grad_norm": 1.25121512293509,
+      "learning_rate": 1.0183639398998332e-06,
+      "loss": 0.4584,
+      "step": 61
+    },
+    {
+      "epoch": 0.020701168614357262,
+      "grad_norm": 1.1451600840691205,
+      "learning_rate": 1.0350584307178632e-06,
+      "loss": 0.4319,
+      "step": 62
+    },
+    {
+      "epoch": 0.021035058430717863,
+      "grad_norm": 1.1826200586553755,
+      "learning_rate": 1.0517529215358932e-06,
+      "loss": 0.4595,
+      "step": 63
+    },
+    {
+      "epoch": 0.021368948247078464,
+      "grad_norm": 1.1451322215240287,
+      "learning_rate": 1.0684474123539232e-06,
+      "loss": 0.4611,
+      "step": 64
+    },
+    {
+      "epoch": 0.021702838063439065,
+      "grad_norm": 1.1529511129575065,
+      "learning_rate": 1.0851419031719533e-06,
+      "loss": 0.4506,
+      "step": 65
+    },
+    {
+      "epoch": 0.022036727879799666,
+      "grad_norm": 1.253718899110828,
+      "learning_rate": 1.1018363939899835e-06,
+      "loss": 0.429,
+      "step": 66
+    },
+    {
+      "epoch": 0.022370617696160267,
+      "grad_norm": 1.2051271713931608,
+      "learning_rate": 1.1185308848080135e-06,
+      "loss": 0.4362,
+      "step": 67
+    },
+    {
+      "epoch": 0.02270450751252087,
+      "grad_norm": 1.1196326735133482,
+      "learning_rate": 1.1352253756260435e-06,
+      "loss": 0.4374,
+      "step": 68
+    },
+    {
+      "epoch": 0.02303839732888147,
+      "grad_norm": 1.0803986512239283,
+      "learning_rate": 1.1519198664440735e-06,
+      "loss": 0.4178,
+      "step": 69
+    },
+    {
+      "epoch": 0.02337228714524207,
+      "grad_norm": 1.126471107313093,
+      "learning_rate": 1.1686143572621036e-06,
+      "loss": 0.4215,
+      "step": 70
+    },
+    {
+      "epoch": 0.02370617696160267,
+      "grad_norm": 1.2640876465676731,
+      "learning_rate": 1.1853088480801336e-06,
+      "loss": 0.4268,
+      "step": 71
+    },
+    {
+      "epoch": 0.024040066777963272,
+      "grad_norm": 1.0739720528356986,
+      "learning_rate": 1.2020033388981638e-06,
+      "loss": 0.4244,
+      "step": 72
+    },
+    {
+      "epoch": 0.024373956594323874,
+      "grad_norm": 1.0016987947298317,
+      "learning_rate": 1.2186978297161938e-06,
+      "loss": 0.4074,
+      "step": 73
+    },
+    {
+      "epoch": 0.024707846410684475,
+      "grad_norm": 1.0393104641598188,
+      "learning_rate": 1.2353923205342238e-06,
+      "loss": 0.412,
+      "step": 74
+    },
+    {
+      "epoch": 0.025041736227045076,
+      "grad_norm": 1.0437708038694786,
+      "learning_rate": 1.2520868113522539e-06,
+      "loss": 0.4157,
+      "step": 75
+    },
+    {
+      "epoch": 0.025375626043405677,
+      "grad_norm": 1.0279809301172993,
+      "learning_rate": 1.2687813021702839e-06,
+      "loss": 0.4169,
+      "step": 76
+    },
+    {
+      "epoch": 0.025709515859766278,
+      "grad_norm": 0.9749760135419846,
+      "learning_rate": 1.285475792988314e-06,
+      "loss": 0.4262,
+      "step": 77
+    },
+    {
+      "epoch": 0.02604340567612688,
+      "grad_norm": 1.0635502355215352,
+      "learning_rate": 1.3021702838063441e-06,
+      "loss": 0.4102,
+      "step": 78
+    },
+    {
+      "epoch": 0.02637729549248748,
+      "grad_norm": 1.0568462216267,
+      "learning_rate": 1.3188647746243741e-06,
+      "loss": 0.4094,
+      "step": 79
+    },
+    {
+      "epoch": 0.02671118530884808,
+      "grad_norm": 1.0320858365475376,
+      "learning_rate": 1.3355592654424042e-06,
+      "loss": 0.4058,
+      "step": 80
+    },
+    {
+      "epoch": 0.027045075125208682,
+      "grad_norm": 0.9972316549030238,
+      "learning_rate": 1.352253756260434e-06,
+      "loss": 0.4013,
+      "step": 81
+    },
+    {
+      "epoch": 0.027378964941569283,
+      "grad_norm": 1.0941783108724712,
+      "learning_rate": 1.3689482470784644e-06,
+      "loss": 0.4021,
+      "step": 82
+    },
+    {
+      "epoch": 0.027712854757929884,
+      "grad_norm": 0.9923112460821326,
+      "learning_rate": 1.3856427378964942e-06,
+      "loss": 0.4087,
+      "step": 83
+    },
+    {
+      "epoch": 0.028046744574290485,
+      "grad_norm": 1.0241705371540237,
+      "learning_rate": 1.4023372287145242e-06,
+      "loss": 0.3961,
+      "step": 84
+    },
+    {
+      "epoch": 0.028380634390651086,
+      "grad_norm": 1.0699530855181592,
+      "learning_rate": 1.4190317195325542e-06,
+      "loss": 0.4075,
+      "step": 85
+    },
+    {
+      "epoch": 0.028714524207011687,
+      "grad_norm": 1.0894268479562248,
+      "learning_rate": 1.4357262103505845e-06,
+      "loss": 0.3991,
+      "step": 86
+    },
+    {
+      "epoch": 0.02904841402337229,
+      "grad_norm": 1.0656080946365216,
+      "learning_rate": 1.4524207011686145e-06,
+      "loss": 0.3986,
+      "step": 87
+    },
+    {
+      "epoch": 0.02938230383973289,
+      "grad_norm": 0.9804844769287893,
+      "learning_rate": 1.4691151919866445e-06,
+      "loss": 0.416,
+      "step": 88
+    },
+    {
+      "epoch": 0.02971619365609349,
+      "grad_norm": 1.1601174508623375,
+      "learning_rate": 1.4858096828046745e-06,
+      "loss": 0.4043,
+      "step": 89
+    },
+    {
+      "epoch": 0.03005008347245409,
+      "grad_norm": 1.0587181139726112,
+      "learning_rate": 1.5025041736227048e-06,
+      "loss": 0.4093,
+      "step": 90
+    },
+    {
+      "epoch": 0.030383973288814693,
+      "grad_norm": 1.0405172685829454,
+      "learning_rate": 1.5191986644407348e-06,
+      "loss": 0.3945,
+      "step": 91
+    },
+    {
+      "epoch": 0.030717863105175294,
+      "grad_norm": 1.0473166176960913,
+      "learning_rate": 1.5358931552587648e-06,
+      "loss": 0.3876,
+      "step": 92
+    },
+    {
+      "epoch": 0.031051752921535895,
+      "grad_norm": 1.0407718848445233,
+      "learning_rate": 1.5525876460767946e-06,
+      "loss": 0.3903,
+      "step": 93
+    },
+    {
+      "epoch": 0.03138564273789649,
+      "grad_norm": 1.11288667655594,
+      "learning_rate": 1.569282136894825e-06,
+      "loss": 0.3918,
+      "step": 94
+    },
+    {
+      "epoch": 0.03171953255425709,
+      "grad_norm": 1.0470971385439736,
+      "learning_rate": 1.5859766277128548e-06,
+      "loss": 0.4018,
+      "step": 95
+    },
+    {
+      "epoch": 0.032053422370617694,
+      "grad_norm": 1.0933798298299329,
+      "learning_rate": 1.6026711185308849e-06,
+      "loss": 0.4138,
+      "step": 96
+    },
+    {
+      "epoch": 0.032387312186978295,
+      "grad_norm": 1.0504094866144102,
+      "learning_rate": 1.6193656093489149e-06,
+      "loss": 0.3921,
+      "step": 97
+    },
+    {
+      "epoch": 0.032721202003338896,
+      "grad_norm": 1.0541848623259034,
+      "learning_rate": 1.636060100166945e-06,
+      "loss": 0.3812,
+      "step": 98
+    },
+    {
+      "epoch": 0.0330550918196995,
+      "grad_norm": 1.0531173697100127,
+      "learning_rate": 1.6527545909849751e-06,
+      "loss": 0.3969,
+      "step": 99
+    },
+    {
+      "epoch": 0.0333889816360601,
+      "grad_norm": 0.9885597378007323,
+      "learning_rate": 1.6694490818030051e-06,
+      "loss": 0.3912,
+      "step": 100
+    },
+    {
+      "epoch": 0.0337228714524207,
+      "grad_norm": 1.0368530772417002,
+      "learning_rate": 1.6861435726210352e-06,
+      "loss": 0.3894,
+      "step": 101
+    },
+    {
+      "epoch": 0.0340567612687813,
+      "grad_norm": 1.129744774812659,
+      "learning_rate": 1.7028380634390654e-06,
+      "loss": 0.3901,
+      "step": 102
+    },
+    {
+      "epoch": 0.0343906510851419,
+      "grad_norm": 1.1175935058711781,
+      "learning_rate": 1.7195325542570954e-06,
+      "loss": 0.3856,
+      "step": 103
+    },
+    {
+      "epoch": 0.0347245409015025,
+      "grad_norm": 0.9878020321666762,
+      "learning_rate": 1.7362270450751254e-06,
+      "loss": 0.3861,
+      "step": 104
+    },
+    {
+      "epoch": 0.035058430717863104,
+      "grad_norm": 0.9950210047550576,
+      "learning_rate": 1.7529215358931554e-06,
+      "loss": 0.3731,
+      "step": 105
+    },
+    {
+      "epoch": 0.035392320534223705,
+      "grad_norm": 1.1294037734463782,
+      "learning_rate": 1.7696160267111852e-06,
+      "loss": 0.3768,
+      "step": 106
+    },
+    {
+      "epoch": 0.035726210350584306,
+      "grad_norm": 1.003221522287488,
+      "learning_rate": 1.7863105175292157e-06,
+      "loss": 0.4086,
+      "step": 107
+    },
+    {
+      "epoch": 0.03606010016694491,
+      "grad_norm": 1.0291958727377342,
+      "learning_rate": 1.8030050083472455e-06,
+      "loss": 0.3899,
+      "step": 108
+    },
+    {
+      "epoch": 0.03639398998330551,
+      "grad_norm": 1.071788738118953,
+      "learning_rate": 1.8196994991652755e-06,
+      "loss": 0.3871,
+      "step": 109
+    },
+    {
+      "epoch": 0.03672787979966611,
+      "grad_norm": 1.009047977529497,
+      "learning_rate": 1.8363939899833055e-06,
+      "loss": 0.3878,
+      "step": 110
+    },
+    {
+      "epoch": 0.03706176961602671,
+      "grad_norm": 1.0808834484545555,
+      "learning_rate": 1.8530884808013358e-06,
+      "loss": 0.3889,
+      "step": 111
+    },
+    {
+      "epoch": 0.03739565943238731,
+      "grad_norm": 0.9841900530307557,
+      "learning_rate": 1.8697829716193658e-06,
+      "loss": 0.37,
+      "step": 112
+    },
+    {
+      "epoch": 0.03772954924874791,
+      "grad_norm": 0.987101803284381,
+      "learning_rate": 1.8864774624373958e-06,
+      "loss": 0.3775,
+      "step": 113
+    },
+    {
+      "epoch": 0.03806343906510851,
+      "grad_norm": 1.0140972005874942,
+      "learning_rate": 1.9031719532554258e-06,
+      "loss": 0.3742,
+      "step": 114
+    },
+    {
+      "epoch": 0.038397328881469114,
+      "grad_norm": 0.9907900303261296,
+      "learning_rate": 1.919866444073456e-06,
+      "loss": 0.3829,
+      "step": 115
+    },
+    {
+      "epoch": 0.038731218697829715,
+      "grad_norm": 0.9524403784131362,
+      "learning_rate": 1.936560934891486e-06,
+      "loss": 0.3593,
+      "step": 116
+    },
+    {
+      "epoch": 0.039065108514190316,
+      "grad_norm": 0.9285482019437789,
+      "learning_rate": 1.953255425709516e-06,
+      "loss": 0.3761,
+      "step": 117
+    },
+    {
+      "epoch": 0.03939899833055092,
+      "grad_norm": 1.0255636705419258,
+      "learning_rate": 1.969949916527546e-06,
+      "loss": 0.3743,
+      "step": 118
+    },
+    {
+      "epoch": 0.03973288814691152,
+      "grad_norm": 1.0033161476626764,
+      "learning_rate": 1.9866444073455763e-06,
+      "loss": 0.3741,
+      "step": 119
+    },
+    {
+      "epoch": 0.04006677796327212,
+      "grad_norm": 0.9804320215021829,
+      "learning_rate": 2.0033388981636063e-06,
+      "loss": 0.3804,
+      "step": 120
+    },
+    {
+      "epoch": 0.04040066777963272,
+      "grad_norm": 1.0013065561990921,
+      "learning_rate": 2.0200333889816364e-06,
+      "loss": 0.3728,
+      "step": 121
+    },
+    {
+      "epoch": 0.04073455759599332,
+      "grad_norm": 0.9758935864064443,
+      "learning_rate": 2.0367278797996664e-06,
+      "loss": 0.3767,
+      "step": 122
+    },
+    {
+      "epoch": 0.04106844741235392,
+      "grad_norm": 1.053705703124745,
+      "learning_rate": 2.0534223706176964e-06,
+      "loss": 0.3846,
+      "step": 123
+    },
+    {
+      "epoch": 0.041402337228714524,
+      "grad_norm": 0.9983587996574941,
+      "learning_rate": 2.0701168614357264e-06,
+      "loss": 0.3829,
+      "step": 124
+    },
+    {
+      "epoch": 0.041736227045075125,
+      "grad_norm": 0.9808004135545196,
+      "learning_rate": 2.0868113522537564e-06,
+      "loss": 0.384,
+      "step": 125
+    },
+    {
+      "epoch": 0.042070116861435726,
+      "grad_norm": 1.0279822401331336,
+      "learning_rate": 2.1035058430717864e-06,
+      "loss": 0.3597,
+      "step": 126
+    },
+    {
+      "epoch": 0.04240400667779633,
+      "grad_norm": 1.0534411548755027,
+      "learning_rate": 2.1202003338898165e-06,
+      "loss": 0.3685,
+      "step": 127
+    },
+    {
+      "epoch": 0.04273789649415693,
+      "grad_norm": 1.0232733499632214,
+      "learning_rate": 2.1368948247078465e-06,
+      "loss": 0.377,
+      "step": 128
+    },
+    {
+      "epoch": 0.04307178631051753,
+      "grad_norm": 0.9494272401589524,
+      "learning_rate": 2.1535893155258765e-06,
+      "loss": 0.3602,
+      "step": 129
+    },
+    {
+      "epoch": 0.04340567612687813,
+      "grad_norm": 1.0782203719142778,
+      "learning_rate": 2.1702838063439065e-06,
+      "loss": 0.3768,
+      "step": 130
+    },
+    {
+      "epoch": 0.04373956594323873,
+      "grad_norm": 0.959238775163698,
+      "learning_rate": 2.1869782971619365e-06,
+      "loss": 0.3774,
+      "step": 131
+    },
+    {
+      "epoch": 0.04407345575959933,
+      "grad_norm": 0.9648879956027536,
+      "learning_rate": 2.203672787979967e-06,
+      "loss": 0.3672,
+      "step": 132
+    },
+    {
+      "epoch": 0.04440734557595993,
+      "grad_norm": 0.9732432531162775,
+      "learning_rate": 2.220367278797997e-06,
+      "loss": 0.3744,
+      "step": 133
+    },
+    {
+      "epoch": 0.044741235392320534,
+      "grad_norm": 1.064060661429251,
+      "learning_rate": 2.237061769616027e-06,
+      "loss": 0.3795,
+      "step": 134
+    },
+    {
+      "epoch": 0.045075125208681135,
+      "grad_norm": 1.0374796066545173,
+      "learning_rate": 2.253756260434057e-06,
+      "loss": 0.3772,
+      "step": 135
+    },
+    {
+      "epoch": 0.04540901502504174,
+      "grad_norm": 0.9944826115257457,
+      "learning_rate": 2.270450751252087e-06,
+      "loss": 0.369,
+      "step": 136
+    },
+    {
+      "epoch": 0.04574290484140234,
+      "grad_norm": 1.0213700376765071,
+      "learning_rate": 2.287145242070117e-06,
+      "loss": 0.3584,
+      "step": 137
+    },
+    {
+      "epoch": 0.04607679465776294,
+      "grad_norm": 0.9700428378282023,
+      "learning_rate": 2.303839732888147e-06,
+      "loss": 0.3551,
+      "step": 138
+    },
+    {
+      "epoch": 0.04641068447412354,
+      "grad_norm": 1.0788447146617486,
+      "learning_rate": 2.320534223706177e-06,
+      "loss": 0.3633,
+      "step": 139
+    },
+    {
+      "epoch": 0.04674457429048414,
+      "grad_norm": 1.0252083135131975,
+      "learning_rate": 2.337228714524207e-06,
+      "loss": 0.3641,
+      "step": 140
+    },
+    {
+      "epoch": 0.04707846410684474,
+      "grad_norm": 0.9862081146917875,
+      "learning_rate": 2.353923205342237e-06,
+      "loss": 0.363,
+      "step": 141
+    },
+    {
+      "epoch": 0.04741235392320534,
+      "grad_norm": 1.0653771735164275,
+      "learning_rate": 2.370617696160267e-06,
+      "loss": 0.372,
+      "step": 142
+    },
+    {
+      "epoch": 0.047746243739565944,
+      "grad_norm": 1.1352747537042513,
+      "learning_rate": 2.387312186978297e-06,
+      "loss": 0.3653,
+      "step": 143
+    },
+    {
+      "epoch": 0.048080133555926545,
+      "grad_norm": 1.0886245831306176,
+      "learning_rate": 2.4040066777963276e-06,
+      "loss": 0.3512,
+      "step": 144
+    },
+    {
+      "epoch": 0.048414023372287146,
+      "grad_norm": 1.081505609928735,
+      "learning_rate": 2.4207011686143576e-06,
+      "loss": 0.3521,
+      "step": 145
+    },
+    {
+      "epoch": 0.04874791318864775,
+      "grad_norm": 1.2541034473944053,
+      "learning_rate": 2.4373956594323876e-06,
+      "loss": 0.3694,
+      "step": 146
+    },
+    {
+      "epoch": 0.04908180300500835,
+      "grad_norm": 1.103644065446481,
+      "learning_rate": 2.4540901502504177e-06,
+      "loss": 0.3645,
+      "step": 147
+    },
+    {
+      "epoch": 0.04941569282136895,
+      "grad_norm": 1.0199778329488214,
+      "learning_rate": 2.4707846410684477e-06,
+      "loss": 0.364,
+      "step": 148
+    },
+    {
+      "epoch": 0.04974958263772955,
+      "grad_norm": 1.0302437384211593,
+      "learning_rate": 2.4874791318864777e-06,
+      "loss": 0.3603,
+      "step": 149
+    },
+    {
+      "epoch": 0.05008347245409015,
+      "grad_norm": 1.0304048225067552,
+      "learning_rate": 2.5041736227045077e-06,
+      "loss": 0.3587,
+      "step": 150
+    },
+    {
+      "epoch": 0.05041736227045075,
+      "grad_norm": 1.0069711156098722,
+      "learning_rate": 2.5208681135225377e-06,
+      "loss": 0.364,
+      "step": 151
+    },
+    {
+      "epoch": 0.05075125208681135,
+      "grad_norm": 1.0434841540104263,
+      "learning_rate": 2.5375626043405677e-06,
+      "loss": 0.3558,
+      "step": 152
+    },
+    {
+      "epoch": 0.051085141903171954,
+      "grad_norm": 1.0901316675320962,
+      "learning_rate": 2.5542570951585978e-06,
+      "loss": 0.3548,
+      "step": 153
+    },
+    {
+      "epoch": 0.051419031719532556,
+      "grad_norm": 1.0160826783323413,
+      "learning_rate": 2.570951585976628e-06,
+      "loss": 0.3637,
+      "step": 154
+    },
+    {
+      "epoch": 0.05175292153589316,
+      "grad_norm": 1.0366948107138263,
+      "learning_rate": 2.5876460767946582e-06,
+      "loss": 0.3564,
+      "step": 155
+    },
+    {
+      "epoch": 0.05208681135225376,
+      "grad_norm": 0.9790026991641014,
+      "learning_rate": 2.6043405676126882e-06,
+      "loss": 0.3533,
+      "step": 156
+    },
+    {
+      "epoch": 0.05242070116861436,
+      "grad_norm": 0.9907810573817663,
+      "learning_rate": 2.6210350584307183e-06,
+      "loss": 0.3666,
+      "step": 157
+    },
+    {
+      "epoch": 0.05275459098497496,
+      "grad_norm": 0.9729528142445464,
+      "learning_rate": 2.6377295492487483e-06,
+      "loss": 0.354,
+      "step": 158
+    },
+    {
+      "epoch": 0.05308848080133556,
+      "grad_norm": 1.0811710094004001,
+      "learning_rate": 2.6544240400667783e-06,
+      "loss": 0.3572,
+      "step": 159
+    },
+    {
+      "epoch": 0.05342237061769616,
+      "grad_norm": 1.1217527008322312,
+      "learning_rate": 2.6711185308848083e-06,
+      "loss": 0.3577,
+      "step": 160
+    },
+    {
+      "epoch": 0.05375626043405676,
+      "grad_norm": 0.9994182649926079,
+      "learning_rate": 2.687813021702838e-06,
+      "loss": 0.3347,
+      "step": 161
+    },
+    {
+      "epoch": 0.054090150250417364,
+      "grad_norm": 0.9797507731887032,
+      "learning_rate": 2.704507512520868e-06,
+      "loss": 0.3581,
+      "step": 162
+    },
+    {
+      "epoch": 0.054424040066777965,
+      "grad_norm": 1.1290007199905177,
+      "learning_rate": 2.7212020033388988e-06,
+      "loss": 0.3614,
+      "step": 163
+    },
+    {
+      "epoch": 0.054757929883138566,
+      "grad_norm": 1.0031495772870516,
+      "learning_rate": 2.737896494156929e-06,
+      "loss": 0.356,
+      "step": 164
+    },
+    {
+      "epoch": 0.05509181969949917,
+      "grad_norm": 0.9637894890574632,
+      "learning_rate": 2.7545909849749584e-06,
+      "loss": 0.361,
+      "step": 165
+    },
+    {
+      "epoch": 0.05542570951585977,
+      "grad_norm": 1.0072845365626804,
+      "learning_rate": 2.7712854757929884e-06,
+      "loss": 0.3505,
+      "step": 166
+    },
+    {
+      "epoch": 0.05575959933222037,
+      "grad_norm": 1.0768788270422753,
+      "learning_rate": 2.7879799666110184e-06,
+      "loss": 0.3568,
+      "step": 167
+    },
+    {
+      "epoch": 0.05609348914858097,
+      "grad_norm": 1.0548664322555334,
+      "learning_rate": 2.8046744574290484e-06,
+      "loss": 0.3599,
+      "step": 168
+    },
+    {
+      "epoch": 0.05642737896494157,
+      "grad_norm": 0.9464119649751291,
+      "learning_rate": 2.8213689482470785e-06,
+      "loss": 0.3587,
+      "step": 169
+    },
+    {
+      "epoch": 0.05676126878130217,
+      "grad_norm": 1.0125435958219176,
+      "learning_rate": 2.8380634390651085e-06,
+      "loss": 0.3458,
+      "step": 170
+    },
+    {
+      "epoch": 0.05709515859766277,
+      "grad_norm": 1.0214997863138064,
+      "learning_rate": 2.854757929883139e-06,
+      "loss": 0.3403,
+      "step": 171
+    },
+    {
+      "epoch": 0.057429048414023375,
+      "grad_norm": 0.9626106859311915,
+      "learning_rate": 2.871452420701169e-06,
+      "loss": 0.3653,
+      "step": 172
+    },
+    {
+      "epoch": 0.057762938230383976,
+      "grad_norm": 0.9908053766252171,
+      "learning_rate": 2.888146911519199e-06,
+      "loss": 0.3545,
+      "step": 173
+    },
+    {
+      "epoch": 0.05809682804674458,
+      "grad_norm": 0.9998982637726064,
+      "learning_rate": 2.904841402337229e-06,
+      "loss": 0.35,
+      "step": 174
+    },
+    {
+      "epoch": 0.05843071786310518,
+      "grad_norm": 0.994549674708493,
+      "learning_rate": 2.921535893155259e-06,
+      "loss": 0.3554,
+      "step": 175
+    },
+    {
+      "epoch": 0.05876460767946578,
+      "grad_norm": 0.9459191873892051,
+      "learning_rate": 2.938230383973289e-06,
+      "loss": 0.3693,
+      "step": 176
+    },
+    {
+      "epoch": 0.05909849749582638,
+      "grad_norm": 0.9411219919933045,
+      "learning_rate": 2.954924874791319e-06,
+      "loss": 0.3408,
+      "step": 177
+    },
+    {
+      "epoch": 0.05943238731218698,
+      "grad_norm": 0.967669386266833,
+      "learning_rate": 2.971619365609349e-06,
+      "loss": 0.37,
+      "step": 178
+    },
+    {
+      "epoch": 0.05976627712854758,
+      "grad_norm": 0.9750095503557731,
+      "learning_rate": 2.9883138564273795e-06,
+      "loss": 0.3518,
+      "step": 179
+    },
+    {
+      "epoch": 0.06010016694490818,
+      "grad_norm": 0.9869992334152199,
+      "learning_rate": 3.0050083472454095e-06,
+      "loss": 0.3596,
+      "step": 180
+    },
+    {
+      "epoch": 0.060434056761268784,
+      "grad_norm": 0.9941235196882385,
+      "learning_rate": 3.0217028380634395e-06,
+      "loss": 0.3559,
+      "step": 181
+    },
+    {
+      "epoch": 0.060767946577629385,
+      "grad_norm": 0.9478911053006368,
+      "learning_rate": 3.0383973288814695e-06,
+      "loss": 0.3572,
+      "step": 182
+    },
+    {
+      "epoch": 0.061101836393989986,
+      "grad_norm": 0.9439517398276193,
+      "learning_rate": 3.0550918196994996e-06,
+      "loss": 0.3438,
+      "step": 183
+    },
+    {
+      "epoch": 0.06143572621035059,
+      "grad_norm": 1.054985996601042,
+      "learning_rate": 3.0717863105175296e-06,
+      "loss": 0.3492,
+      "step": 184
+    },
+    {
+      "epoch": 0.06176961602671119,
+      "grad_norm": 1.0008519931147941,
+      "learning_rate": 3.0884808013355596e-06,
+      "loss": 0.3554,
+      "step": 185
+    },
+    {
+      "epoch": 0.06210350584307179,
+      "grad_norm": 1.0383206448639142,
+      "learning_rate": 3.105175292153589e-06,
+      "loss": 0.3737,
+      "step": 186
+    },
+    {
+      "epoch": 0.06243739565943239,
+      "grad_norm": 0.951233970592366,
+      "learning_rate": 3.121869782971619e-06,
+      "loss": 0.3631,
+      "step": 187
+    },
+    {
+      "epoch": 0.06277128547579298,
+      "grad_norm": 0.9399845544816917,
+      "learning_rate": 3.13856427378965e-06,
+      "loss": 0.3569,
+      "step": 188
+    },
+    {
+      "epoch": 0.06310517529215359,
+      "grad_norm": 0.9398189327271754,
+      "learning_rate": 3.15525876460768e-06,
+      "loss": 0.3505,
+      "step": 189
+    },
+    {
+      "epoch": 0.06343906510851419,
+      "grad_norm": 0.9640702315168751,
+      "learning_rate": 3.1719532554257097e-06,
+      "loss": 0.36,
+      "step": 190
+    },
+    {
+      "epoch": 0.0637729549248748,
+      "grad_norm": 0.984878576045248,
+      "learning_rate": 3.1886477462437397e-06,
+      "loss": 0.3602,
+      "step": 191
+    },
+    {
+      "epoch": 0.06410684474123539,
+      "grad_norm": 0.9326105611561197,
+      "learning_rate": 3.2053422370617697e-06,
+      "loss": 0.3489,
+      "step": 192
+    },
+    {
+      "epoch": 0.064440734557596,
+      "grad_norm": 0.871513512768245,
+      "learning_rate": 3.2220367278797997e-06,
+      "loss": 0.3546,
+      "step": 193
+    },
+    {
+      "epoch": 0.06477462437395659,
+      "grad_norm": 0.9419481324220705,
+      "learning_rate": 3.2387312186978297e-06,
+      "loss": 0.3512,
+      "step": 194
+    },
+    {
+      "epoch": 0.0651085141903172,
+      "grad_norm": 1.0242403896723964,
+      "learning_rate": 3.2554257095158598e-06,
+      "loss": 0.3455,
+      "step": 195
+    },
+    {
+      "epoch": 0.06544240400667779,
+      "grad_norm": 0.9402397508522083,
+      "learning_rate": 3.27212020033389e-06,
+      "loss": 0.3472,
+      "step": 196
+    },
+    {
+      "epoch": 0.0657762938230384,
+      "grad_norm": 0.9089400970434404,
+      "learning_rate": 3.2888146911519202e-06,
+      "loss": 0.3436,
+      "step": 197
+    },
+    {
+      "epoch": 0.066110183639399,
+      "grad_norm": 0.9866750435687955,
+      "learning_rate": 3.3055091819699502e-06,
+      "loss": 0.3412,
+      "step": 198
+    },
+    {
+      "epoch": 0.0664440734557596,
+      "grad_norm": 0.911926129378434,
+      "learning_rate": 3.3222036727879803e-06,
+      "loss": 0.3537,
+      "step": 199
+    },
+    {
+      "epoch": 0.0667779632721202,
+      "grad_norm": 0.9725619714909971,
+      "learning_rate": 3.3388981636060103e-06,
+      "loss": 0.3461,
+      "step": 200
+    },
+    {
+      "epoch": 0.0671118530884808,
+      "grad_norm": 1.0279558346616509,
+      "learning_rate": 3.3555926544240403e-06,
+      "loss": 0.3322,
+      "step": 201
+    },
+    {
+      "epoch": 0.0674457429048414,
+      "grad_norm": 0.939918705954264,
+      "learning_rate": 3.3722871452420703e-06,
+      "loss": 0.3419,
+      "step": 202
+    },
+    {
+      "epoch": 0.06777963272120201,
+      "grad_norm": 0.9382322613619366,
+      "learning_rate": 3.3889816360601003e-06,
+      "loss": 0.3366,
+      "step": 203
+    },
+    {
+      "epoch": 0.0681135225375626,
+      "grad_norm": 0.9566616841118353,
+      "learning_rate": 3.4056761268781308e-06,
+      "loss": 0.3595,
+      "step": 204
+    },
+    {
+      "epoch": 0.06844741235392321,
+      "grad_norm": 0.9438871036275341,
+      "learning_rate": 3.422370617696161e-06,
+      "loss": 0.3532,
+      "step": 205
+    },
+    {
+      "epoch": 0.0687813021702838,
+      "grad_norm": 0.9192692809929508,
+      "learning_rate": 3.439065108514191e-06,
+      "loss": 0.3383,
+      "step": 206
+    },
+    {
+      "epoch": 0.06911519198664441,
+      "grad_norm": 1.048775910779933,
+      "learning_rate": 3.455759599332221e-06,
+      "loss": 0.3426,
+      "step": 207
+    },
+    {
+      "epoch": 0.069449081803005,
+      "grad_norm": 1.026369951012414,
+      "learning_rate": 3.472454090150251e-06,
+      "loss": 0.356,
+      "step": 208
+    },
+    {
+      "epoch": 0.06978297161936561,
+      "grad_norm": 0.9196593629912507,
+      "learning_rate": 3.489148580968281e-06,
+      "loss": 0.3452,
+      "step": 209
+    },
+    {
+      "epoch": 0.07011686143572621,
+      "grad_norm": 0.9406661121369319,
+      "learning_rate": 3.505843071786311e-06,
+      "loss": 0.3482,
+      "step": 210
+    },
+    {
+      "epoch": 0.07045075125208682,
+      "grad_norm": 1.0278277374600837,
+      "learning_rate": 3.5225375626043405e-06,
+      "loss": 0.3482,
+      "step": 211
+    },
+    {
+      "epoch": 0.07078464106844741,
+      "grad_norm": 0.9531665078517667,
+      "learning_rate": 3.5392320534223705e-06,
+      "loss": 0.3504,
+      "step": 212
+    },
+    {
+      "epoch": 0.07111853088480802,
+      "grad_norm": 0.9625467154992694,
+      "learning_rate": 3.5559265442404014e-06,
+      "loss": 0.3456,
+      "step": 213
+    },
+    {
+      "epoch": 0.07145242070116861,
+      "grad_norm": 0.8807832141891921,
+      "learning_rate": 3.5726210350584314e-06,
+      "loss": 0.3467,
+      "step": 214
+    },
+    {
+      "epoch": 0.07178631051752922,
+      "grad_norm": 0.9484531598505749,
+      "learning_rate": 3.589315525876461e-06,
+      "loss": 0.3386,
+      "step": 215
+    },
+    {
+      "epoch": 0.07212020033388981,
+      "grad_norm": 0.9621440385975512,
+      "learning_rate": 3.606010016694491e-06,
+      "loss": 0.3406,
+      "step": 216
+    },
+    {
+      "epoch": 0.07245409015025042,
+      "grad_norm": 0.9739140813117508,
+      "learning_rate": 3.622704507512521e-06,
+      "loss": 0.3401,
+      "step": 217
+    },
+    {
+      "epoch": 0.07278797996661102,
+      "grad_norm": 1.045376795312426,
+      "learning_rate": 3.639398998330551e-06,
+      "loss": 0.3567,
+      "step": 218
+    },
+    {
+      "epoch": 0.07312186978297162,
+      "grad_norm": 0.9326856447010807,
+      "learning_rate": 3.656093489148581e-06,
+      "loss": 0.344,
+      "step": 219
+    },
+    {
+      "epoch": 0.07345575959933222,
+      "grad_norm": 1.033511525385637,
+      "learning_rate": 3.672787979966611e-06,
+      "loss": 0.343,
+      "step": 220
+    },
+    {
+      "epoch": 0.07378964941569283,
+      "grad_norm": 0.9195543347869846,
+      "learning_rate": 3.6894824707846415e-06,
+      "loss": 0.3468,
+      "step": 221
+    },
+    {
+      "epoch": 0.07412353923205342,
+      "grad_norm": 0.9164815845532063,
+      "learning_rate": 3.7061769616026715e-06,
+      "loss": 0.3313,
+      "step": 222
+    },
+    {
+      "epoch": 0.07445742904841403,
+      "grad_norm": 0.9790256224733047,
+      "learning_rate": 3.7228714524207015e-06,
+      "loss": 0.3481,
+      "step": 223
+    },
+    {
+      "epoch": 0.07479131886477462,
+      "grad_norm": 0.9919236790720759,
+      "learning_rate": 3.7395659432387315e-06,
+      "loss": 0.3543,
+      "step": 224
+    },
+    {
+      "epoch": 0.07512520868113523,
+      "grad_norm": 0.9433967527993085,
+      "learning_rate": 3.7562604340567616e-06,
+      "loss": 0.3393,
+      "step": 225
+    },
+    {
+      "epoch": 0.07545909849749582,
+      "grad_norm": 1.030450622443354,
+      "learning_rate": 3.7729549248747916e-06,
+      "loss": 0.3485,
+      "step": 226
+    },
+    {
+      "epoch": 0.07579298831385643,
+      "grad_norm": 1.006505540995668,
+      "learning_rate": 3.7896494156928216e-06,
+      "loss": 0.3603,
+      "step": 227
+    },
+    {
+      "epoch": 0.07612687813021703,
+      "grad_norm": 0.9754555797361768,
+      "learning_rate": 3.8063439065108516e-06,
+      "loss": 0.3458,
+      "step": 228
+    },
+    {
+      "epoch": 0.07646076794657763,
+      "grad_norm": 1.0178654617663587,
+      "learning_rate": 3.823038397328882e-06,
+      "loss": 0.3386,
+      "step": 229
+    },
+    {
+      "epoch": 0.07679465776293823,
+      "grad_norm": 0.9404673155461375,
+      "learning_rate": 3.839732888146912e-06,
+      "loss": 0.3489,
+      "step": 230
+    },
+    {
+      "epoch": 0.07712854757929884,
+      "grad_norm": 0.9148869164459803,
+      "learning_rate": 3.856427378964942e-06,
+      "loss": 0.35,
+      "step": 231
+    },
+    {
+      "epoch": 0.07746243739565943,
+      "grad_norm": 1.0615960517926055,
+      "learning_rate": 3.873121869782972e-06,
+      "loss": 0.3294,
+      "step": 232
+    },
+    {
+      "epoch": 0.07779632721202004,
+      "grad_norm": 0.9412402084700151,
+      "learning_rate": 3.889816360601002e-06,
+      "loss": 0.3411,
+      "step": 233
+    },
+    {
+      "epoch": 0.07813021702838063,
+      "grad_norm": 0.883065946909578,
+      "learning_rate": 3.906510851419032e-06,
+      "loss": 0.3377,
+      "step": 234
+    },
+    {
+      "epoch": 0.07846410684474124,
+      "grad_norm": 0.9608751450882079,
+      "learning_rate": 3.923205342237062e-06,
+      "loss": 0.3393,
+      "step": 235
+    },
+    {
+      "epoch": 0.07879799666110184,
+      "grad_norm": 0.9392199302436275,
+      "learning_rate": 3.939899833055092e-06,
+      "loss": 0.3501,
+      "step": 236
+    },
+    {
+      "epoch": 0.07913188647746244,
+      "grad_norm": 0.9172520724032756,
+      "learning_rate": 3.956594323873122e-06,
+      "loss": 0.3421,
+      "step": 237
+    },
+    {
+      "epoch": 0.07946577629382304,
+      "grad_norm": 0.8952203398564054,
+      "learning_rate": 3.973288814691153e-06,
+      "loss": 0.3448,
+      "step": 238
+    },
+    {
+      "epoch": 0.07979966611018365,
+      "grad_norm": 0.9804813133039779,
+      "learning_rate": 3.989983305509183e-06,
+      "loss": 0.343,
+      "step": 239
+    },
+    {
+      "epoch": 0.08013355592654424,
+      "grad_norm": 0.8573500639064191,
+      "learning_rate": 4.006677796327213e-06,
+      "loss": 0.3313,
+      "step": 240
+    },
+    {
+      "epoch": 0.08046744574290485,
+      "grad_norm": 0.9218133267409421,
+      "learning_rate": 4.023372287145243e-06,
+      "loss": 0.3351,
+      "step": 241
+    },
+    {
+      "epoch": 0.08080133555926544,
+      "grad_norm": 0.9464205868254937,
+      "learning_rate": 4.040066777963273e-06,
+      "loss": 0.3469,
+      "step": 242
+    },
+    {
+      "epoch": 0.08113522537562605,
+      "grad_norm": 0.9547743624927089,
+      "learning_rate": 4.056761268781303e-06,
+      "loss": 0.3333,
+      "step": 243
+    },
+    {
+      "epoch": 0.08146911519198664,
+      "grad_norm": 0.9385401995076604,
+      "learning_rate": 4.073455759599333e-06,
+      "loss": 0.3416,
+      "step": 244
+    },
+    {
+      "epoch": 0.08180300500834725,
+      "grad_norm": 0.9087946853489387,
+      "learning_rate": 4.090150250417363e-06,
+      "loss": 0.3444,
+      "step": 245
+    },
+    {
+      "epoch": 0.08213689482470785,
+      "grad_norm": 0.9226187378521882,
+      "learning_rate": 4.106844741235393e-06,
+      "loss": 0.3531,
+      "step": 246
+    },
+    {
+      "epoch": 0.08247078464106845,
+      "grad_norm": 0.9316388099316362,
+      "learning_rate": 4.123539232053423e-06,
+      "loss": 0.3339,
+      "step": 247
+    },
+    {
+      "epoch": 0.08280467445742905,
+      "grad_norm": 0.9027192877815473,
+      "learning_rate": 4.140233722871453e-06,
+      "loss": 0.3405,
+      "step": 248
+    },
+    {
+      "epoch": 0.08313856427378966,
+      "grad_norm": 0.9357213335479192,
+      "learning_rate": 4.156928213689483e-06,
+      "loss": 0.3388,
+      "step": 249
+    },
+    {
+      "epoch": 0.08347245409015025,
+      "grad_norm": 0.9605851645041922,
+      "learning_rate": 4.173622704507513e-06,
+      "loss": 0.3327,
+      "step": 250
+    },
+    {
+      "epoch": 0.08380634390651086,
+      "grad_norm": 0.9331326912802664,
+      "learning_rate": 4.190317195325543e-06,
+      "loss": 0.3494,
+      "step": 251
+    },
+    {
+      "epoch": 0.08414023372287145,
+      "grad_norm": 0.9693211160880641,
+      "learning_rate": 4.207011686143573e-06,
+      "loss": 0.3266,
+      "step": 252
+    },
+    {
+      "epoch": 0.08447412353923206,
+      "grad_norm": 0.9902022599960142,
+      "learning_rate": 4.223706176961603e-06,
+      "loss": 0.3611,
+      "step": 253
+    },
+    {
+      "epoch": 0.08480801335559265,
+      "grad_norm": 0.9374039100644388,
+      "learning_rate": 4.240400667779633e-06,
+      "loss": 0.3328,
+      "step": 254
+    },
+    {
+      "epoch": 0.08514190317195326,
+      "grad_norm": 0.8719108117732441,
+      "learning_rate": 4.257095158597663e-06,
+      "loss": 0.3297,
+      "step": 255
+    },
+    {
+      "epoch": 0.08547579298831386,
+      "grad_norm": 0.9656402225041684,
+      "learning_rate": 4.273789649415693e-06,
+      "loss": 0.3491,
+      "step": 256
+    },
+    {
+      "epoch": 0.08580968280467446,
+      "grad_norm": 0.9756659114156737,
+      "learning_rate": 4.290484140233723e-06,
+      "loss": 0.3335,
+      "step": 257
+    },
+    {
+      "epoch": 0.08614357262103506,
+      "grad_norm": 0.9490187471908624,
+      "learning_rate": 4.307178631051753e-06,
+      "loss": 0.3188,
+      "step": 258
+    },
+    {
+      "epoch": 0.08647746243739567,
+      "grad_norm": 0.9366799012301222,
+      "learning_rate": 4.323873121869783e-06,
+      "loss": 0.3343,
+      "step": 259
+    },
+    {
+      "epoch": 0.08681135225375626,
+      "grad_norm": 0.9743587902721987,
+      "learning_rate": 4.340567612687813e-06,
+      "loss": 0.339,
+      "step": 260
+    },
+    {
+      "epoch": 0.08714524207011685,
+      "grad_norm": 0.9589212242214878,
+      "learning_rate": 4.357262103505843e-06,
+      "loss": 0.3444,
+      "step": 261
+    },
+    {
+      "epoch": 0.08747913188647746,
+      "grad_norm": 0.9347175795280139,
+      "learning_rate": 4.373956594323873e-06,
+      "loss": 0.3401,
+      "step": 262
+    },
+    {
+      "epoch": 0.08781302170283806,
+      "grad_norm": 0.9417959543736641,
+      "learning_rate": 4.390651085141904e-06,
+      "loss": 0.3411,
+      "step": 263
+    },
+    {
+      "epoch": 0.08814691151919866,
+      "grad_norm": 0.9874440512252058,
+      "learning_rate": 4.407345575959934e-06,
+      "loss": 0.3439,
+      "step": 264
+    },
+    {
+      "epoch": 0.08848080133555926,
+      "grad_norm": 0.8996768541675062,
+      "learning_rate": 4.424040066777964e-06,
+      "loss": 0.3141,
+      "step": 265
+    },
+    {
+      "epoch": 0.08881469115191987,
+      "grad_norm": 1.0011414903014286,
+      "learning_rate": 4.440734557595994e-06,
+      "loss": 0.3499,
+      "step": 266
+    },
+    {
+      "epoch": 0.08914858096828046,
+      "grad_norm": 0.8777138959955474,
+      "learning_rate": 4.457429048414024e-06,
+      "loss": 0.3265,
+      "step": 267
+    },
+    {
+      "epoch": 0.08948247078464107,
+      "grad_norm": 0.9356437897244687,
+      "learning_rate": 4.474123539232054e-06,
+      "loss": 0.3438,
+      "step": 268
+    },
+    {
+      "epoch": 0.08981636060100166,
+      "grad_norm": 1.0229224211478263,
+      "learning_rate": 4.490818030050084e-06,
+      "loss": 0.3536,
+      "step": 269
+    },
+    {
+      "epoch": 0.09015025041736227,
+      "grad_norm": 1.0128053057399193,
+      "learning_rate": 4.507512520868114e-06,
+      "loss": 0.3443,
+      "step": 270
+    },
+    {
+      "epoch": 0.09048414023372287,
+      "grad_norm": 0.9510172987827382,
+      "learning_rate": 4.524207011686144e-06,
+      "loss": 0.3391,
+      "step": 271
+    },
+    {
+      "epoch": 0.09081803005008347,
+      "grad_norm": 0.9158615578958605,
+      "learning_rate": 4.540901502504174e-06,
+      "loss": 0.3512,
+      "step": 272
+    },
+    {
+      "epoch": 0.09115191986644407,
+      "grad_norm": 0.9288795485722017,
+      "learning_rate": 4.557595993322204e-06,
+      "loss": 0.3431,
+      "step": 273
+    },
+    {
+      "epoch": 0.09148580968280468,
+      "grad_norm": 0.9272573551788351,
+      "learning_rate": 4.574290484140234e-06,
+      "loss": 0.3382,
+      "step": 274
+    },
+    {
+      "epoch": 0.09181969949916527,
+      "grad_norm": 0.8710734500522298,
+      "learning_rate": 4.590984974958264e-06,
+      "loss": 0.3437,
+      "step": 275
+    },
+    {
+      "epoch": 0.09215358931552588,
+      "grad_norm": 0.9040830570290983,
+      "learning_rate": 4.607679465776294e-06,
+      "loss": 0.3444,
+      "step": 276
+    },
+    {
+      "epoch": 0.09248747913188647,
+      "grad_norm": 0.8628406600181594,
+      "learning_rate": 4.624373956594324e-06,
+      "loss": 0.3307,
+      "step": 277
+    },
+    {
+      "epoch": 0.09282136894824708,
+      "grad_norm": 0.8878070339574036,
+      "learning_rate": 4.641068447412354e-06,
+      "loss": 0.3385,
+      "step": 278
+    },
+    {
+      "epoch": 0.09315525876460767,
+      "grad_norm": 0.8804933602011825,
+      "learning_rate": 4.657762938230384e-06,
+      "loss": 0.3388,
+      "step": 279
+    },
+    {
+      "epoch": 0.09348914858096828,
+      "grad_norm": 0.8813847897759036,
+      "learning_rate": 4.674457429048414e-06,
+      "loss": 0.3346,
+      "step": 280
+    },
+    {
+      "epoch": 0.09382303839732888,
+      "grad_norm": 0.8586844444765143,
+      "learning_rate": 4.691151919866444e-06,
+      "loss": 0.3265,
+      "step": 281
+    },
+    {
+      "epoch": 0.09415692821368948,
+      "grad_norm": 0.9919861871341583,
+      "learning_rate": 4.707846410684474e-06,
+      "loss": 0.3322,
+      "step": 282
+    },
+    {
+      "epoch": 0.09449081803005008,
+      "grad_norm": 0.8717971953103075,
+      "learning_rate": 4.724540901502504e-06,
+      "loss": 0.334,
+      "step": 283
+    },
+    {
+      "epoch": 0.09482470784641069,
+      "grad_norm": 0.8803852013028539,
+      "learning_rate": 4.741235392320534e-06,
+      "loss": 0.337,
+      "step": 284
+    },
+    {
+      "epoch": 0.09515859766277128,
+      "grad_norm": 0.9095189838984219,
+      "learning_rate": 4.757929883138564e-06,
+      "loss": 0.3392,
+      "step": 285
+    },
+    {
+      "epoch": 0.09549248747913189,
+      "grad_norm": 0.8795834575787537,
+      "learning_rate": 4.774624373956594e-06,
+      "loss": 0.3352,
+      "step": 286
+    },
+    {
+      "epoch": 0.09582637729549248,
+      "grad_norm": 0.8813550524932001,
+      "learning_rate": 4.791318864774624e-06,
+      "loss": 0.3434,
+      "step": 287
+    },
+    {
+      "epoch": 0.09616026711185309,
+      "grad_norm": 0.9274146456820654,
+      "learning_rate": 4.808013355592655e-06,
+      "loss": 0.329,
+      "step": 288
+    },
+    {
+      "epoch": 0.09649415692821368,
+      "grad_norm": 0.9095456432538594,
+      "learning_rate": 4.824707846410685e-06,
+      "loss": 0.3356,
+      "step": 289
+    },
+    {
+      "epoch": 0.09682804674457429,
+      "grad_norm": 0.8370920422361592,
+      "learning_rate": 4.841402337228715e-06,
+      "loss": 0.3318,
+      "step": 290
+    },
+    {
+      "epoch": 0.09716193656093489,
+      "grad_norm": 0.9238323588559383,
+      "learning_rate": 4.858096828046745e-06,
+      "loss": 0.334,
+      "step": 291
+    },
+    {
+      "epoch": 0.0974958263772955,
+      "grad_norm": 0.9314728435890227,
+      "learning_rate": 4.874791318864775e-06,
+      "loss": 0.3422,
+      "step": 292
+    },
+    {
+      "epoch": 0.09782971619365609,
+      "grad_norm": 0.8827040918351866,
+      "learning_rate": 4.891485809682805e-06,
+      "loss": 0.3329,
+      "step": 293
+    },
+    {
+      "epoch": 0.0981636060100167,
+      "grad_norm": 0.8426052454144822,
+      "learning_rate": 4.908180300500835e-06,
+      "loss": 0.3196,
+      "step": 294
+    },
+    {
+      "epoch": 0.09849749582637729,
+      "grad_norm": 0.8544809605886926,
+      "learning_rate": 4.924874791318865e-06,
+      "loss": 0.3378,
+      "step": 295
+    },
+    {
+      "epoch": 0.0988313856427379,
+      "grad_norm": 0.8988047255093645,
+      "learning_rate": 4.941569282136895e-06,
+      "loss": 0.3366,
+      "step": 296
+    },
+    {
+      "epoch": 0.09916527545909849,
+      "grad_norm": 0.9359177125018384,
+      "learning_rate": 4.958263772954925e-06,
+      "loss": 0.3404,
+      "step": 297
+    },
+    {
+      "epoch": 0.0994991652754591,
+      "grad_norm": 0.8194715931707465,
+      "learning_rate": 4.974958263772955e-06,
+      "loss": 0.3339,
+      "step": 298
+    },
+    {
+      "epoch": 0.0998330550918197,
+      "grad_norm": 0.9237740331682973,
+      "learning_rate": 4.991652754590985e-06,
+      "loss": 0.3342,
+      "step": 299
+    },
+    {
+      "epoch": 0.1001669449081803,
+      "grad_norm": 0.8955735733856511,
+      "learning_rate": 5.008347245409015e-06,
+      "loss": 0.3239,
+      "step": 300
+    },
+    {
+      "epoch": 0.1005008347245409,
+      "grad_norm": 0.8374201196142677,
+      "learning_rate": 5.0250417362270454e-06,
+      "loss": 0.3364,
+      "step": 301
+    },
+    {
+      "epoch": 0.1008347245409015,
+      "grad_norm": 0.862899346860263,
+      "learning_rate": 5.0417362270450755e-06,
+      "loss": 0.3175,
+      "step": 302
+    },
+    {
+      "epoch": 0.1011686143572621,
+      "grad_norm": 0.9031415078652251,
+      "learning_rate": 5.0584307178631055e-06,
+      "loss": 0.3326,
+      "step": 303
+    },
+    {
+      "epoch": 0.1015025041736227,
+      "grad_norm": 0.9244500797974233,
+      "learning_rate": 5.0751252086811355e-06,
+      "loss": 0.3461,
+      "step": 304
+    },
+    {
+      "epoch": 0.1018363939899833,
+      "grad_norm": 0.8470037101720705,
+      "learning_rate": 5.0918196994991655e-06,
+      "loss": 0.3423,
+      "step": 305
+    },
+    {
+      "epoch": 0.10217028380634391,
+      "grad_norm": 0.9166133686456392,
+      "learning_rate": 5.1085141903171955e-06,
+      "loss": 0.3492,
+      "step": 306
+    },
+    {
+      "epoch": 0.1025041736227045,
+      "grad_norm": 0.8736064797036586,
+      "learning_rate": 5.1252086811352255e-06,
+      "loss": 0.3373,
+      "step": 307
+    },
+    {
+      "epoch": 0.10283806343906511,
+      "grad_norm": 0.8985765014970434,
+      "learning_rate": 5.141903171953256e-06,
+      "loss": 0.3349,
+      "step": 308
+    },
+    {
+      "epoch": 0.1031719532554257,
+      "grad_norm": 0.8722142872498423,
+      "learning_rate": 5.158597662771286e-06,
+      "loss": 0.3431,
+      "step": 309
+    },
+    {
+      "epoch": 0.10350584307178631,
+      "grad_norm": 0.8501484158698268,
+      "learning_rate": 5.1752921535893164e-06,
+      "loss": 0.3265,
+      "step": 310
+    },
+    {
+      "epoch": 0.10383973288814691,
+      "grad_norm": 0.863314020736203,
+      "learning_rate": 5.1919866444073465e-06,
+      "loss": 0.3106,
+      "step": 311
+    },
+    {
+      "epoch": 0.10417362270450752,
+      "grad_norm": 0.8681570491923405,
+      "learning_rate": 5.2086811352253765e-06,
+      "loss": 0.3314,
+      "step": 312
+    },
+    {
+      "epoch": 0.10450751252086811,
+      "grad_norm": 0.8323052054120158,
+      "learning_rate": 5.2253756260434065e-06,
+      "loss": 0.3212,
+      "step": 313
+    },
+    {
+      "epoch": 0.10484140233722872,
+      "grad_norm": 0.8930784795043445,
+      "learning_rate": 5.2420701168614365e-06,
+      "loss": 0.3397,
+      "step": 314
+    },
+    {
+      "epoch": 0.10517529215358931,
+      "grad_norm": 0.8505993101463951,
+      "learning_rate": 5.2587646076794665e-06,
+      "loss": 0.3308,
+      "step": 315
+    },
+    {
+      "epoch": 0.10550918196994992,
+      "grad_norm": 0.885050504946131,
+      "learning_rate": 5.2754590984974965e-06,
+      "loss": 0.3358,
+      "step": 316
+    },
+    {
+      "epoch": 0.10584307178631051,
+      "grad_norm": 0.8651718208346788,
+      "learning_rate": 5.2921535893155266e-06,
+      "loss": 0.3374,
+      "step": 317
+    },
+    {
+      "epoch": 0.10617696160267112,
+      "grad_norm": 0.9555447901738221,
+      "learning_rate": 5.308848080133557e-06,
+      "loss": 0.3308,
+      "step": 318
+    },
+    {
+      "epoch": 0.10651085141903172,
+      "grad_norm": 0.9138511524531121,
+      "learning_rate": 5.325542570951587e-06,
+      "loss": 0.3297,
+      "step": 319
+    },
+    {
+      "epoch": 0.10684474123539232,
+      "grad_norm": 0.9700328999480945,
+      "learning_rate": 5.342237061769617e-06,
+      "loss": 0.3288,
+      "step": 320
+    },
+    {
+      "epoch": 0.10717863105175292,
+      "grad_norm": 0.9099633252583451,
+      "learning_rate": 5.358931552587646e-06,
+      "loss": 0.3246,
+      "step": 321
+    },
+    {
+      "epoch": 0.10751252086811353,
+      "grad_norm": 0.8086273809052509,
+      "learning_rate": 5.375626043405676e-06,
+      "loss": 0.3187,
+      "step": 322
+    },
+    {
+      "epoch": 0.10784641068447412,
+      "grad_norm": 0.8618148246901657,
+      "learning_rate": 5.392320534223706e-06,
+      "loss": 0.3221,
+      "step": 323
+    },
+    {
+      "epoch": 0.10818030050083473,
+      "grad_norm": 0.885127736316702,
+      "learning_rate": 5.409015025041736e-06,
+      "loss": 0.3277,
+      "step": 324
+    },
+    {
+      "epoch": 0.10851419031719532,
+      "grad_norm": 0.9559705539471627,
+      "learning_rate": 5.4257095158597675e-06,
+      "loss": 0.3293,
+      "step": 325
+    },
+    {
+      "epoch": 0.10884808013355593,
+      "grad_norm": 0.8669142578027386,
+      "learning_rate": 5.4424040066777976e-06,
+      "loss": 0.3245,
+      "step": 326
+    },
+    {
+      "epoch": 0.10918196994991652,
+      "grad_norm": 0.8798483912404614,
+      "learning_rate": 5.459098497495828e-06,
+      "loss": 0.3378,
+      "step": 327
+    },
+    {
+      "epoch": 0.10951585976627713,
+      "grad_norm": 0.9350198857672756,
+      "learning_rate": 5.475792988313858e-06,
+      "loss": 0.3509,
+      "step": 328
+    },
+    {
+      "epoch": 0.10984974958263773,
+      "grad_norm": 0.8849926711815125,
+      "learning_rate": 5.492487479131887e-06,
+      "loss": 0.3324,
+      "step": 329
+    },
+    {
+      "epoch": 0.11018363939899833,
+      "grad_norm": 0.832014509081207,
+      "learning_rate": 5.509181969949917e-06,
+      "loss": 0.3254,
+      "step": 330
+    },
+    {
+      "epoch": 0.11051752921535893,
+      "grad_norm": 0.8479129663932236,
+      "learning_rate": 5.525876460767947e-06,
+      "loss": 0.3256,
+      "step": 331
+    },
+    {
+      "epoch": 0.11085141903171954,
+      "grad_norm": 0.843078683169519,
+      "learning_rate": 5.542570951585977e-06,
+      "loss": 0.3318,
+      "step": 332
+    },
+    {
+      "epoch": 0.11118530884808013,
+      "grad_norm": 0.8165267281915661,
+      "learning_rate": 5.559265442404007e-06,
+      "loss": 0.3171,
+      "step": 333
+    },
+    {
+      "epoch": 0.11151919866444074,
+      "grad_norm": 0.8611047426380982,
+      "learning_rate": 5.575959933222037e-06,
+      "loss": 0.3358,
+      "step": 334
+    },
+    {
+      "epoch": 0.11185308848080133,
+      "grad_norm": 0.8914213947700891,
+      "learning_rate": 5.592654424040067e-06,
+      "loss": 0.334,
+      "step": 335
+    },
+    {
+      "epoch": 0.11218697829716194,
+      "grad_norm": 0.9182326281799634,
+      "learning_rate": 5.609348914858097e-06,
+      "loss": 0.3423,
+      "step": 336
+    },
+    {
+      "epoch": 0.11252086811352253,
+      "grad_norm": 0.8551380975916479,
+      "learning_rate": 5.626043405676127e-06,
+      "loss": 0.3362,
+      "step": 337
+    },
+    {
+      "epoch": 0.11285475792988314,
+      "grad_norm": 0.8839789146451351,
+      "learning_rate": 5.642737896494157e-06,
+      "loss": 0.3372,
+      "step": 338
+    },
+    {
+      "epoch": 0.11318864774624374,
+      "grad_norm": 0.8497586628565239,
+      "learning_rate": 5.659432387312187e-06,
+      "loss": 0.3249,
+      "step": 339
+    },
+    {
+      "epoch": 0.11352253756260434,
+      "grad_norm": 0.9300333060776206,
+      "learning_rate": 5.676126878130217e-06,
+      "loss": 0.332,
+      "step": 340
+    },
+    {
+      "epoch": 0.11385642737896494,
+      "grad_norm": 0.8517294811954167,
+      "learning_rate": 5.692821368948247e-06,
+      "loss": 0.3251,
+      "step": 341
+    },
+    {
+      "epoch": 0.11419031719532555,
+      "grad_norm": 0.8841064373467393,
+      "learning_rate": 5.709515859766278e-06,
+      "loss": 0.335,
+      "step": 342
+    },
+    {
+      "epoch": 0.11452420701168614,
+      "grad_norm": 0.8615988636014796,
+      "learning_rate": 5.726210350584308e-06,
+      "loss": 0.3291,
+      "step": 343
+    },
+    {
+      "epoch": 0.11485809682804675,
+      "grad_norm": 0.8992300184458956,
+      "learning_rate": 5.742904841402338e-06,
+      "loss": 0.3238,
+      "step": 344
+    },
+    {
+      "epoch": 0.11519198664440734,
+      "grad_norm": 0.8091108210502342,
+      "learning_rate": 5.759599332220368e-06,
+      "loss": 0.3313,
+      "step": 345
+    },
+    {
+      "epoch": 0.11552587646076795,
+      "grad_norm": 0.7967637928592147,
+      "learning_rate": 5.776293823038398e-06,
+      "loss": 0.3279,
+      "step": 346
+    },
+    {
+      "epoch": 0.11585976627712855,
+      "grad_norm": 0.8483921636690027,
+      "learning_rate": 5.792988313856428e-06,
+      "loss": 0.3394,
+      "step": 347
+    },
+    {
+      "epoch": 0.11619365609348915,
+      "grad_norm": 0.9336016012794437,
+      "learning_rate": 5.809682804674458e-06,
+      "loss": 0.3307,
+      "step": 348
+    },
+    {
+      "epoch": 0.11652754590984975,
+      "grad_norm": 0.8242300171750219,
+      "learning_rate": 5.826377295492488e-06,
+      "loss": 0.3374,
+      "step": 349
+    },
+    {
+      "epoch": 0.11686143572621036,
+      "grad_norm": 0.8732799099302696,
+      "learning_rate": 5.843071786310518e-06,
+      "loss": 0.34,
+      "step": 350
+    },
+    {
+      "epoch": 0.11719532554257095,
+      "grad_norm": 0.9524028800851025,
+      "learning_rate": 5.859766277128548e-06,
+      "loss": 0.327,
+      "step": 351
+    },
+    {
+      "epoch": 0.11752921535893156,
+      "grad_norm": 0.8203820607817844,
+      "learning_rate": 5.876460767946578e-06,
+      "loss": 0.326,
+      "step": 352
+    },
+    {
+      "epoch": 0.11786310517529215,
+      "grad_norm": 0.8839010145306939,
+      "learning_rate": 5.893155258764608e-06,
+      "loss": 0.3363,
+      "step": 353
+    },
+    {
+      "epoch": 0.11819699499165276,
+      "grad_norm": 0.8400974423876234,
+      "learning_rate": 5.909849749582638e-06,
+      "loss": 0.321,
+      "step": 354
+    },
+    {
+      "epoch": 0.11853088480801335,
+      "grad_norm": 0.8277867238185553,
+      "learning_rate": 5.926544240400668e-06,
+      "loss": 0.3391,
+      "step": 355
+    },
+    {
+      "epoch": 0.11886477462437396,
+      "grad_norm": 0.8669826682474135,
+      "learning_rate": 5.943238731218698e-06,
+      "loss": 0.3416,
+      "step": 356
+    },
+    {
+      "epoch": 0.11919866444073456,
+      "grad_norm": 0.8361315489200994,
+      "learning_rate": 5.959933222036728e-06,
+      "loss": 0.3293,
+      "step": 357
+    },
+    {
+      "epoch": 0.11953255425709516,
+      "grad_norm": 0.8139526386169346,
+      "learning_rate": 5.976627712854759e-06,
+      "loss": 0.3201,
+      "step": 358
+    },
+    {
+      "epoch": 0.11986644407345576,
+      "grad_norm": 0.8156776887159959,
+      "learning_rate": 5.993322203672789e-06,
+      "loss": 0.3251,
+      "step": 359
+    },
+    {
+      "epoch": 0.12020033388981637,
+      "grad_norm": 0.8323024643142797,
+      "learning_rate": 6.010016694490819e-06,
+      "loss": 0.348,
+      "step": 360
+    },
+    {
+      "epoch": 0.12053422370617696,
+      "grad_norm": 0.8101522136526093,
+      "learning_rate": 6.026711185308849e-06,
+      "loss": 0.3309,
+      "step": 361
+    },
+    {
+      "epoch": 0.12086811352253757,
+      "grad_norm": 0.8613296269390823,
+      "learning_rate": 6.043405676126879e-06,
+      "loss": 0.3272,
+      "step": 362
+    },
+    {
+      "epoch": 0.12120200333889816,
+      "grad_norm": 0.8342761548223996,
+      "learning_rate": 6.060100166944909e-06,
+      "loss": 0.3252,
+      "step": 363
+    },
+    {
+      "epoch": 0.12153589315525877,
+      "grad_norm": 0.9603215372730219,
+      "learning_rate": 6.076794657762939e-06,
+      "loss": 0.3344,
+      "step": 364
+    },
+    {
+      "epoch": 0.12186978297161936,
+      "grad_norm": 0.8399472067661772,
+      "learning_rate": 6.093489148580969e-06,
+      "loss": 0.3262,
+      "step": 365
+    },
+    {
+      "epoch": 0.12220367278797997,
+      "grad_norm": 0.9395085715539048,
+      "learning_rate": 6.110183639398999e-06,
+      "loss": 0.3296,
+      "step": 366
+    },
+    {
+      "epoch": 0.12253756260434057,
+      "grad_norm": 0.8478767739933473,
+      "learning_rate": 6.126878130217029e-06,
+      "loss": 0.3313,
+      "step": 367
+    },
+    {
+      "epoch": 0.12287145242070117,
+      "grad_norm": 0.8458582851025415,
+      "learning_rate": 6.143572621035059e-06,
+      "loss": 0.3329,
+      "step": 368
+    },
+    {
+      "epoch": 0.12320534223706177,
+      "grad_norm": 0.847754041553699,
+      "learning_rate": 6.160267111853089e-06,
+      "loss": 0.3213,
+      "step": 369
+    },
+    {
+      "epoch": 0.12353923205342238,
+      "grad_norm": 0.8901021705687011,
+      "learning_rate": 6.176961602671119e-06,
+      "loss": 0.3231,
+      "step": 370
+    },
+    {
+      "epoch": 0.12387312186978297,
+      "grad_norm": 0.894224041517593,
+      "learning_rate": 6.193656093489148e-06,
+      "loss": 0.3277,
+      "step": 371
+    },
+    {
+      "epoch": 0.12420701168614358,
+      "grad_norm": 0.8984980381936205,
+      "learning_rate": 6.210350584307178e-06,
+      "loss": 0.3184,
+      "step": 372
+    },
+    {
+      "epoch": 0.12454090150250417,
+      "grad_norm": 0.8378129983347837,
+      "learning_rate": 6.227045075125208e-06,
+      "loss": 0.3287,
+      "step": 373
+    },
+    {
+      "epoch": 0.12487479131886478,
+      "grad_norm": 0.8040378657152218,
+      "learning_rate": 6.243739565943238e-06,
+      "loss": 0.3078,
+      "step": 374
+    },
+    {
+      "epoch": 0.12520868113522537,
+      "grad_norm": 0.8024014142736975,
+      "learning_rate": 6.26043405676127e-06,
+      "loss": 0.3254,
+      "step": 375
+    },
+    {
+      "epoch": 0.12554257095158597,
+      "grad_norm": 0.7881066130712853,
+      "learning_rate": 6.2771285475793e-06,
+      "loss": 0.321,
+      "step": 376
+    },
+    {
+      "epoch": 0.1258764607679466,
+      "grad_norm": 0.7905211515313185,
+      "learning_rate": 6.29382303839733e-06,
+      "loss": 0.3327,
+      "step": 377
+    },
+    {
+      "epoch": 0.12621035058430718,
+      "grad_norm": 0.7865288324329349,
+      "learning_rate": 6.31051752921536e-06,
+      "loss": 0.3358,
+      "step": 378
+    },
+    {
+      "epoch": 0.12654424040066778,
+      "grad_norm": 0.7992280254051921,
+      "learning_rate": 6.327212020033389e-06,
+      "loss": 0.3272,
+      "step": 379
+    },
+    {
+      "epoch": 0.12687813021702837,
+      "grad_norm": 0.833131684698246,
+      "learning_rate": 6.343906510851419e-06,
+      "loss": 0.3312,
+      "step": 380
+    },
+    {
+      "epoch": 0.127212020033389,
+      "grad_norm": 0.8138838467824974,
+      "learning_rate": 6.360601001669449e-06,
+      "loss": 0.3296,
+      "step": 381
+    },
+    {
+      "epoch": 0.1275459098497496,
+      "grad_norm": 0.8303293815483702,
+      "learning_rate": 6.377295492487479e-06,
+      "loss": 0.319,
+      "step": 382
+    },
+    {
+      "epoch": 0.12787979966611018,
+      "grad_norm": 0.7927602043420368,
+      "learning_rate": 6.393989983305509e-06,
+      "loss": 0.3334,
+      "step": 383
+    },
+    {
+      "epoch": 0.12821368948247078,
+      "grad_norm": 0.8176248891060838,
+      "learning_rate": 6.4106844741235394e-06,
+      "loss": 0.3353,
+      "step": 384
+    },
+    {
+      "epoch": 0.1285475792988314,
+      "grad_norm": 0.830269248700592,
+      "learning_rate": 6.4273789649415694e-06,
+      "loss": 0.3201,
+      "step": 385
+    },
+    {
+      "epoch": 0.128881469115192,
+      "grad_norm": 0.8527742949726782,
+      "learning_rate": 6.4440734557595995e-06,
+      "loss": 0.3426,
+      "step": 386
+    },
+    {
+      "epoch": 0.1292153589315526,
+      "grad_norm": 0.7974703676139379,
+      "learning_rate": 6.4607679465776295e-06,
+      "loss": 0.3248,
+      "step": 387
+    },
+    {
+      "epoch": 0.12954924874791318,
+      "grad_norm": 0.7631270830728287,
+      "learning_rate": 6.4774624373956595e-06,
+      "loss": 0.3299,
+      "step": 388
+    },
+    {
+      "epoch": 0.1298831385642738,
+      "grad_norm": 0.8188752212842851,
+      "learning_rate": 6.4941569282136895e-06,
+      "loss": 0.3224,
+      "step": 389
+    },
+    {
+      "epoch": 0.1302170283806344,
+      "grad_norm": 0.8159275831991325,
+      "learning_rate": 6.5108514190317195e-06,
+      "loss": 0.3119,
+      "step": 390
+    },
+    {
+      "epoch": 0.130550918196995,
+      "grad_norm": 0.8039999142528687,
+      "learning_rate": 6.5275459098497496e-06,
+      "loss": 0.3304,
+      "step": 391
+    },
+    {
+      "epoch": 0.13088480801335559,
+      "grad_norm": 0.793285514729878,
+      "learning_rate": 6.54424040066778e-06,
+      "loss": 0.3322,
+      "step": 392
+    },
+    {
+      "epoch": 0.13121869782971618,
+      "grad_norm": 0.7827475194852668,
+      "learning_rate": 6.5609348914858104e-06,
+      "loss": 0.3218,
+      "step": 393
+    },
+    {
+      "epoch": 0.1315525876460768,
+      "grad_norm": 0.7929845712903717,
+      "learning_rate": 6.5776293823038405e-06,
+      "loss": 0.3362,
+      "step": 394
+    },
+    {
+      "epoch": 0.1318864774624374,
+      "grad_norm": 0.8461054866909821,
+      "learning_rate": 6.5943238731218705e-06,
+      "loss": 0.3304,
+      "step": 395
+    },
+    {
+      "epoch": 0.132220367278798,
+      "grad_norm": 0.793836493927258,
+      "learning_rate": 6.6110183639399005e-06,
+      "loss": 0.3149,
+      "step": 396
+    },
+    {
+      "epoch": 0.13255425709515858,
+      "grad_norm": 0.8223673258072549,
+      "learning_rate": 6.6277128547579305e-06,
+      "loss": 0.3246,
+      "step": 397
+    },
+    {
+      "epoch": 0.1328881469115192,
+      "grad_norm": 0.8027475419925476,
+      "learning_rate": 6.6444073455759605e-06,
+      "loss": 0.3309,
+      "step": 398
+    },
+    {
+      "epoch": 0.1332220367278798,
+      "grad_norm": 0.8495635675039467,
+      "learning_rate": 6.6611018363939905e-06,
+      "loss": 0.3289,
+      "step": 399
+    },
+    {
+      "epoch": 0.1335559265442404,
+      "grad_norm": 0.7849595707347243,
+      "learning_rate": 6.6777963272120206e-06,
+      "loss": 0.3166,
+      "step": 400
+    },
+    {
+      "epoch": 0.133889816360601,
+      "grad_norm": 0.8170059148464306,
+      "learning_rate": 6.6944908180300506e-06,
+      "loss": 0.333,
+      "step": 401
+    },
+    {
+      "epoch": 0.1342237061769616,
+      "grad_norm": 0.8000657865700851,
+      "learning_rate": 6.711185308848081e-06,
+      "loss": 0.333,
+      "step": 402
+    },
+    {
+      "epoch": 0.1345575959933222,
+      "grad_norm": 0.8015462894092409,
+      "learning_rate": 6.727879799666111e-06,
+      "loss": 0.3313,
+      "step": 403
+    },
+    {
+      "epoch": 0.1348914858096828,
+      "grad_norm": 0.7835473923129718,
+      "learning_rate": 6.744574290484141e-06,
+      "loss": 0.3139,
+      "step": 404
+    },
+    {
+      "epoch": 0.1352253756260434,
+      "grad_norm": 0.8289066768646055,
+      "learning_rate": 6.761268781302171e-06,
+      "loss": 0.3352,
+      "step": 405
+    },
+    {
+      "epoch": 0.13555926544240401,
+      "grad_norm": 0.7803857298710676,
+      "learning_rate": 6.777963272120201e-06,
+      "loss": 0.3179,
+      "step": 406
+    },
+    {
+      "epoch": 0.1358931552587646,
+      "grad_norm": 0.8236453121891996,
+      "learning_rate": 6.794657762938231e-06,
+      "loss": 0.3324,
+      "step": 407
+    },
+    {
+      "epoch": 0.1362270450751252,
+      "grad_norm": 0.8454808753809345,
+      "learning_rate": 6.8113522537562615e-06,
+      "loss": 0.328,
+      "step": 408
+    },
+    {
+      "epoch": 0.1365609348914858,
+      "grad_norm": 0.9127121927054462,
+      "learning_rate": 6.8280467445742916e-06,
+      "loss": 0.3358,
+      "step": 409
+    },
+    {
+      "epoch": 0.13689482470784642,
+      "grad_norm": 0.7846041649291898,
+      "learning_rate": 6.844741235392322e-06,
+      "loss": 0.3233,
+      "step": 410
+    },
+    {
+      "epoch": 0.137228714524207,
+      "grad_norm": 0.9201235291106307,
+      "learning_rate": 6.861435726210352e-06,
+      "loss": 0.321,
+      "step": 411
+    },
+    {
+      "epoch": 0.1375626043405676,
+      "grad_norm": 0.8033872399797178,
+      "learning_rate": 6.878130217028382e-06,
+      "loss": 0.316,
+      "step": 412
+    },
+    {
+      "epoch": 0.1378964941569282,
+      "grad_norm": 0.8345684715033209,
+      "learning_rate": 6.894824707846412e-06,
+      "loss": 0.3175,
+      "step": 413
+    },
+    {
+      "epoch": 0.13823038397328882,
+      "grad_norm": 0.9529747232536198,
+      "learning_rate": 6.911519198664442e-06,
+      "loss": 0.3155,
+      "step": 414
+    },
+    {
+      "epoch": 0.13856427378964942,
+      "grad_norm": 0.7734469728122102,
+      "learning_rate": 6.928213689482472e-06,
+      "loss": 0.3194,
+      "step": 415
+    },
+    {
+      "epoch": 0.13889816360601,
+      "grad_norm": 0.8126513847176664,
+      "learning_rate": 6.944908180300502e-06,
+      "loss": 0.3276,
+      "step": 416
+    },
+    {
+      "epoch": 0.1392320534223706,
+      "grad_norm": 0.8905634812889437,
+      "learning_rate": 6.961602671118532e-06,
+      "loss": 0.3317,
+      "step": 417
+    },
+    {
+      "epoch": 0.13956594323873123,
+      "grad_norm": 0.9074523021240545,
+      "learning_rate": 6.978297161936562e-06,
+      "loss": 0.3179,
+      "step": 418
+    },
+    {
+      "epoch": 0.13989983305509182,
+      "grad_norm": 0.784057088957602,
+      "learning_rate": 6.994991652754592e-06,
+      "loss": 0.3154,
+      "step": 419
+    },
+    {
+      "epoch": 0.14023372287145242,
+      "grad_norm": 0.8423356411690469,
+      "learning_rate": 7.011686143572622e-06,
+      "loss": 0.3218,
+      "step": 420
+    },
+    {
+      "epoch": 0.140567612687813,
+      "grad_norm": 0.7882599075449199,
+      "learning_rate": 7.028380634390651e-06,
+      "loss": 0.3133,
+      "step": 421
+    },
+    {
+      "epoch": 0.14090150250417363,
+      "grad_norm": 0.8942925936919841,
+      "learning_rate": 7.045075125208681e-06,
+      "loss": 0.3264,
+      "step": 422
+    },
+    {
+      "epoch": 0.14123539232053423,
+      "grad_norm": 0.79770279744943,
+      "learning_rate": 7.061769616026711e-06,
+      "loss": 0.32,
+      "step": 423
+    },
+    {
+      "epoch": 0.14156928213689482,
+      "grad_norm": 0.8039261032935318,
+      "learning_rate": 7.078464106844741e-06,
+      "loss": 0.3278,
+      "step": 424
+    },
+    {
+      "epoch": 0.1419031719532554,
+      "grad_norm": 0.9005342314717675,
+      "learning_rate": 7.095158597662773e-06,
+      "loss": 0.3165,
+      "step": 425
+    },
+    {
+      "epoch": 0.14223706176961604,
+      "grad_norm": 0.8227540600826516,
+      "learning_rate": 7.111853088480803e-06,
+      "loss": 0.3321,
+      "step": 426
+    },
+    {
+      "epoch": 0.14257095158597663,
+      "grad_norm": 0.7793771980143581,
+      "learning_rate": 7.128547579298833e-06,
+      "loss": 0.3251,
+      "step": 427
+    },
+    {
+      "epoch": 0.14290484140233722,
+      "grad_norm": 0.8787617175332001,
+      "learning_rate": 7.145242070116863e-06,
+      "loss": 0.3369,
+      "step": 428
+    },
+    {
+      "epoch": 0.14323873121869782,
+      "grad_norm": 0.7856322905582587,
+      "learning_rate": 7.161936560934892e-06,
+      "loss": 0.3206,
+      "step": 429
+    },
+    {
+      "epoch": 0.14357262103505844,
+      "grad_norm": 0.8008132156366593,
+      "learning_rate": 7.178631051752922e-06,
+      "loss": 0.3178,
+      "step": 430
+    },
+    {
+      "epoch": 0.14390651085141903,
+      "grad_norm": 0.7921460163083631,
+      "learning_rate": 7.195325542570952e-06,
+      "loss": 0.326,
+      "step": 431
+    },
+    {
+      "epoch": 0.14424040066777963,
+      "grad_norm": 0.7943930677123332,
+      "learning_rate": 7.212020033388982e-06,
+      "loss": 0.3137,
+      "step": 432
+    },
+    {
+      "epoch": 0.14457429048414022,
+      "grad_norm": 0.8867565416973608,
+      "learning_rate": 7.228714524207012e-06,
+      "loss": 0.3323,
+      "step": 433
+    },
+    {
+      "epoch": 0.14490818030050084,
+      "grad_norm": 0.8481774194851605,
+      "learning_rate": 7.245409015025042e-06,
+      "loss": 0.3214,
+      "step": 434
+    },
+    {
+      "epoch": 0.14524207011686144,
+      "grad_norm": 0.8152815606649613,
+      "learning_rate": 7.262103505843072e-06,
+      "loss": 0.3234,
+      "step": 435
+    },
+    {
+      "epoch": 0.14557595993322203,
+      "grad_norm": 0.8485079375718337,
+      "learning_rate": 7.278797996661102e-06,
+      "loss": 0.3312,
+      "step": 436
+    },
+    {
+      "epoch": 0.14590984974958263,
+      "grad_norm": 0.8316215407015559,
+      "learning_rate": 7.295492487479132e-06,
+      "loss": 0.3409,
+      "step": 437
+    },
+    {
+      "epoch": 0.14624373956594325,
+      "grad_norm": 0.7866131242096099,
+      "learning_rate": 7.312186978297162e-06,
+      "loss": 0.334,
+      "step": 438
+    },
+    {
+      "epoch": 0.14657762938230384,
+      "grad_norm": 0.8374183015975639,
+      "learning_rate": 7.328881469115192e-06,
+      "loss": 0.3155,
+      "step": 439
+    },
+    {
+      "epoch": 0.14691151919866444,
+      "grad_norm": 0.8965252500439357,
+      "learning_rate": 7.345575959933222e-06,
+      "loss": 0.3291,
+      "step": 440
+    },
+    {
+      "epoch": 0.14724540901502503,
+      "grad_norm": 0.7877070534314122,
+      "learning_rate": 7.362270450751253e-06,
+      "loss": 0.3338,
+      "step": 441
+    },
+    {
+      "epoch": 0.14757929883138565,
+      "grad_norm": 0.7693836245477367,
+      "learning_rate": 7.378964941569283e-06,
+      "loss": 0.3187,
+      "step": 442
+    },
+    {
+      "epoch": 0.14791318864774625,
+      "grad_norm": 0.7811194616198989,
+      "learning_rate": 7.395659432387313e-06,
+      "loss": 0.3218,
+      "step": 443
+    },
+    {
+      "epoch": 0.14824707846410684,
+      "grad_norm": 0.8067196421370193,
+      "learning_rate": 7.412353923205343e-06,
+      "loss": 0.3191,
+      "step": 444
+    },
+    {
+      "epoch": 0.14858096828046743,
+      "grad_norm": 0.745007993584021,
+      "learning_rate": 7.429048414023373e-06,
+      "loss": 0.3178,
+      "step": 445
+    },
+    {
+      "epoch": 0.14891485809682806,
+      "grad_norm": 0.7860736411083058,
+      "learning_rate": 7.445742904841403e-06,
+      "loss": 0.3265,
+      "step": 446
+    },
+    {
+      "epoch": 0.14924874791318865,
+      "grad_norm": 0.7730455247915992,
+      "learning_rate": 7.462437395659433e-06,
+      "loss": 0.3249,
+      "step": 447
+    },
+    {
+      "epoch": 0.14958263772954924,
+      "grad_norm": 0.8138521974216645,
+      "learning_rate": 7.479131886477463e-06,
+      "loss": 0.325,
+      "step": 448
+    },
+    {
+      "epoch": 0.14991652754590984,
+      "grad_norm": 0.8661456040993474,
+      "learning_rate": 7.495826377295493e-06,
+      "loss": 0.3298,
+      "step": 449
+    },
+    {
+      "epoch": 0.15025041736227046,
+      "grad_norm": 0.8051739894421988,
+      "learning_rate": 7.512520868113523e-06,
+      "loss": 0.3414,
+      "step": 450
+    },
+    {
+      "epoch": 0.15058430717863105,
+      "grad_norm": 0.7360449030147277,
+      "learning_rate": 7.529215358931553e-06,
+      "loss": 0.326,
+      "step": 451
+    },
+    {
+      "epoch": 0.15091819699499165,
+      "grad_norm": 0.7753369666450849,
+      "learning_rate": 7.545909849749583e-06,
+      "loss": 0.3284,
+      "step": 452
+    },
+    {
+      "epoch": 0.15125208681135224,
+      "grad_norm": 0.8034462011906657,
+      "learning_rate": 7.562604340567613e-06,
+      "loss": 0.3172,
+      "step": 453
+    },
+    {
+      "epoch": 0.15158597662771287,
+      "grad_norm": 0.816594979135993,
+      "learning_rate": 7.579298831385643e-06,
+      "loss": 0.3276,
+      "step": 454
+    },
+    {
+      "epoch": 0.15191986644407346,
+      "grad_norm": 0.7559299456203845,
+      "learning_rate": 7.595993322203673e-06,
+      "loss": 0.3155,
+      "step": 455
+    },
+    {
+      "epoch": 0.15225375626043405,
+      "grad_norm": 0.8500287197736555,
+      "learning_rate": 7.612687813021703e-06,
+      "loss": 0.3197,
+      "step": 456
+    },
+    {
+      "epoch": 0.15258764607679465,
+      "grad_norm": 0.7969014837631675,
+      "learning_rate": 7.629382303839733e-06,
+      "loss": 0.3261,
+      "step": 457
+    },
+    {
+      "epoch": 0.15292153589315527,
+      "grad_norm": 0.768264708478163,
+      "learning_rate": 7.646076794657763e-06,
+      "loss": 0.3242,
+      "step": 458
+    },
+    {
+      "epoch": 0.15325542570951586,
+      "grad_norm": 0.8069410984531731,
+      "learning_rate": 7.662771285475793e-06,
+      "loss": 0.3288,
+      "step": 459
+    },
+    {
+      "epoch": 0.15358931552587646,
+      "grad_norm": 0.7920850646453554,
+      "learning_rate": 7.679465776293823e-06,
+      "loss": 0.3371,
+      "step": 460
+    },
+    {
+      "epoch": 0.15392320534223705,
+      "grad_norm": 0.7785000821700567,
+      "learning_rate": 7.696160267111853e-06,
+      "loss": 0.3033,
+      "step": 461
+    },
+    {
+      "epoch": 0.15425709515859767,
+      "grad_norm": 0.7735052227404169,
+      "learning_rate": 7.712854757929883e-06,
+      "loss": 0.3223,
+      "step": 462
+    },
+    {
+      "epoch": 0.15459098497495827,
+      "grad_norm": 0.7739240492974162,
+      "learning_rate": 7.729549248747913e-06,
+      "loss": 0.3184,
+      "step": 463
+    },
+    {
+      "epoch": 0.15492487479131886,
+      "grad_norm": 0.7673134772151452,
+      "learning_rate": 7.746243739565943e-06,
+      "loss": 0.3313,
+      "step": 464
+    },
+    {
+      "epoch": 0.15525876460767946,
+      "grad_norm": 0.8003298238897926,
+      "learning_rate": 7.762938230383973e-06,
+      "loss": 0.3203,
+      "step": 465
+    },
+    {
+      "epoch": 0.15559265442404008,
+      "grad_norm": 0.8069234085629442,
+      "learning_rate": 7.779632721202003e-06,
+      "loss": 0.3284,
+      "step": 466
+    },
+    {
+      "epoch": 0.15592654424040067,
+      "grad_norm": 0.7382892924896265,
+      "learning_rate": 7.796327212020033e-06,
+      "loss": 0.3132,
+      "step": 467
+    },
+    {
+      "epoch": 0.15626043405676127,
+      "grad_norm": 0.810581367586647,
+      "learning_rate": 7.813021702838063e-06,
+      "loss": 0.3204,
+      "step": 468
+    },
+    {
+      "epoch": 0.15659432387312186,
+      "grad_norm": 0.7966533826357051,
+      "learning_rate": 7.829716193656093e-06,
+      "loss": 0.3115,
+      "step": 469
+    },
+    {
+      "epoch": 0.15692821368948248,
+      "grad_norm": 0.8269845067196703,
+      "learning_rate": 7.846410684474123e-06,
+      "loss": 0.3157,
+      "step": 470
+    },
+    {
+      "epoch": 0.15726210350584308,
+      "grad_norm": 0.7050096437435556,
+      "learning_rate": 7.863105175292153e-06,
+      "loss": 0.3131,
+      "step": 471
+    },
+    {
+      "epoch": 0.15759599332220367,
+      "grad_norm": 0.9129246065051041,
+      "learning_rate": 7.879799666110184e-06,
+      "loss": 0.3248,
+      "step": 472
+    },
+    {
+      "epoch": 0.15792988313856426,
+      "grad_norm": 0.7645190322893277,
+      "learning_rate": 7.896494156928214e-06,
+      "loss": 0.3153,
+      "step": 473
+    },
+    {
+      "epoch": 0.1582637729549249,
+      "grad_norm": 0.8088851332376162,
+      "learning_rate": 7.913188647746244e-06,
+      "loss": 0.3299,
+      "step": 474
+    },
+    {
+      "epoch": 0.15859766277128548,
+      "grad_norm": 0.8328128755199847,
+      "learning_rate": 7.929883138564275e-06,
+      "loss": 0.3321,
+      "step": 475
+    },
+    {
+      "epoch": 0.15893155258764607,
+      "grad_norm": 0.8103019168538868,
+      "learning_rate": 7.946577629382305e-06,
+      "loss": 0.3358,
+      "step": 476
+    },
+    {
+      "epoch": 0.15926544240400667,
+      "grad_norm": 0.7310047276567335,
+      "learning_rate": 7.963272120200335e-06,
+      "loss": 0.3287,
+      "step": 477
+    },
+    {
+      "epoch": 0.1595993322203673,
+      "grad_norm": 0.7583400690765941,
+      "learning_rate": 7.979966611018365e-06,
+      "loss": 0.3209,
+      "step": 478
+    },
+    {
+      "epoch": 0.15993322203672788,
+      "grad_norm": 0.8275382387475719,
+      "learning_rate": 7.996661101836395e-06,
+      "loss": 0.3226,
+      "step": 479
+    },
+    {
+      "epoch": 0.16026711185308848,
+      "grad_norm": 0.8324627172393388,
+      "learning_rate": 8.013355592654425e-06,
+      "loss": 0.3162,
+      "step": 480
+    },
+    {
+      "epoch": 0.16060100166944907,
+      "grad_norm": 0.7658134787985736,
+      "learning_rate": 8.030050083472455e-06,
+      "loss": 0.3217,
+      "step": 481
+    },
+    {
+      "epoch": 0.1609348914858097,
+      "grad_norm": 0.8334447051987756,
+      "learning_rate": 8.046744574290485e-06,
+      "loss": 0.3105,
+      "step": 482
+    },
+    {
+      "epoch": 0.1612687813021703,
+      "grad_norm": 0.7640159041956591,
+      "learning_rate": 8.063439065108515e-06,
+      "loss": 0.3237,
+      "step": 483
+    },
+    {
+      "epoch": 0.16160267111853088,
+      "grad_norm": 0.7935719290550487,
+      "learning_rate": 8.080133555926545e-06,
+      "loss": 0.3183,
+      "step": 484
+    },
+    {
+      "epoch": 0.16193656093489148,
+      "grad_norm": 0.8042613292069457,
+      "learning_rate": 8.096828046744575e-06,
+      "loss": 0.3088,
+      "step": 485
+    },
+    {
+      "epoch": 0.1622704507512521,
+      "grad_norm": 0.7298193294098594,
+      "learning_rate": 8.113522537562605e-06,
+      "loss": 0.3289,
+      "step": 486
+    },
+    {
+      "epoch": 0.1626043405676127,
+      "grad_norm": 0.7760296007604826,
+      "learning_rate": 8.130217028380635e-06,
+      "loss": 0.3392,
+      "step": 487
+    },
+    {
+      "epoch": 0.1629382303839733,
+      "grad_norm": 0.7998813780637646,
+      "learning_rate": 8.146911519198665e-06,
+      "loss": 0.3216,
+      "step": 488
+    },
+    {
+      "epoch": 0.16327212020033388,
+      "grad_norm": 0.754659475225854,
+      "learning_rate": 8.163606010016696e-06,
+      "loss": 0.3283,
+      "step": 489
+    },
+    {
+      "epoch": 0.1636060100166945,
+      "grad_norm": 0.828770460621221,
+      "learning_rate": 8.180300500834726e-06,
+      "loss": 0.3297,
+      "step": 490
+    },
+    {
+      "epoch": 0.1639398998330551,
+      "grad_norm": 0.8721226587333579,
+      "learning_rate": 8.196994991652756e-06,
+      "loss": 0.3243,
+      "step": 491
+    },
+    {
+      "epoch": 0.1642737896494157,
+      "grad_norm": 0.7943843715584529,
+      "learning_rate": 8.213689482470786e-06,
+      "loss": 0.3182,
+      "step": 492
+    },
+    {
+      "epoch": 0.16460767946577629,
+      "grad_norm": 0.7372749126835477,
+      "learning_rate": 8.230383973288816e-06,
+      "loss": 0.3141,
+      "step": 493
+    },
+    {
+      "epoch": 0.1649415692821369,
+      "grad_norm": 0.7431419416002658,
+      "learning_rate": 8.247078464106846e-06,
+      "loss": 0.3142,
+      "step": 494
+    },
+    {
+      "epoch": 0.1652754590984975,
+      "grad_norm": 0.7942554696379105,
+      "learning_rate": 8.263772954924876e-06,
+      "loss": 0.3237,
+      "step": 495
+    },
+    {
+      "epoch": 0.1656093489148581,
+      "grad_norm": 0.7834064114003052,
+      "learning_rate": 8.280467445742906e-06,
+      "loss": 0.3155,
+      "step": 496
+    },
+    {
+      "epoch": 0.1659432387312187,
+      "grad_norm": 0.8211077834088286,
+      "learning_rate": 8.297161936560936e-06,
+      "loss": 0.3227,
+      "step": 497
+    },
+    {
+      "epoch": 0.1662771285475793,
+      "grad_norm": 0.7303338802378014,
+      "learning_rate": 8.313856427378966e-06,
+      "loss": 0.3161,
+      "step": 498
+    },
+    {
+      "epoch": 0.1666110183639399,
+      "grad_norm": 0.8150909854544564,
+      "learning_rate": 8.330550918196996e-06,
+      "loss": 0.3391,
+      "step": 499
+    },
+    {
+      "epoch": 0.1669449081803005,
+      "grad_norm": 0.8268594090322386,
+      "learning_rate": 8.347245409015026e-06,
+      "loss": 0.3225,
+      "step": 500
+    },
+    {
+      "epoch": 0.1672787979966611,
+      "grad_norm": 0.709422401662154,
+      "learning_rate": 8.363939899833056e-06,
+      "loss": 0.3235,
+      "step": 501
+    },
+    {
+      "epoch": 0.16761268781302172,
+      "grad_norm": 0.7810288298274758,
+      "learning_rate": 8.380634390651086e-06,
+      "loss": 0.3297,
+      "step": 502
+    },
+    {
+      "epoch": 0.1679465776293823,
+      "grad_norm": 0.8113004242138308,
+      "learning_rate": 8.397328881469116e-06,
+      "loss": 0.3336,
+      "step": 503
+    },
+    {
+      "epoch": 0.1682804674457429,
+      "grad_norm": 0.7130387818490599,
+      "learning_rate": 8.414023372287146e-06,
+      "loss": 0.3232,
+      "step": 504
+    },
+    {
+      "epoch": 0.1686143572621035,
+      "grad_norm": 0.763332933237767,
+      "learning_rate": 8.430717863105176e-06,
+      "loss": 0.3207,
+      "step": 505
+    },
+    {
+      "epoch": 0.16894824707846412,
+      "grad_norm": 0.7292264834092494,
+      "learning_rate": 8.447412353923206e-06,
+      "loss": 0.3288,
+      "step": 506
+    },
+    {
+      "epoch": 0.16928213689482471,
+      "grad_norm": 0.7215523065461271,
+      "learning_rate": 8.464106844741236e-06,
+      "loss": 0.3376,
+      "step": 507
+    },
+    {
+      "epoch": 0.1696160267111853,
+      "grad_norm": 0.7443791631173745,
+      "learning_rate": 8.480801335559266e-06,
+      "loss": 0.3173,
+      "step": 508
+    },
+    {
+      "epoch": 0.1699499165275459,
+      "grad_norm": 0.7533075498019624,
+      "learning_rate": 8.497495826377296e-06,
+      "loss": 0.33,
+      "step": 509
+    },
+    {
+      "epoch": 0.17028380634390652,
+      "grad_norm": 0.8091104652472866,
+      "learning_rate": 8.514190317195326e-06,
+      "loss": 0.3253,
+      "step": 510
+    },
+    {
+      "epoch": 0.17061769616026712,
+      "grad_norm": 0.7705793740777093,
+      "learning_rate": 8.530884808013356e-06,
+      "loss": 0.3226,
+      "step": 511
+    },
+    {
+      "epoch": 0.1709515859766277,
+      "grad_norm": 0.7523663640248952,
+      "learning_rate": 8.547579298831386e-06,
+      "loss": 0.3253,
+      "step": 512
+    },
+    {
+      "epoch": 0.1712854757929883,
+      "grad_norm": 0.8318150540485132,
+      "learning_rate": 8.564273789649416e-06,
+      "loss": 0.3226,
+      "step": 513
+    },
+    {
+      "epoch": 0.17161936560934893,
+      "grad_norm": 0.8551152867595071,
+      "learning_rate": 8.580968280467446e-06,
+      "loss": 0.3214,
+      "step": 514
+    },
+    {
+      "epoch": 0.17195325542570952,
+      "grad_norm": 0.6992123748676768,
+      "learning_rate": 8.597662771285476e-06,
+      "loss": 0.3302,
+      "step": 515
+    },
+    {
+      "epoch": 0.17228714524207012,
+      "grad_norm": 0.8109200213995847,
+      "learning_rate": 8.614357262103506e-06,
+      "loss": 0.3183,
+      "step": 516
+    },
+    {
+      "epoch": 0.1726210350584307,
+      "grad_norm": 0.7154099679151257,
+      "learning_rate": 8.631051752921536e-06,
+      "loss": 0.325,
+      "step": 517
+    },
+    {
+      "epoch": 0.17295492487479133,
+      "grad_norm": 0.7042520936376871,
+      "learning_rate": 8.647746243739566e-06,
+      "loss": 0.3258,
+      "step": 518
+    },
+    {
+      "epoch": 0.17328881469115193,
+      "grad_norm": 0.8356673455228761,
+      "learning_rate": 8.664440734557596e-06,
+      "loss": 0.3398,
+      "step": 519
+    },
+    {
+      "epoch": 0.17362270450751252,
+      "grad_norm": 0.7424543791374282,
+      "learning_rate": 8.681135225375626e-06,
+      "loss": 0.3318,
+      "step": 520
+    },
+    {
+      "epoch": 0.17395659432387311,
+      "grad_norm": 0.7296263140020599,
+      "learning_rate": 8.697829716193656e-06,
+      "loss": 0.3308,
+      "step": 521
+    },
+    {
+      "epoch": 0.1742904841402337,
+      "grad_norm": 0.7792422670957555,
+      "learning_rate": 8.714524207011686e-06,
+      "loss": 0.3134,
+      "step": 522
+    },
+    {
+      "epoch": 0.17462437395659433,
+      "grad_norm": 0.7446025853270303,
+      "learning_rate": 8.731218697829716e-06,
+      "loss": 0.3117,
+      "step": 523
+    },
+    {
+      "epoch": 0.17495826377295493,
+      "grad_norm": 0.796990104914787,
+      "learning_rate": 8.747913188647746e-06,
+      "loss": 0.3262,
+      "step": 524
+    },
+    {
+      "epoch": 0.17529215358931552,
+      "grad_norm": 0.7834536124966097,
+      "learning_rate": 8.764607679465778e-06,
+      "loss": 0.3199,
+      "step": 525
+    },
+    {
+      "epoch": 0.1756260434056761,
+      "grad_norm": 0.7678132233195324,
+      "learning_rate": 8.781302170283808e-06,
+      "loss": 0.3332,
+      "step": 526
+    },
+    {
+      "epoch": 0.17595993322203674,
+      "grad_norm": 0.7767830958785583,
+      "learning_rate": 8.797996661101838e-06,
+      "loss": 0.3141,
+      "step": 527
+    },
+    {
+      "epoch": 0.17629382303839733,
+      "grad_norm": 0.7852387118311777,
+      "learning_rate": 8.814691151919868e-06,
+      "loss": 0.3315,
+      "step": 528
+    },
+    {
+      "epoch": 0.17662771285475792,
+      "grad_norm": 0.7255280477671688,
+      "learning_rate": 8.831385642737898e-06,
+      "loss": 0.3221,
+      "step": 529
+    },
+    {
+      "epoch": 0.17696160267111852,
+      "grad_norm": 0.7482161691510137,
+      "learning_rate": 8.848080133555928e-06,
+      "loss": 0.3194,
+      "step": 530
+    },
+    {
+      "epoch": 0.17729549248747914,
+      "grad_norm": 0.7495820981919216,
+      "learning_rate": 8.864774624373958e-06,
+      "loss": 0.3203,
+      "step": 531
+    },
+    {
+      "epoch": 0.17762938230383973,
+      "grad_norm": 0.7427985679840543,
+      "learning_rate": 8.881469115191988e-06,
+      "loss": 0.318,
+      "step": 532
+    },
+    {
+      "epoch": 0.17796327212020033,
+      "grad_norm": 0.7192721123905805,
+      "learning_rate": 8.898163606010018e-06,
+      "loss": 0.3186,
+      "step": 533
+    },
+    {
+      "epoch": 0.17829716193656092,
+      "grad_norm": 0.7863163084541621,
+      "learning_rate": 8.914858096828048e-06,
+      "loss": 0.3217,
+      "step": 534
+    },
+    {
+      "epoch": 0.17863105175292154,
+      "grad_norm": 0.7441736096907753,
+      "learning_rate": 8.931552587646078e-06,
+      "loss": 0.3135,
+      "step": 535
+    },
+    {
+      "epoch": 0.17896494156928214,
+      "grad_norm": 0.7600406189062465,
+      "learning_rate": 8.948247078464108e-06,
+      "loss": 0.3074,
+      "step": 536
+    },
+    {
+      "epoch": 0.17929883138564273,
+      "grad_norm": 0.7225882836034713,
+      "learning_rate": 8.964941569282138e-06,
+      "loss": 0.3103,
+      "step": 537
+    },
+    {
+      "epoch": 0.17963272120200333,
+      "grad_norm": 0.8154269463833613,
+      "learning_rate": 8.981636060100168e-06,
+      "loss": 0.3449,
+      "step": 538
+    },
+    {
+      "epoch": 0.17996661101836395,
+      "grad_norm": 0.8303496759620166,
+      "learning_rate": 8.998330550918198e-06,
+      "loss": 0.3377,
+      "step": 539
+    },
+    {
+      "epoch": 0.18030050083472454,
+      "grad_norm": 0.717322451758441,
+      "learning_rate": 9.015025041736228e-06,
+      "loss": 0.3102,
+      "step": 540
+    },
+    {
+      "epoch": 0.18063439065108514,
+      "grad_norm": 0.780512877554671,
+      "learning_rate": 9.031719532554258e-06,
+      "loss": 0.3078,
+      "step": 541
+    },
+    {
+      "epoch": 0.18096828046744573,
+      "grad_norm": 0.7769371870889908,
+      "learning_rate": 9.048414023372288e-06,
+      "loss": 0.3218,
+      "step": 542
+    },
+    {
+      "epoch": 0.18130217028380635,
+      "grad_norm": 0.699570978770357,
+      "learning_rate": 9.065108514190318e-06,
+      "loss": 0.3046,
+      "step": 543
+    },
+    {
+      "epoch": 0.18163606010016695,
+      "grad_norm": 0.7587635230325729,
+      "learning_rate": 9.081803005008348e-06,
+      "loss": 0.3226,
+      "step": 544
+    },
+    {
+      "epoch": 0.18196994991652754,
+      "grad_norm": 0.7576445186041849,
+      "learning_rate": 9.098497495826378e-06,
+      "loss": 0.3137,
+      "step": 545
+    },
+    {
+      "epoch": 0.18230383973288813,
+      "grad_norm": 0.7239651984797366,
+      "learning_rate": 9.115191986644408e-06,
+      "loss": 0.3096,
+      "step": 546
+    },
+    {
+      "epoch": 0.18263772954924876,
+      "grad_norm": 0.8109770776440577,
+      "learning_rate": 9.131886477462438e-06,
+      "loss": 0.3363,
+      "step": 547
+    },
+    {
+      "epoch": 0.18297161936560935,
+      "grad_norm": 0.7160237652592348,
+      "learning_rate": 9.148580968280468e-06,
+      "loss": 0.321,
+      "step": 548
+    },
+    {
+      "epoch": 0.18330550918196994,
+      "grad_norm": 0.758755928135249,
+      "learning_rate": 9.165275459098498e-06,
+      "loss": 0.3208,
+      "step": 549
+    },
+    {
+      "epoch": 0.18363939899833054,
+      "grad_norm": 0.7133666636629313,
+      "learning_rate": 9.181969949916528e-06,
+      "loss": 0.3209,
+      "step": 550
+    },
+    {
+      "epoch": 0.18397328881469116,
+      "grad_norm": 0.7676565185405027,
+      "learning_rate": 9.198664440734558e-06,
+      "loss": 0.3185,
+      "step": 551
+    },
+    {
+      "epoch": 0.18430717863105175,
+      "grad_norm": 0.7361446038958627,
+      "learning_rate": 9.215358931552588e-06,
+      "loss": 0.3201,
+      "step": 552
+    },
+    {
+      "epoch": 0.18464106844741235,
+      "grad_norm": 0.7204730215612852,
+      "learning_rate": 9.232053422370618e-06,
+      "loss": 0.3263,
+      "step": 553
+    },
+    {
+      "epoch": 0.18497495826377294,
+      "grad_norm": 0.7361208545355521,
+      "learning_rate": 9.248747913188648e-06,
+      "loss": 0.324,
+      "step": 554
+    },
+    {
+      "epoch": 0.18530884808013356,
+      "grad_norm": 0.738179843713012,
+      "learning_rate": 9.265442404006678e-06,
+      "loss": 0.3356,
+      "step": 555
+    },
+    {
+      "epoch": 0.18564273789649416,
+      "grad_norm": 0.806880536497213,
+      "learning_rate": 9.282136894824708e-06,
+      "loss": 0.318,
+      "step": 556
+    },
+    {
+      "epoch": 0.18597662771285475,
+      "grad_norm": 0.7388701261078625,
+      "learning_rate": 9.298831385642738e-06,
+      "loss": 0.3331,
+      "step": 557
+    },
+    {
+      "epoch": 0.18631051752921535,
+      "grad_norm": 0.7506732460629391,
+      "learning_rate": 9.315525876460768e-06,
+      "loss": 0.3224,
+      "step": 558
+    },
+    {
+      "epoch": 0.18664440734557597,
+      "grad_norm": 0.8175305799080922,
+      "learning_rate": 9.332220367278798e-06,
+      "loss": 0.3298,
+      "step": 559
+    },
+    {
+      "epoch": 0.18697829716193656,
+      "grad_norm": 0.747961803709094,
+      "learning_rate": 9.348914858096828e-06,
+      "loss": 0.3194,
+      "step": 560
+    },
+    {
+      "epoch": 0.18731218697829716,
+      "grad_norm": 0.7328982858114919,
+      "learning_rate": 9.365609348914858e-06,
+      "loss": 0.3156,
+      "step": 561
+    },
+    {
+      "epoch": 0.18764607679465775,
+      "grad_norm": 0.7284983768990672,
+      "learning_rate": 9.382303839732888e-06,
+      "loss": 0.3268,
+      "step": 562
+    },
+    {
+      "epoch": 0.18797996661101837,
+      "grad_norm": 0.750376940412196,
+      "learning_rate": 9.398998330550918e-06,
+      "loss": 0.3179,
+      "step": 563
+    },
+    {
+      "epoch": 0.18831385642737897,
+      "grad_norm": 0.8061560063673782,
+      "learning_rate": 9.415692821368949e-06,
+      "loss": 0.308,
+      "step": 564
+    },
+    {
+      "epoch": 0.18864774624373956,
+      "grad_norm": 0.7217806194796712,
+      "learning_rate": 9.432387312186979e-06,
+      "loss": 0.3171,
+      "step": 565
+    },
+    {
+      "epoch": 0.18898163606010016,
+      "grad_norm": 0.7122433729475894,
+      "learning_rate": 9.449081803005009e-06,
+      "loss": 0.3168,
+      "step": 566
+    },
+    {
+      "epoch": 0.18931552587646078,
+      "grad_norm": 0.8409398462794953,
+      "learning_rate": 9.465776293823039e-06,
+      "loss": 0.3292,
+      "step": 567
+    },
+    {
+      "epoch": 0.18964941569282137,
+      "grad_norm": 0.7304005281015755,
+      "learning_rate": 9.482470784641069e-06,
+      "loss": 0.3248,
+      "step": 568
+    },
+    {
+      "epoch": 0.18998330550918197,
+      "grad_norm": 0.8060093948742936,
+      "learning_rate": 9.499165275459099e-06,
+      "loss": 0.3286,
+      "step": 569
+    },
+    {
+      "epoch": 0.19031719532554256,
+      "grad_norm": 0.8118415514399662,
+      "learning_rate": 9.515859766277129e-06,
+      "loss": 0.3372,
+      "step": 570
+    },
+    {
+      "epoch": 0.19065108514190318,
+      "grad_norm": 0.7187778947555116,
+      "learning_rate": 9.532554257095159e-06,
+      "loss": 0.3232,
+      "step": 571
+    },
+    {
+      "epoch": 0.19098497495826378,
+      "grad_norm": 0.7201219366455208,
+      "learning_rate": 9.549248747913189e-06,
+      "loss": 0.307,
+      "step": 572
+    },
+    {
+      "epoch": 0.19131886477462437,
+      "grad_norm": 0.7460863117892408,
+      "learning_rate": 9.565943238731219e-06,
+      "loss": 0.3102,
+      "step": 573
+    },
+    {
+      "epoch": 0.19165275459098496,
+      "grad_norm": 0.7122303794885726,
+      "learning_rate": 9.582637729549249e-06,
+      "loss": 0.3132,
+      "step": 574
+    },
+    {
+      "epoch": 0.19198664440734559,
+      "grad_norm": 0.7213911082242184,
+      "learning_rate": 9.59933222036728e-06,
+      "loss": 0.3243,
+      "step": 575
+    },
+    {
+      "epoch": 0.19232053422370618,
+      "grad_norm": 0.7007687675261132,
+      "learning_rate": 9.61602671118531e-06,
+      "loss": 0.3187,
+      "step": 576
+    },
+    {
+      "epoch": 0.19265442404006677,
+      "grad_norm": 0.6758436526554914,
+      "learning_rate": 9.63272120200334e-06,
+      "loss": 0.3189,
+      "step": 577
+    },
+    {
+      "epoch": 0.19298831385642737,
+      "grad_norm": 0.7289262297511554,
+      "learning_rate": 9.64941569282137e-06,
+      "loss": 0.3389,
+      "step": 578
+    },
+    {
+      "epoch": 0.193322203672788,
+      "grad_norm": 0.7386672049446915,
+      "learning_rate": 9.6661101836394e-06,
+      "loss": 0.3244,
+      "step": 579
+    },
+    {
+      "epoch": 0.19365609348914858,
+      "grad_norm": 0.7389719351622202,
+      "learning_rate": 9.68280467445743e-06,
+      "loss": 0.3136,
+      "step": 580
+    },
+    {
+      "epoch": 0.19398998330550918,
+      "grad_norm": 0.6874035268647203,
+      "learning_rate": 9.69949916527546e-06,
+      "loss": 0.3051,
+      "step": 581
+    },
+    {
+      "epoch": 0.19432387312186977,
+      "grad_norm": 0.7092118422661184,
+      "learning_rate": 9.71619365609349e-06,
+      "loss": 0.3164,
+      "step": 582
+    },
+    {
+      "epoch": 0.1946577629382304,
+      "grad_norm": 0.6929817344088215,
+      "learning_rate": 9.73288814691152e-06,
+      "loss": 0.3155,
+      "step": 583
+    },
+    {
+      "epoch": 0.194991652754591,
+      "grad_norm": 0.6722396603362585,
+      "learning_rate": 9.74958263772955e-06,
+      "loss": 0.3077,
+      "step": 584
+    },
+    {
+      "epoch": 0.19532554257095158,
+      "grad_norm": 0.708142486327623,
+      "learning_rate": 9.76627712854758e-06,
+      "loss": 0.3223,
+      "step": 585
+    },
+    {
+      "epoch": 0.19565943238731218,
+      "grad_norm": 1.3081377288851128,
+      "learning_rate": 9.78297161936561e-06,
+      "loss": 0.3233,
+      "step": 586
+    },
+    {
+      "epoch": 0.1959933222036728,
+      "grad_norm": 0.7448352942678667,
+      "learning_rate": 9.79966611018364e-06,
+      "loss": 0.3114,
+      "step": 587
+    },
+    {
+      "epoch": 0.1963272120200334,
+      "grad_norm": 0.7203850987359588,
+      "learning_rate": 9.81636060100167e-06,
+      "loss": 0.3223,
+      "step": 588
+    },
+    {
+      "epoch": 0.196661101836394,
+      "grad_norm": 0.7255559036917043,
+      "learning_rate": 9.8330550918197e-06,
+      "loss": 0.3146,
+      "step": 589
+    },
+    {
+      "epoch": 0.19699499165275458,
+      "grad_norm": 0.6601251342382983,
+      "learning_rate": 9.84974958263773e-06,
+      "loss": 0.3155,
+      "step": 590
+    },
+    {
+      "epoch": 0.1973288814691152,
+      "grad_norm": 0.6816346382680538,
+      "learning_rate": 9.86644407345576e-06,
+      "loss": 0.3141,
+      "step": 591
+    },
+    {
+      "epoch": 0.1976627712854758,
+      "grad_norm": 0.6818783959287772,
+      "learning_rate": 9.88313856427379e-06,
+      "loss": 0.3266,
+      "step": 592
+    },
+    {
+      "epoch": 0.1979966611018364,
+      "grad_norm": 0.69102653793309,
+      "learning_rate": 9.89983305509182e-06,
+      "loss": 0.3332,
+      "step": 593
+    },
+    {
+      "epoch": 0.19833055091819699,
+      "grad_norm": 0.746668659618295,
+      "learning_rate": 9.91652754590985e-06,
+      "loss": 0.3269,
+      "step": 594
+    },
+    {
+      "epoch": 0.1986644407345576,
+      "grad_norm": 0.7160358119470471,
+      "learning_rate": 9.93322203672788e-06,
+      "loss": 0.3363,
+      "step": 595
+    },
+    {
+      "epoch": 0.1989983305509182,
+      "grad_norm": 0.7268426823725448,
+      "learning_rate": 9.94991652754591e-06,
+      "loss": 0.3084,
+      "step": 596
+    },
+    {
+      "epoch": 0.1993322203672788,
+      "grad_norm": 0.7010785200293571,
+      "learning_rate": 9.96661101836394e-06,
+      "loss": 0.3179,
+      "step": 597
+    },
+    {
+      "epoch": 0.1996661101836394,
+      "grad_norm": 0.6959453607875318,
+      "learning_rate": 9.98330550918197e-06,
+      "loss": 0.3232,
+      "step": 598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7239064723796267,
+      "learning_rate": 1e-05,
+      "loss": 0.3091,
+      "step": 599
+    },
+    {
+      "epoch": 0.2003338898163606,
+      "grad_norm": 0.7745232285807138,
+      "learning_rate": 9.999999151012927e-06,
+      "loss": 0.3291,
+      "step": 600
+    },
+    {
+      "epoch": 0.2006677796327212,
+      "grad_norm": 0.7223958043958382,
+      "learning_rate": 9.999996604051993e-06,
+      "loss": 0.3246,
+      "step": 601
+    },
+    {
+      "epoch": 0.2010016694490818,
+      "grad_norm": 0.751606064379803,
+      "learning_rate": 9.999992359118065e-06,
+      "loss": 0.3091,
+      "step": 602
+    },
+    {
+      "epoch": 0.20133555926544242,
+      "grad_norm": 0.694971562221629,
+      "learning_rate": 9.999986416212582e-06,
+      "loss": 0.3133,
+      "step": 603
+    },
+    {
+      "epoch": 0.201669449081803,
+      "grad_norm": 0.7185164462644095,
+      "learning_rate": 9.999978775337564e-06,
+      "loss": 0.3322,
+      "step": 604
+    },
+    {
+      "epoch": 0.2020033388981636,
+      "grad_norm": 0.681013937105291,
+      "learning_rate": 9.999969436495606e-06,
+      "loss": 0.3158,
+      "step": 605
+    },
+    {
+      "epoch": 0.2023372287145242,
+      "grad_norm": 0.7329800880854658,
+      "learning_rate": 9.99995839968988e-06,
+      "loss": 0.3201,
+      "step": 606
+    },
+    {
+      "epoch": 0.20267111853088482,
+      "grad_norm": 0.6970542074311304,
+      "learning_rate": 9.999945664924132e-06,
+      "loss": 0.3059,
+      "step": 607
+    },
+    {
+      "epoch": 0.2030050083472454,
+      "grad_norm": 0.7434452529264757,
+      "learning_rate": 9.999931232202689e-06,
+      "loss": 0.3181,
+      "step": 608
+    },
+    {
+      "epoch": 0.203338898163606,
+      "grad_norm": 0.7090013111128189,
+      "learning_rate": 9.99991510153045e-06,
+      "loss": 0.321,
+      "step": 609
+    },
+    {
+      "epoch": 0.2036727879799666,
+      "grad_norm": 0.6736158443893092,
+      "learning_rate": 9.999897272912895e-06,
+      "loss": 0.3163,
+      "step": 610
+    },
+    {
+      "epoch": 0.20400667779632722,
+      "grad_norm": 0.8676037312437357,
+      "learning_rate": 9.999877746356074e-06,
+      "loss": 0.3295,
+      "step": 611
+    },
+    {
+      "epoch": 0.20434056761268782,
+      "grad_norm": 0.6886411003063958,
+      "learning_rate": 9.999856521866625e-06,
+      "loss": 0.3026,
+      "step": 612
+    },
+    {
+      "epoch": 0.2046744574290484,
+      "grad_norm": 0.7343183934886601,
+      "learning_rate": 9.999833599451751e-06,
+      "loss": 0.3249,
+      "step": 613
+    },
+    {
+      "epoch": 0.205008347245409,
+      "grad_norm": 0.731862891684194,
+      "learning_rate": 9.999808979119239e-06,
+      "loss": 0.3117,
+      "step": 614
+    },
+    {
+      "epoch": 0.20534223706176963,
+      "grad_norm": 0.718079797696134,
+      "learning_rate": 9.999782660877448e-06,
+      "loss": 0.3179,
+      "step": 615
+    },
+    {
+      "epoch": 0.20567612687813022,
+      "grad_norm": 0.648150337420773,
+      "learning_rate": 9.999754644735316e-06,
+      "loss": 0.3152,
+      "step": 616
+    },
+    {
+      "epoch": 0.20601001669449082,
+      "grad_norm": 0.7025225659194315,
+      "learning_rate": 9.999724930702358e-06,
+      "loss": 0.3329,
+      "step": 617
+    },
+    {
+      "epoch": 0.2063439065108514,
+      "grad_norm": 0.7784510600869818,
+      "learning_rate": 9.999693518788662e-06,
+      "loss": 0.3097,
+      "step": 618
+    },
+    {
+      "epoch": 0.20667779632721203,
+      "grad_norm": 0.6795263653491268,
+      "learning_rate": 9.999660409004896e-06,
+      "loss": 0.3222,
+      "step": 619
+    },
+    {
+      "epoch": 0.20701168614357263,
+      "grad_norm": 0.7672246726091313,
+      "learning_rate": 9.999625601362308e-06,
+      "loss": 0.3215,
+      "step": 620
+    },
+    {
+      "epoch": 0.20734557595993322,
+      "grad_norm": 0.7766948716039482,
+      "learning_rate": 9.999589095872715e-06,
+      "loss": 0.3198,
+      "step": 621
+    },
+    {
+      "epoch": 0.20767946577629381,
+      "grad_norm": 0.7108533952581885,
+      "learning_rate": 9.999550892548515e-06,
+      "loss": 0.3286,
+      "step": 622
+    },
+    {
+      "epoch": 0.20801335559265444,
+      "grad_norm": 0.7553845334763537,
+      "learning_rate": 9.999510991402681e-06,
+      "loss": 0.3343,
+      "step": 623
+    },
+    {
+      "epoch": 0.20834724540901503,
+      "grad_norm": 0.7169164019201072,
+      "learning_rate": 9.999469392448764e-06,
+      "loss": 0.3167,
+      "step": 624
+    },
+    {
+      "epoch": 0.20868113522537562,
+      "grad_norm": 0.6863166674086179,
+      "learning_rate": 9.99942609570089e-06,
+      "loss": 0.3251,
+      "step": 625
+    },
+    {
+      "epoch": 0.20901502504173622,
+      "grad_norm": 0.7122117748912319,
+      "learning_rate": 9.999381101173765e-06,
+      "loss": 0.316,
+      "step": 626
+    },
+    {
+      "epoch": 0.20934891485809684,
+      "grad_norm": 0.7560006847234226,
+      "learning_rate": 9.999334408882662e-06,
+      "loss": 0.328,
+      "step": 627
+    },
+    {
+      "epoch": 0.20968280467445743,
+      "grad_norm": 0.6882516185141836,
+      "learning_rate": 9.999286018843447e-06,
+      "loss": 0.303,
+      "step": 628
+    },
+    {
+      "epoch": 0.21001669449081803,
+      "grad_norm": 0.7085792479897741,
+      "learning_rate": 9.999235931072544e-06,
+      "loss": 0.3151,
+      "step": 629
+    },
+    {
+      "epoch": 0.21035058430717862,
+      "grad_norm": 0.7034103781406703,
+      "learning_rate": 9.999184145586968e-06,
+      "loss": 0.3117,
+      "step": 630
+    },
+    {
+      "epoch": 0.21068447412353924,
+      "grad_norm": 0.6549817701843921,
+      "learning_rate": 9.999130662404306e-06,
+      "loss": 0.3066,
+      "step": 631
+    },
+    {
+      "epoch": 0.21101836393989984,
+      "grad_norm": 0.6853147424902278,
+      "learning_rate": 9.999075481542717e-06,
+      "loss": 0.3191,
+      "step": 632
+    },
+    {
+      "epoch": 0.21135225375626043,
+      "grad_norm": 0.6665715334694383,
+      "learning_rate": 9.99901860302094e-06,
+      "loss": 0.3267,
+      "step": 633
+    },
+    {
+      "epoch": 0.21168614357262103,
+      "grad_norm": 0.6913590815490434,
+      "learning_rate": 9.998960026858296e-06,
+      "loss": 0.3137,
+      "step": 634
+    },
+    {
+      "epoch": 0.21202003338898165,
+      "grad_norm": 0.6674423705712718,
+      "learning_rate": 9.99889975307467e-06,
+      "loss": 0.3179,
+      "step": 635
+    },
+    {
+      "epoch": 0.21235392320534224,
+      "grad_norm": 0.6568011636091821,
+      "learning_rate": 9.998837781690535e-06,
+      "loss": 0.312,
+      "step": 636
+    },
+    {
+      "epoch": 0.21268781302170284,
+      "grad_norm": 0.6809155664611595,
+      "learning_rate": 9.998774112726935e-06,
+      "loss": 0.3071,
+      "step": 637
+    },
+    {
+      "epoch": 0.21302170283806343,
+      "grad_norm": 0.6937891792941617,
+      "learning_rate": 9.998708746205493e-06,
+      "loss": 0.3108,
+      "step": 638
+    },
+    {
+      "epoch": 0.21335559265442405,
+      "grad_norm": 0.6945179624728876,
+      "learning_rate": 9.998641682148404e-06,
+      "loss": 0.3141,
+      "step": 639
+    },
+    {
+      "epoch": 0.21368948247078465,
+      "grad_norm": 0.6845633698508677,
+      "learning_rate": 9.998572920578444e-06,
+      "loss": 0.3236,
+      "step": 640
+    },
+    {
+      "epoch": 0.21402337228714524,
+      "grad_norm": 0.7408509325257491,
+      "learning_rate": 9.998502461518966e-06,
+      "loss": 0.3317,
+      "step": 641
+    },
+    {
+      "epoch": 0.21435726210350584,
+      "grad_norm": 0.6965688356917853,
+      "learning_rate": 9.998430304993899e-06,
+      "loss": 0.3157,
+      "step": 642
+    },
+    {
+      "epoch": 0.21469115191986646,
+      "grad_norm": 0.6588378424406256,
+      "learning_rate": 9.99835645102774e-06,
+      "loss": 0.3156,
+      "step": 643
+    },
+    {
+      "epoch": 0.21502504173622705,
+      "grad_norm": 0.6680774795595205,
+      "learning_rate": 9.998280899645575e-06,
+      "loss": 0.3172,
+      "step": 644
+    },
+    {
+      "epoch": 0.21535893155258765,
+      "grad_norm": 0.6884547044807267,
+      "learning_rate": 9.99820365087306e-06,
+      "loss": 0.3191,
+      "step": 645
+    },
+    {
+      "epoch": 0.21569282136894824,
+      "grad_norm": 0.6469820809013782,
+      "learning_rate": 9.998124704736426e-06,
+      "loss": 0.311,
+      "step": 646
+    },
+    {
+      "epoch": 0.21602671118530886,
+      "grad_norm": 0.6510578291967526,
+      "learning_rate": 9.998044061262486e-06,
+      "loss": 0.3103,
+      "step": 647
+    },
+    {
+      "epoch": 0.21636060100166946,
+      "grad_norm": 0.7062528941189932,
+      "learning_rate": 9.997961720478623e-06,
+      "loss": 0.3244,
+      "step": 648
+    },
+    {
+      "epoch": 0.21669449081803005,
+      "grad_norm": 0.6798412153493817,
+      "learning_rate": 9.997877682412804e-06,
+      "loss": 0.3122,
+      "step": 649
+    },
+    {
+      "epoch": 0.21702838063439064,
+      "grad_norm": 0.6608217515577914,
+      "learning_rate": 9.997791947093562e-06,
+      "loss": 0.3216,
+      "step": 650
+    },
+    {
+      "epoch": 0.21736227045075127,
+      "grad_norm": 0.6528286326671019,
+      "learning_rate": 9.997704514550017e-06,
+      "loss": 0.3183,
+      "step": 651
+    },
+    {
+      "epoch": 0.21769616026711186,
+      "grad_norm": 0.6605088333903899,
+      "learning_rate": 9.997615384811857e-06,
+      "loss": 0.3095,
+      "step": 652
+    },
+    {
+      "epoch": 0.21803005008347245,
+      "grad_norm": 0.7074573664905897,
+      "learning_rate": 9.997524557909353e-06,
+      "loss": 0.328,
+      "step": 653
+    },
+    {
+      "epoch": 0.21836393989983305,
+      "grad_norm": 0.6298243505710597,
+      "learning_rate": 9.997432033873348e-06,
+      "loss": 0.3069,
+      "step": 654
+    },
+    {
+      "epoch": 0.21869782971619364,
+      "grad_norm": 0.7141163520859374,
+      "learning_rate": 9.997337812735263e-06,
+      "loss": 0.3284,
+      "step": 655
+    },
+    {
+      "epoch": 0.21903171953255426,
+      "grad_norm": 0.7116306013120169,
+      "learning_rate": 9.997241894527096e-06,
+      "loss": 0.3177,
+      "step": 656
+    },
+    {
+      "epoch": 0.21936560934891486,
+      "grad_norm": 0.6691335487575459,
+      "learning_rate": 9.997144279281416e-06,
+      "loss": 0.3177,
+      "step": 657
+    },
+    {
+      "epoch": 0.21969949916527545,
+      "grad_norm": 0.6500904355711586,
+      "learning_rate": 9.99704496703138e-06,
+      "loss": 0.305,
+      "step": 658
+    },
+    {
+      "epoch": 0.22003338898163605,
+      "grad_norm": 0.7084194167262223,
+      "learning_rate": 9.996943957810708e-06,
+      "loss": 0.3043,
+      "step": 659
+    },
+    {
+      "epoch": 0.22036727879799667,
+      "grad_norm": 0.6830261131516596,
+      "learning_rate": 9.996841251653701e-06,
+      "loss": 0.3174,
+      "step": 660
+    },
+    {
+      "epoch": 0.22070116861435726,
+      "grad_norm": 0.6981676365378354,
+      "learning_rate": 9.996736848595244e-06,
+      "loss": 0.3161,
+      "step": 661
+    },
+    {
+      "epoch": 0.22103505843071786,
+      "grad_norm": 0.7338241211464367,
+      "learning_rate": 9.996630748670788e-06,
+      "loss": 0.3226,
+      "step": 662
+    },
+    {
+      "epoch": 0.22136894824707845,
+      "grad_norm": 0.6891460328994479,
+      "learning_rate": 9.996522951916363e-06,
+      "loss": 0.3056,
+      "step": 663
+    },
+    {
+      "epoch": 0.22170283806343907,
+      "grad_norm": 0.6662408900907806,
+      "learning_rate": 9.996413458368578e-06,
+      "loss": 0.3066,
+      "step": 664
+    },
+    {
+      "epoch": 0.22203672787979967,
+      "grad_norm": 0.7158827926019817,
+      "learning_rate": 9.996302268064615e-06,
+      "loss": 0.3239,
+      "step": 665
+    },
+    {
+      "epoch": 0.22237061769616026,
+      "grad_norm": 0.6726852004414356,
+      "learning_rate": 9.996189381042236e-06,
+      "loss": 0.3146,
+      "step": 666
+    },
+    {
+      "epoch": 0.22270450751252086,
+      "grad_norm": 0.6475011139242043,
+      "learning_rate": 9.996074797339774e-06,
+      "loss": 0.2997,
+      "step": 667
+    },
+    {
+      "epoch": 0.22303839732888148,
+      "grad_norm": 0.6787911819217508,
+      "learning_rate": 9.995958516996143e-06,
+      "loss": 0.3117,
+      "step": 668
+    },
+    {
+      "epoch": 0.22337228714524207,
+      "grad_norm": 0.7363344381162076,
+      "learning_rate": 9.99584054005083e-06,
+      "loss": 0.3138,
+      "step": 669
+    },
+    {
+      "epoch": 0.22370617696160267,
+      "grad_norm": 0.6635505152204142,
+      "learning_rate": 9.9957208665439e-06,
+      "loss": 0.3233,
+      "step": 670
+    },
+    {
+      "epoch": 0.22404006677796326,
+      "grad_norm": 0.6788439989646472,
+      "learning_rate": 9.995599496515996e-06,
+      "loss": 0.3082,
+      "step": 671
+    },
+    {
+      "epoch": 0.22437395659432388,
+      "grad_norm": 0.718099648428951,
+      "learning_rate": 9.995476430008328e-06,
+      "loss": 0.3152,
+      "step": 672
+    },
+    {
+      "epoch": 0.22470784641068448,
+      "grad_norm": 0.6723598767616826,
+      "learning_rate": 9.995351667062697e-06,
+      "loss": 0.3038,
+      "step": 673
+    },
+    {
+      "epoch": 0.22504173622704507,
+      "grad_norm": 0.6206965718044407,
+      "learning_rate": 9.995225207721465e-06,
+      "loss": 0.3081,
+      "step": 674
+    },
+    {
+      "epoch": 0.22537562604340566,
+      "grad_norm": 0.6826516233631149,
+      "learning_rate": 9.995097052027578e-06,
+      "loss": 0.296,
+      "step": 675
+    },
+    {
+      "epoch": 0.22570951585976629,
+      "grad_norm": 0.67408120339682,
+      "learning_rate": 9.994967200024562e-06,
+      "loss": 0.293,
+      "step": 676
+    },
+    {
+      "epoch": 0.22604340567612688,
+      "grad_norm": 0.6539606742871614,
+      "learning_rate": 9.99483565175651e-06,
+      "loss": 0.3069,
+      "step": 677
+    },
+    {
+      "epoch": 0.22637729549248747,
+      "grad_norm": 0.6508817751958446,
+      "learning_rate": 9.994702407268093e-06,
+      "loss": 0.3156,
+      "step": 678
+    },
+    {
+      "epoch": 0.22671118530884807,
+      "grad_norm": 0.6922346563671706,
+      "learning_rate": 9.994567466604566e-06,
+      "loss": 0.3189,
+      "step": 679
+    },
+    {
+      "epoch": 0.2270450751252087,
+      "grad_norm": 0.7538722012509047,
+      "learning_rate": 9.99443082981175e-06,
+      "loss": 0.3186,
+      "step": 680
+    },
+    {
+      "epoch": 0.22737896494156928,
+      "grad_norm": 0.675957804094594,
+      "learning_rate": 9.994292496936048e-06,
+      "loss": 0.3253,
+      "step": 681
+    },
+    {
+      "epoch": 0.22771285475792988,
+      "grad_norm": 0.7351938767111457,
+      "learning_rate": 9.994152468024435e-06,
+      "loss": 0.3086,
+      "step": 682
+    },
+    {
+      "epoch": 0.22804674457429047,
+      "grad_norm": 0.710203449328641,
+      "learning_rate": 9.994010743124466e-06,
+      "loss": 0.3067,
+      "step": 683
+    },
+    {
+      "epoch": 0.2283806343906511,
+      "grad_norm": 0.6812089199080917,
+      "learning_rate": 9.99386732228427e-06,
+      "loss": 0.3131,
+      "step": 684
+    },
+    {
+      "epoch": 0.2287145242070117,
+      "grad_norm": 0.618669305759647,
+      "learning_rate": 9.993722205552552e-06,
+      "loss": 0.302,
+      "step": 685
+    },
+    {
+      "epoch": 0.22904841402337228,
+      "grad_norm": 0.6878170194997228,
+      "learning_rate": 9.993575392978592e-06,
+      "loss": 0.3182,
+      "step": 686
+    },
+    {
+      "epoch": 0.22938230383973288,
+      "grad_norm": 0.7276831538446582,
+      "learning_rate": 9.993426884612246e-06,
+      "loss": 0.3243,
+      "step": 687
+    },
+    {
+      "epoch": 0.2297161936560935,
+      "grad_norm": 0.6528958357714014,
+      "learning_rate": 9.99327668050395e-06,
+      "loss": 0.323,
+      "step": 688
+    },
+    {
+      "epoch": 0.2300500834724541,
+      "grad_norm": 0.6394663341402973,
+      "learning_rate": 9.993124780704707e-06,
+      "loss": 0.3074,
+      "step": 689
+    },
+    {
+      "epoch": 0.2303839732888147,
+      "grad_norm": 0.667139903809351,
+      "learning_rate": 9.992971185266107e-06,
+      "loss": 0.3083,
+      "step": 690
+    },
+    {
+      "epoch": 0.23071786310517528,
+      "grad_norm": 0.6847082339741472,
+      "learning_rate": 9.992815894240309e-06,
+      "loss": 0.3221,
+      "step": 691
+    },
+    {
+      "epoch": 0.2310517529215359,
+      "grad_norm": 0.6605500744268817,
+      "learning_rate": 9.992658907680046e-06,
+      "loss": 0.3046,
+      "step": 692
+    },
+    {
+      "epoch": 0.2313856427378965,
+      "grad_norm": 0.6599478102893481,
+      "learning_rate": 9.992500225638633e-06,
+      "loss": 0.3246,
+      "step": 693
+    },
+    {
+      "epoch": 0.2317195325542571,
+      "grad_norm": 0.7097267858031422,
+      "learning_rate": 9.992339848169956e-06,
+      "loss": 0.3095,
+      "step": 694
+    },
+    {
+      "epoch": 0.23205342237061768,
+      "grad_norm": 0.7335874793337843,
+      "learning_rate": 9.992177775328479e-06,
+      "loss": 0.3086,
+      "step": 695
+    },
+    {
+      "epoch": 0.2323873121869783,
+      "grad_norm": 0.6609989400868227,
+      "learning_rate": 9.992014007169241e-06,
+      "loss": 0.3141,
+      "step": 696
+    },
+    {
+      "epoch": 0.2327212020033389,
+      "grad_norm": 0.7307970178053497,
+      "learning_rate": 9.991848543747857e-06,
+      "loss": 0.3167,
+      "step": 697
+    },
+    {
+      "epoch": 0.2330550918196995,
+      "grad_norm": 0.7305081391056903,
+      "learning_rate": 9.991681385120515e-06,
+      "loss": 0.3209,
+      "step": 698
+    },
+    {
+      "epoch": 0.2333889816360601,
+      "grad_norm": 0.6450372423194908,
+      "learning_rate": 9.991512531343987e-06,
+      "loss": 0.3197,
+      "step": 699
+    },
+    {
+      "epoch": 0.2337228714524207,
+      "grad_norm": 0.7381241391725759,
+      "learning_rate": 9.991341982475609e-06,
+      "loss": 0.3166,
+      "step": 700
+    },
+    {
+      "epoch": 0.2340567612687813,
+      "grad_norm": 0.7590561682502889,
+      "learning_rate": 9.991169738573301e-06,
+      "loss": 0.3043,
+      "step": 701
+    },
+    {
+      "epoch": 0.2343906510851419,
+      "grad_norm": 0.6445162214337684,
+      "learning_rate": 9.990995799695557e-06,
+      "loss": 0.3087,
+      "step": 702
+    },
+    {
+      "epoch": 0.2347245409015025,
+      "grad_norm": 0.6387694717685652,
+      "learning_rate": 9.990820165901444e-06,
+      "loss": 0.3055,
+      "step": 703
+    },
+    {
+      "epoch": 0.23505843071786311,
+      "grad_norm": 0.7481976313018593,
+      "learning_rate": 9.990642837250607e-06,
+      "loss": 0.3167,
+      "step": 704
+    },
+    {
+      "epoch": 0.2353923205342237,
+      "grad_norm": 0.6983907632877976,
+      "learning_rate": 9.990463813803266e-06,
+      "loss": 0.3228,
+      "step": 705
+    },
+    {
+      "epoch": 0.2357262103505843,
+      "grad_norm": 0.6688651583984132,
+      "learning_rate": 9.990283095620216e-06,
+      "loss": 0.3268,
+      "step": 706
+    },
+    {
+      "epoch": 0.2360601001669449,
+      "grad_norm": 0.6770839790504161,
+      "learning_rate": 9.99010068276283e-06,
+      "loss": 0.3181,
+      "step": 707
+    },
+    {
+      "epoch": 0.23639398998330552,
+      "grad_norm": 0.6714156721437909,
+      "learning_rate": 9.989916575293051e-06,
+      "loss": 0.3186,
+      "step": 708
+    },
+    {
+      "epoch": 0.2367278797996661,
+      "grad_norm": 0.6540150370462013,
+      "learning_rate": 9.989730773273404e-06,
+      "loss": 0.3223,
+      "step": 709
+    },
+    {
+      "epoch": 0.2370617696160267,
+      "grad_norm": 0.6609189182450268,
+      "learning_rate": 9.989543276766983e-06,
+      "loss": 0.3162,
+      "step": 710
+    },
+    {
+      "epoch": 0.2373956594323873,
+      "grad_norm": 0.6984241358444584,
+      "learning_rate": 9.989354085837464e-06,
+      "loss": 0.2997,
+      "step": 711
+    },
+    {
+      "epoch": 0.23772954924874792,
+      "grad_norm": 0.6990231879510617,
+      "learning_rate": 9.989163200549095e-06,
+      "loss": 0.2971,
+      "step": 712
+    },
+    {
+      "epoch": 0.23806343906510852,
+      "grad_norm": 0.658541275183071,
+      "learning_rate": 9.988970620966698e-06,
+      "loss": 0.3061,
+      "step": 713
+    },
+    {
+      "epoch": 0.2383973288814691,
+      "grad_norm": 0.7022113400672207,
+      "learning_rate": 9.988776347155673e-06,
+      "loss": 0.3293,
+      "step": 714
+    },
+    {
+      "epoch": 0.2387312186978297,
+      "grad_norm": 0.7106124601542608,
+      "learning_rate": 9.988580379181997e-06,
+      "loss": 0.3224,
+      "step": 715
+    },
+    {
+      "epoch": 0.23906510851419033,
+      "grad_norm": 0.6547807633616505,
+      "learning_rate": 9.988382717112213e-06,
+      "loss": 0.3177,
+      "step": 716
+    },
+    {
+      "epoch": 0.23939899833055092,
+      "grad_norm": 0.641253722802366,
+      "learning_rate": 9.988183361013452e-06,
+      "loss": 0.322,
+      "step": 717
+    },
+    {
+      "epoch": 0.23973288814691152,
+      "grad_norm": 0.7177132192638013,
+      "learning_rate": 9.987982310953412e-06,
+      "loss": 0.3177,
+      "step": 718
+    },
+    {
+      "epoch": 0.2400667779632721,
+      "grad_norm": 0.6789961132917205,
+      "learning_rate": 9.987779567000369e-06,
+      "loss": 0.3298,
+      "step": 719
+    },
+    {
+      "epoch": 0.24040066777963273,
+      "grad_norm": 0.6363754776956632,
+      "learning_rate": 9.987575129223174e-06,
+      "loss": 0.3156,
+      "step": 720
+    },
+    {
+      "epoch": 0.24073455759599333,
+      "grad_norm": 0.6322175319754683,
+      "learning_rate": 9.987368997691253e-06,
+      "loss": 0.3137,
+      "step": 721
+    },
+    {
+      "epoch": 0.24106844741235392,
+      "grad_norm": 0.6570564644152928,
+      "learning_rate": 9.987161172474607e-06,
+      "loss": 0.307,
+      "step": 722
+    },
+    {
+      "epoch": 0.24140233722871451,
+      "grad_norm": 0.6063293126121128,
+      "learning_rate": 9.986951653643811e-06,
+      "loss": 0.3014,
+      "step": 723
+    },
+    {
+      "epoch": 0.24173622704507514,
+      "grad_norm": 0.6750259704940584,
+      "learning_rate": 9.98674044127002e-06,
+      "loss": 0.3067,
+      "step": 724
+    },
+    {
+      "epoch": 0.24207011686143573,
+      "grad_norm": 0.6943272857903212,
+      "learning_rate": 9.986527535424956e-06,
+      "loss": 0.3351,
+      "step": 725
+    },
+    {
+      "epoch": 0.24240400667779632,
+      "grad_norm": 0.639988921501206,
+      "learning_rate": 9.986312936180924e-06,
+      "loss": 0.3096,
+      "step": 726
+    },
+    {
+      "epoch": 0.24273789649415692,
+      "grad_norm": 0.6281533260576224,
+      "learning_rate": 9.9860966436108e-06,
+      "loss": 0.3079,
+      "step": 727
+    },
+    {
+      "epoch": 0.24307178631051754,
+      "grad_norm": 0.706857732853257,
+      "learning_rate": 9.985878657788037e-06,
+      "loss": 0.3144,
+      "step": 728
+    },
+    {
+      "epoch": 0.24340567612687813,
+      "grad_norm": 0.6541506616360614,
+      "learning_rate": 9.985658978786659e-06,
+      "loss": 0.3075,
+      "step": 729
+    },
+    {
+      "epoch": 0.24373956594323873,
+      "grad_norm": 0.6624835311614887,
+      "learning_rate": 9.985437606681269e-06,
+      "loss": 0.3071,
+      "step": 730
+    },
+    {
+      "epoch": 0.24407345575959932,
+      "grad_norm": 0.6473093171281169,
+      "learning_rate": 9.985214541547046e-06,
+      "loss": 0.3168,
+      "step": 731
+    },
+    {
+      "epoch": 0.24440734557595994,
+      "grad_norm": 0.6416368357968834,
+      "learning_rate": 9.98498978345974e-06,
+      "loss": 0.3229,
+      "step": 732
+    },
+    {
+      "epoch": 0.24474123539232054,
+      "grad_norm": 0.6304624138327128,
+      "learning_rate": 9.984763332495677e-06,
+      "loss": 0.3189,
+      "step": 733
+    },
+    {
+      "epoch": 0.24507512520868113,
+      "grad_norm": 0.6621370403008399,
+      "learning_rate": 9.98453518873176e-06,
+      "loss": 0.3144,
+      "step": 734
+    },
+    {
+      "epoch": 0.24540901502504173,
+      "grad_norm": 0.64561048624302,
+      "learning_rate": 9.984305352245465e-06,
+      "loss": 0.3142,
+      "step": 735
+    },
+    {
+      "epoch": 0.24574290484140235,
+      "grad_norm": 0.6864381565497274,
+      "learning_rate": 9.984073823114842e-06,
+      "loss": 0.3214,
+      "step": 736
+    },
+    {
+      "epoch": 0.24607679465776294,
+      "grad_norm": 0.6466422271417273,
+      "learning_rate": 9.983840601418518e-06,
+      "loss": 0.3338,
+      "step": 737
+    },
+    {
+      "epoch": 0.24641068447412354,
+      "grad_norm": 0.6429106154267435,
+      "learning_rate": 9.983605687235695e-06,
+      "loss": 0.3049,
+      "step": 738
+    },
+    {
+      "epoch": 0.24674457429048413,
+      "grad_norm": 0.6634273720609353,
+      "learning_rate": 9.983369080646148e-06,
+      "loss": 0.3071,
+      "step": 739
+    },
+    {
+      "epoch": 0.24707846410684475,
+      "grad_norm": 0.6401166842229258,
+      "learning_rate": 9.983130781730225e-06,
+      "loss": 0.3157,
+      "step": 740
+    },
+    {
+      "epoch": 0.24741235392320535,
+      "grad_norm": 0.7028339845948085,
+      "learning_rate": 9.982890790568855e-06,
+      "loss": 0.3178,
+      "step": 741
+    },
+    {
+      "epoch": 0.24774624373956594,
+      "grad_norm": 0.6547314864818602,
+      "learning_rate": 9.982649107243537e-06,
+      "loss": 0.3097,
+      "step": 742
+    },
+    {
+      "epoch": 0.24808013355592654,
+      "grad_norm": 0.7009206953232453,
+      "learning_rate": 9.982405731836343e-06,
+      "loss": 0.3081,
+      "step": 743
+    },
+    {
+      "epoch": 0.24841402337228716,
+      "grad_norm": 0.6561883332493822,
+      "learning_rate": 9.982160664429921e-06,
+      "loss": 0.3057,
+      "step": 744
+    },
+    {
+      "epoch": 0.24874791318864775,
+      "grad_norm": 0.6782214128217363,
+      "learning_rate": 9.9819139051075e-06,
+      "loss": 0.3201,
+      "step": 745
+    },
+    {
+      "epoch": 0.24908180300500835,
+      "grad_norm": 0.6581627262846691,
+      "learning_rate": 9.981665453952874e-06,
+      "loss": 0.3096,
+      "step": 746
+    },
+    {
+      "epoch": 0.24941569282136894,
+      "grad_norm": 0.6317050915286468,
+      "learning_rate": 9.981415311050417e-06,
+      "loss": 0.3042,
+      "step": 747
+    },
+    {
+      "epoch": 0.24974958263772956,
+      "grad_norm": 0.6731056365355634,
+      "learning_rate": 9.981163476485077e-06,
+      "loss": 0.3179,
+      "step": 748
+    },
+    {
+      "epoch": 0.2500834724540901,
+      "grad_norm": 0.6830510316558891,
+      "learning_rate": 9.980909950342374e-06,
+      "loss": 0.2943,
+      "step": 749
+    },
+    {
+      "epoch": 0.25041736227045075,
+      "grad_norm": 0.6762570902723388,
+      "learning_rate": 9.980654732708404e-06,
+      "loss": 0.3091,
+      "step": 750
+    },
+    {
+      "epoch": 0.25075125208681137,
+      "grad_norm": 0.6572673183049867,
+      "learning_rate": 9.980397823669841e-06,
+      "loss": 0.3138,
+      "step": 751
+    },
+    {
+      "epoch": 0.25108514190317194,
+      "grad_norm": 0.6345262533229002,
+      "learning_rate": 9.980139223313926e-06,
+      "loss": 0.3038,
+      "step": 752
+    },
+    {
+      "epoch": 0.25141903171953256,
+      "grad_norm": 0.6586686805136891,
+      "learning_rate": 9.979878931728479e-06,
+      "loss": 0.3124,
+      "step": 753
+    },
+    {
+      "epoch": 0.2517529215358932,
+      "grad_norm": 0.6763612681109208,
+      "learning_rate": 9.979616949001896e-06,
+      "loss": 0.308,
+      "step": 754
+    },
+    {
+      "epoch": 0.25208681135225375,
+      "grad_norm": 0.6653912885274064,
+      "learning_rate": 9.979353275223145e-06,
+      "loss": 0.3065,
+      "step": 755
+    },
+    {
+      "epoch": 0.25242070116861437,
+      "grad_norm": 0.6710077564536276,
+      "learning_rate": 9.979087910481765e-06,
+      "loss": 0.3306,
+      "step": 756
+    },
+    {
+      "epoch": 0.25275459098497494,
+      "grad_norm": 0.6529009454145228,
+      "learning_rate": 9.978820854867875e-06,
+      "loss": 0.319,
+      "step": 757
+    },
+    {
+      "epoch": 0.25308848080133556,
+      "grad_norm": 0.6637010979194617,
+      "learning_rate": 9.978552108472164e-06,
+      "loss": 0.3238,
+      "step": 758
+    },
+    {
+      "epoch": 0.2534223706176962,
+      "grad_norm": 0.6315031171492718,
+      "learning_rate": 9.978281671385899e-06,
+      "loss": 0.3114,
+      "step": 759
+    },
+    {
+      "epoch": 0.25375626043405675,
+      "grad_norm": 0.6353485947127433,
+      "learning_rate": 9.978009543700918e-06,
+      "loss": 0.3041,
+      "step": 760
+    },
+    {
+      "epoch": 0.25409015025041737,
+      "grad_norm": 0.6329709832202217,
+      "learning_rate": 9.977735725509632e-06,
+      "loss": 0.3171,
+      "step": 761
+    },
+    {
+      "epoch": 0.254424040066778,
+      "grad_norm": 0.6235490819608175,
+      "learning_rate": 9.977460216905032e-06,
+      "loss": 0.3187,
+      "step": 762
+    },
+    {
+      "epoch": 0.25475792988313856,
+      "grad_norm": 0.6243014555048773,
+      "learning_rate": 9.977183017980677e-06,
+      "loss": 0.3153,
+      "step": 763
+    },
+    {
+      "epoch": 0.2550918196994992,
+      "grad_norm": 0.6413256149071491,
+      "learning_rate": 9.976904128830705e-06,
+      "loss": 0.3213,
+      "step": 764
+    },
+    {
+      "epoch": 0.25542570951585974,
+      "grad_norm": 0.6003590097282294,
+      "learning_rate": 9.976623549549821e-06,
+      "loss": 0.313,
+      "step": 765
+    },
+    {
+      "epoch": 0.25575959933222037,
+      "grad_norm": 0.5975427236162498,
+      "learning_rate": 9.97634128023331e-06,
+      "loss": 0.3043,
+      "step": 766
+    },
+    {
+      "epoch": 0.256093489148581,
+      "grad_norm": 0.6180436775054071,
+      "learning_rate": 9.976057320977033e-06,
+      "loss": 0.3129,
+      "step": 767
+    },
+    {
+      "epoch": 0.25642737896494155,
+      "grad_norm": 0.5943941186437867,
+      "learning_rate": 9.975771671877414e-06,
+      "loss": 0.2996,
+      "step": 768
+    },
+    {
+      "epoch": 0.2567612687813022,
+      "grad_norm": 0.6222174655070934,
+      "learning_rate": 9.975484333031465e-06,
+      "loss": 0.3185,
+      "step": 769
+    },
+    {
+      "epoch": 0.2570951585976628,
+      "grad_norm": 0.6041499461634103,
+      "learning_rate": 9.97519530453676e-06,
+      "loss": 0.3128,
+      "step": 770
+    },
+    {
+      "epoch": 0.25742904841402336,
+      "grad_norm": 0.598338206721707,
+      "learning_rate": 9.974904586491455e-06,
+      "loss": 0.3,
+      "step": 771
+    },
+    {
+      "epoch": 0.257762938230384,
+      "grad_norm": 0.5829671831811722,
+      "learning_rate": 9.974612178994271e-06,
+      "loss": 0.2883,
+      "step": 772
+    },
+    {
+      "epoch": 0.25809682804674455,
+      "grad_norm": 0.6359437158094623,
+      "learning_rate": 9.974318082144515e-06,
+      "loss": 0.3158,
+      "step": 773
+    },
+    {
+      "epoch": 0.2584307178631052,
+      "grad_norm": 0.6399141635772083,
+      "learning_rate": 9.974022296042055e-06,
+      "loss": 0.3119,
+      "step": 774
+    },
+    {
+      "epoch": 0.2587646076794658,
+      "grad_norm": 0.6145661038326289,
+      "learning_rate": 9.973724820787343e-06,
+      "loss": 0.3151,
+      "step": 775
+    },
+    {
+      "epoch": 0.25909849749582636,
+      "grad_norm": 0.6376500398627979,
+      "learning_rate": 9.973425656481395e-06,
+      "loss": 0.3174,
+      "step": 776
+    },
+    {
+      "epoch": 0.259432387312187,
+      "grad_norm": 0.6763327380848617,
+      "learning_rate": 9.973124803225811e-06,
+      "loss": 0.3133,
+      "step": 777
+    },
+    {
+      "epoch": 0.2597662771285476,
+      "grad_norm": 0.6160305423608218,
+      "learning_rate": 9.972822261122757e-06,
+      "loss": 0.3,
+      "step": 778
+    },
+    {
+      "epoch": 0.2601001669449082,
+      "grad_norm": 0.5980193667235342,
+      "learning_rate": 9.97251803027497e-06,
+      "loss": 0.3068,
+      "step": 779
+    },
+    {
+      "epoch": 0.2604340567612688,
+      "grad_norm": 0.6146641076334989,
+      "learning_rate": 9.972212110785776e-06,
+      "loss": 0.3081,
+      "step": 780
+    },
+    {
+      "epoch": 0.26076794657762936,
+      "grad_norm": 0.7199431203525698,
+      "learning_rate": 9.971904502759053e-06,
+      "loss": 0.3232,
+      "step": 781
+    },
+    {
+      "epoch": 0.26110183639399,
+      "grad_norm": 0.6597146934365115,
+      "learning_rate": 9.97159520629927e-06,
+      "loss": 0.3033,
+      "step": 782
+    },
+    {
+      "epoch": 0.2614357262103506,
+      "grad_norm": 0.6290728071935953,
+      "learning_rate": 9.971284221511459e-06,
+      "loss": 0.3093,
+      "step": 783
+    },
+    {
+      "epoch": 0.26176961602671117,
+      "grad_norm": 0.6695340028487,
+      "learning_rate": 9.970971548501229e-06,
+      "loss": 0.3081,
+      "step": 784
+    },
+    {
+      "epoch": 0.2621035058430718,
+      "grad_norm": 0.7036250941493171,
+      "learning_rate": 9.970657187374762e-06,
+      "loss": 0.3131,
+      "step": 785
+    },
+    {
+      "epoch": 0.26243739565943236,
+      "grad_norm": 0.6418561751369942,
+      "learning_rate": 9.970341138238817e-06,
+      "loss": 0.3074,
+      "step": 786
+    },
+    {
+      "epoch": 0.262771285475793,
+      "grad_norm": 0.6690664220595359,
+      "learning_rate": 9.970023401200719e-06,
+      "loss": 0.3235,
+      "step": 787
+    },
+    {
+      "epoch": 0.2631051752921536,
+      "grad_norm": 0.6054361060490654,
+      "learning_rate": 9.969703976368368e-06,
+      "loss": 0.3199,
+      "step": 788
+    },
+    {
+      "epoch": 0.26343906510851417,
+      "grad_norm": 0.6730688364410078,
+      "learning_rate": 9.969382863850245e-06,
+      "loss": 0.3156,
+      "step": 789
+    },
+    {
+      "epoch": 0.2637729549248748,
+      "grad_norm": 0.6323579620758437,
+      "learning_rate": 9.969060063755395e-06,
+      "loss": 0.3066,
+      "step": 790
+    },
+    {
+      "epoch": 0.2641068447412354,
+      "grad_norm": 0.6170550053584963,
+      "learning_rate": 9.968735576193438e-06,
+      "loss": 0.3072,
+      "step": 791
+    },
+    {
+      "epoch": 0.264440734557596,
+      "grad_norm": 0.6617785829149546,
+      "learning_rate": 9.96840940127457e-06,
+      "loss": 0.3047,
+      "step": 792
+    },
+    {
+      "epoch": 0.2647746243739566,
+      "grad_norm": 0.6720624802363009,
+      "learning_rate": 9.968081539109557e-06,
+      "loss": 0.3206,
+      "step": 793
+    },
+    {
+      "epoch": 0.26510851419031717,
+      "grad_norm": 0.6133015488725736,
+      "learning_rate": 9.967751989809741e-06,
+      "loss": 0.3062,
+      "step": 794
+    },
+    {
+      "epoch": 0.2654424040066778,
+      "grad_norm": 0.6220779862572104,
+      "learning_rate": 9.967420753487035e-06,
+      "loss": 0.3123,
+      "step": 795
+    },
+    {
+      "epoch": 0.2657762938230384,
+      "grad_norm": 0.6188687328063293,
+      "learning_rate": 9.967087830253923e-06,
+      "loss": 0.3056,
+      "step": 796
+    },
+    {
+      "epoch": 0.266110183639399,
+      "grad_norm": 0.632067572297075,
+      "learning_rate": 9.966753220223466e-06,
+      "loss": 0.3204,
+      "step": 797
+    },
+    {
+      "epoch": 0.2664440734557596,
+      "grad_norm": 0.6408091321082425,
+      "learning_rate": 9.966416923509296e-06,
+      "loss": 0.3197,
+      "step": 798
+    },
+    {
+      "epoch": 0.2667779632721202,
+      "grad_norm": 0.6596373030760121,
+      "learning_rate": 9.966078940225617e-06,
+      "loss": 0.3071,
+      "step": 799
+    },
+    {
+      "epoch": 0.2671118530884808,
+      "grad_norm": 0.6174257657310275,
+      "learning_rate": 9.965739270487207e-06,
+      "loss": 0.312,
+      "step": 800
+    },
+    {
+      "epoch": 0.2674457429048414,
+      "grad_norm": 0.6324539857700394,
+      "learning_rate": 9.965397914409414e-06,
+      "loss": 0.2956,
+      "step": 801
+    },
+    {
+      "epoch": 0.267779632721202,
+      "grad_norm": 0.6016148330714408,
+      "learning_rate": 9.965054872108161e-06,
+      "loss": 0.309,
+      "step": 802
+    },
+    {
+      "epoch": 0.2681135225375626,
+      "grad_norm": 0.6149517546121147,
+      "learning_rate": 9.964710143699948e-06,
+      "loss": 0.3022,
+      "step": 803
+    },
+    {
+      "epoch": 0.2684474123539232,
+      "grad_norm": 0.6046688492459947,
+      "learning_rate": 9.964363729301838e-06,
+      "loss": 0.3013,
+      "step": 804
+    },
+    {
+      "epoch": 0.2687813021702838,
+      "grad_norm": 0.6545950425004171,
+      "learning_rate": 9.964015629031472e-06,
+      "loss": 0.3167,
+      "step": 805
+    },
+    {
+      "epoch": 0.2691151919866444,
+      "grad_norm": 0.5938991067621764,
+      "learning_rate": 9.963665843007066e-06,
+      "loss": 0.307,
+      "step": 806
+    },
+    {
+      "epoch": 0.26944908180300503,
+      "grad_norm": 0.6128778217572743,
+      "learning_rate": 9.963314371347403e-06,
+      "loss": 0.3054,
+      "step": 807
+    },
+    {
+      "epoch": 0.2697829716193656,
+      "grad_norm": 0.6592426203760766,
+      "learning_rate": 9.96296121417184e-06,
+      "loss": 0.315,
+      "step": 808
+    },
+    {
+      "epoch": 0.2701168614357262,
+      "grad_norm": 0.5735881514735979,
+      "learning_rate": 9.962606371600311e-06,
+      "loss": 0.3133,
+      "step": 809
+    },
+    {
+      "epoch": 0.2704507512520868,
+      "grad_norm": 0.6133261444199122,
+      "learning_rate": 9.962249843753317e-06,
+      "loss": 0.3173,
+      "step": 810
+    },
+    {
+      "epoch": 0.2707846410684474,
+      "grad_norm": 0.58760751289117,
+      "learning_rate": 9.961891630751931e-06,
+      "loss": 0.3189,
+      "step": 811
+    },
+    {
+      "epoch": 0.27111853088480803,
+      "grad_norm": 0.6436516239764292,
+      "learning_rate": 9.961531732717803e-06,
+      "loss": 0.3198,
+      "step": 812
+    },
+    {
+      "epoch": 0.2714524207011686,
+      "grad_norm": 0.6087599452838778,
+      "learning_rate": 9.961170149773152e-06,
+      "loss": 0.3036,
+      "step": 813
+    },
+    {
+      "epoch": 0.2717863105175292,
+      "grad_norm": 0.6326523541121202,
+      "learning_rate": 9.960806882040767e-06,
+      "loss": 0.3112,
+      "step": 814
+    },
+    {
+      "epoch": 0.27212020033388984,
+      "grad_norm": 0.6033965111455138,
+      "learning_rate": 9.960441929644017e-06,
+      "loss": 0.3126,
+      "step": 815
+    },
+    {
+      "epoch": 0.2724540901502504,
+      "grad_norm": 0.6644104609308248,
+      "learning_rate": 9.960075292706834e-06,
+      "loss": 0.303,
+      "step": 816
+    },
+    {
+      "epoch": 0.272787979966611,
+      "grad_norm": 0.6371955514725776,
+      "learning_rate": 9.959706971353727e-06,
+      "loss": 0.3014,
+      "step": 817
+    },
+    {
+      "epoch": 0.2731218697829716,
+      "grad_norm": 0.6235706062744508,
+      "learning_rate": 9.959336965709776e-06,
+      "loss": 0.3084,
+      "step": 818
+    },
+    {
+      "epoch": 0.2734557595993322,
+      "grad_norm": 0.6672261753900592,
+      "learning_rate": 9.958965275900632e-06,
+      "loss": 0.3219,
+      "step": 819
+    },
+    {
+      "epoch": 0.27378964941569284,
+      "grad_norm": 0.6607695254803473,
+      "learning_rate": 9.958591902052521e-06,
+      "loss": 0.3039,
+      "step": 820
+    },
+    {
+      "epoch": 0.2741235392320534,
+      "grad_norm": 0.6108089911354792,
+      "learning_rate": 9.958216844292238e-06,
+      "loss": 0.314,
+      "step": 821
+    },
+    {
+      "epoch": 0.274457429048414,
+      "grad_norm": 0.6287309789420155,
+      "learning_rate": 9.957840102747149e-06,
+      "loss": 0.3046,
+      "step": 822
+    },
+    {
+      "epoch": 0.27479131886477465,
+      "grad_norm": 0.6532920319675541,
+      "learning_rate": 9.957461677545198e-06,
+      "loss": 0.3026,
+      "step": 823
+    },
+    {
+      "epoch": 0.2751252086811352,
+      "grad_norm": 0.6246621105164231,
+      "learning_rate": 9.95708156881489e-06,
+      "loss": 0.3118,
+      "step": 824
+    },
+    {
+      "epoch": 0.27545909849749584,
+      "grad_norm": 0.5939778012257733,
+      "learning_rate": 9.956699776685312e-06,
+      "loss": 0.3162,
+      "step": 825
+    },
+    {
+      "epoch": 0.2757929883138564,
+      "grad_norm": 0.6629949622161345,
+      "learning_rate": 9.95631630128612e-06,
+      "loss": 0.2978,
+      "step": 826
+    },
+    {
+      "epoch": 0.276126878130217,
+      "grad_norm": 0.6576030374253635,
+      "learning_rate": 9.955931142747534e-06,
+      "loss": 0.3072,
+      "step": 827
+    },
+    {
+      "epoch": 0.27646076794657765,
+      "grad_norm": 0.639397090187686,
+      "learning_rate": 9.95554430120036e-06,
+      "loss": 0.305,
+      "step": 828
+    },
+    {
+      "epoch": 0.2767946577629382,
+      "grad_norm": 0.612633753276582,
+      "learning_rate": 9.955155776775961e-06,
+      "loss": 0.3077,
+      "step": 829
+    },
+    {
+      "epoch": 0.27712854757929883,
+      "grad_norm": 0.622340442933951,
+      "learning_rate": 9.95476556960628e-06,
+      "loss": 0.3022,
+      "step": 830
+    },
+    {
+      "epoch": 0.27746243739565946,
+      "grad_norm": 0.6956024322322204,
+      "learning_rate": 9.954373679823831e-06,
+      "loss": 0.327,
+      "step": 831
+    },
+    {
+      "epoch": 0.27779632721202,
+      "grad_norm": 0.6353332509335413,
+      "learning_rate": 9.953980107561695e-06,
+      "loss": 0.3031,
+      "step": 832
+    },
+    {
+      "epoch": 0.27813021702838064,
+      "grad_norm": 0.6446973340749373,
+      "learning_rate": 9.95358485295353e-06,
+      "loss": 0.3082,
+      "step": 833
+    },
+    {
+      "epoch": 0.2784641068447412,
+      "grad_norm": 0.6193258689477105,
+      "learning_rate": 9.95318791613356e-06,
+      "loss": 0.306,
+      "step": 834
+    },
+    {
+      "epoch": 0.27879799666110183,
+      "grad_norm": 0.6540728439482395,
+      "learning_rate": 9.952789297236583e-06,
+      "loss": 0.3145,
+      "step": 835
+    },
+    {
+      "epoch": 0.27913188647746245,
+      "grad_norm": 0.6135161298895859,
+      "learning_rate": 9.952388996397968e-06,
+      "loss": 0.297,
+      "step": 836
+    },
+    {
+      "epoch": 0.279465776293823,
+      "grad_norm": 0.6508669142838964,
+      "learning_rate": 9.951987013753657e-06,
+      "loss": 0.3154,
+      "step": 837
+    },
+    {
+      "epoch": 0.27979966611018364,
+      "grad_norm": 0.6372224087132735,
+      "learning_rate": 9.951583349440157e-06,
+      "loss": 0.315,
+      "step": 838
+    },
+    {
+      "epoch": 0.28013355592654426,
+      "grad_norm": 0.6432842664350981,
+      "learning_rate": 9.951178003594556e-06,
+      "loss": 0.3294,
+      "step": 839
+    },
+    {
+      "epoch": 0.28046744574290483,
+      "grad_norm": 0.5985164583810018,
+      "learning_rate": 9.950770976354503e-06,
+      "loss": 0.3154,
+      "step": 840
+    },
+    {
+      "epoch": 0.28080133555926545,
+      "grad_norm": 0.6050407051105324,
+      "learning_rate": 9.950362267858225e-06,
+      "loss": 0.3104,
+      "step": 841
+    },
+    {
+      "epoch": 0.281135225375626,
+      "grad_norm": 0.652899183286552,
+      "learning_rate": 9.949951878244514e-06,
+      "loss": 0.3103,
+      "step": 842
+    },
+    {
+      "epoch": 0.28146911519198664,
+      "grad_norm": 0.6486628187527617,
+      "learning_rate": 9.949539807652739e-06,
+      "loss": 0.3103,
+      "step": 843
+    },
+    {
+      "epoch": 0.28180300500834726,
+      "grad_norm": 0.5967143729280339,
+      "learning_rate": 9.949126056222837e-06,
+      "loss": 0.3124,
+      "step": 844
+    },
+    {
+      "epoch": 0.28213689482470783,
+      "grad_norm": 0.6360280330541884,
+      "learning_rate": 9.948710624095315e-06,
+      "loss": 0.2989,
+      "step": 845
+    },
+    {
+      "epoch": 0.28247078464106845,
+      "grad_norm": 0.6386664945793225,
+      "learning_rate": 9.948293511411251e-06,
+      "loss": 0.3117,
+      "step": 846
+    },
+    {
+      "epoch": 0.2828046744574291,
+      "grad_norm": 0.6452501535544478,
+      "learning_rate": 9.947874718312295e-06,
+      "loss": 0.3097,
+      "step": 847
+    },
+    {
+      "epoch": 0.28313856427378964,
+      "grad_norm": 0.6137756495691662,
+      "learning_rate": 9.947454244940668e-06,
+      "loss": 0.3129,
+      "step": 848
+    },
+    {
+      "epoch": 0.28347245409015026,
+      "grad_norm": 0.6268646511627051,
+      "learning_rate": 9.947032091439159e-06,
+      "loss": 0.314,
+      "step": 849
+    },
+    {
+      "epoch": 0.2838063439065108,
+      "grad_norm": 0.5993640220506576,
+      "learning_rate": 9.94660825795113e-06,
+      "loss": 0.2941,
+      "step": 850
+    },
+    {
+      "epoch": 0.28414023372287145,
+      "grad_norm": 0.6367649376877658,
+      "learning_rate": 9.946182744620512e-06,
+      "loss": 0.3156,
+      "step": 851
+    },
+    {
+      "epoch": 0.28447412353923207,
+      "grad_norm": 0.6326244480084043,
+      "learning_rate": 9.945755551591808e-06,
+      "loss": 0.2981,
+      "step": 852
+    },
+    {
+      "epoch": 0.28480801335559264,
+      "grad_norm": 0.6174511458601781,
+      "learning_rate": 9.94532667901009e-06,
+      "loss": 0.3113,
+      "step": 853
+    },
+    {
+      "epoch": 0.28514190317195326,
+      "grad_norm": 0.6169030263186861,
+      "learning_rate": 9.944896127020999e-06,
+      "loss": 0.3018,
+      "step": 854
+    },
+    {
+      "epoch": 0.2854757929883139,
+      "grad_norm": 0.6300643798136133,
+      "learning_rate": 9.944463895770754e-06,
+      "loss": 0.314,
+      "step": 855
+    },
+    {
+      "epoch": 0.28580968280467445,
+      "grad_norm": 0.6082254546577714,
+      "learning_rate": 9.944029985406131e-06,
+      "loss": 0.3098,
+      "step": 856
+    },
+    {
+      "epoch": 0.28614357262103507,
+      "grad_norm": 0.6074909418739824,
+      "learning_rate": 9.94359439607449e-06,
+      "loss": 0.3054,
+      "step": 857
+    },
+    {
+      "epoch": 0.28647746243739564,
+      "grad_norm": 0.652448043642393,
+      "learning_rate": 9.943157127923751e-06,
+      "loss": 0.3214,
+      "step": 858
+    },
+    {
+      "epoch": 0.28681135225375626,
+      "grad_norm": 0.5877136474093703,
+      "learning_rate": 9.942718181102408e-06,
+      "loss": 0.3126,
+      "step": 859
+    },
+    {
+      "epoch": 0.2871452420701169,
+      "grad_norm": 0.6257533942404674,
+      "learning_rate": 9.94227755575953e-06,
+      "loss": 0.3075,
+      "step": 860
+    },
+    {
+      "epoch": 0.28747913188647745,
+      "grad_norm": 0.5791993280370373,
+      "learning_rate": 9.941835252044744e-06,
+      "loss": 0.2881,
+      "step": 861
+    },
+    {
+      "epoch": 0.28781302170283807,
+      "grad_norm": 0.6005164764425798,
+      "learning_rate": 9.941391270108261e-06,
+      "loss": 0.3006,
+      "step": 862
+    },
+    {
+      "epoch": 0.2881469115191987,
+      "grad_norm": 0.701833392140988,
+      "learning_rate": 9.940945610100849e-06,
+      "loss": 0.3139,
+      "step": 863
+    },
+    {
+      "epoch": 0.28848080133555926,
+      "grad_norm": 0.6247902317359704,
+      "learning_rate": 9.940498272173855e-06,
+      "loss": 0.3169,
+      "step": 864
+    },
+    {
+      "epoch": 0.2888146911519199,
+      "grad_norm": 0.6032069294518242,
+      "learning_rate": 9.94004925647919e-06,
+      "loss": 0.3079,
+      "step": 865
+    },
+    {
+      "epoch": 0.28914858096828044,
+      "grad_norm": 0.7275877516031075,
+      "learning_rate": 9.939598563169342e-06,
+      "loss": 0.3164,
+      "step": 866
+    },
+    {
+      "epoch": 0.28948247078464107,
+      "grad_norm": 0.6681350452707889,
+      "learning_rate": 9.93914619239736e-06,
+      "loss": 0.3274,
+      "step": 867
+    },
+    {
+      "epoch": 0.2898163606010017,
+      "grad_norm": 0.6370789861891926,
+      "learning_rate": 9.938692144316869e-06,
+      "loss": 0.3271,
+      "step": 868
+    },
+    {
+      "epoch": 0.29015025041736225,
+      "grad_norm": 0.5995903697964612,
+      "learning_rate": 9.93823641908206e-06,
+      "loss": 0.3083,
+      "step": 869
+    },
+    {
+      "epoch": 0.2904841402337229,
+      "grad_norm": 0.614804563003791,
+      "learning_rate": 9.937779016847698e-06,
+      "loss": 0.3178,
+      "step": 870
+    },
+    {
+      "epoch": 0.2908180300500835,
+      "grad_norm": 0.6357584795866368,
+      "learning_rate": 9.937319937769108e-06,
+      "loss": 0.3061,
+      "step": 871
+    },
+    {
+      "epoch": 0.29115191986644406,
+      "grad_norm": 0.6045326716093058,
+      "learning_rate": 9.936859182002199e-06,
+      "loss": 0.3118,
+      "step": 872
+    },
+    {
+      "epoch": 0.2914858096828047,
+      "grad_norm": 0.610026464735996,
+      "learning_rate": 9.936396749703435e-06,
+      "loss": 0.3091,
+      "step": 873
+    },
+    {
+      "epoch": 0.29181969949916525,
+      "grad_norm": 0.6263634988192537,
+      "learning_rate": 9.935932641029857e-06,
+      "loss": 0.316,
+      "step": 874
+    },
+    {
+      "epoch": 0.2921535893155259,
+      "grad_norm": 0.61245514624297,
+      "learning_rate": 9.935466856139075e-06,
+      "loss": 0.2904,
+      "step": 875
+    },
+    {
+      "epoch": 0.2924874791318865,
+      "grad_norm": 0.6406169406117876,
+      "learning_rate": 9.934999395189269e-06,
+      "loss": 0.2954,
+      "step": 876
+    },
+    {
+      "epoch": 0.29282136894824706,
+      "grad_norm": 0.6337912417121114,
+      "learning_rate": 9.93453025833918e-06,
+      "loss": 0.3103,
+      "step": 877
+    },
+    {
+      "epoch": 0.2931552587646077,
+      "grad_norm": 0.603214360656413,
+      "learning_rate": 9.934059445748134e-06,
+      "loss": 0.2991,
+      "step": 878
+    },
+    {
+      "epoch": 0.2934891485809683,
+      "grad_norm": 0.6048660815291924,
+      "learning_rate": 9.933586957576006e-06,
+      "loss": 0.3093,
+      "step": 879
+    },
+    {
+      "epoch": 0.2938230383973289,
+      "grad_norm": 0.5821535940407092,
+      "learning_rate": 9.93311279398326e-06,
+      "loss": 0.303,
+      "step": 880
+    },
+    {
+      "epoch": 0.2941569282136895,
+      "grad_norm": 0.5924884634943405,
+      "learning_rate": 9.932636955130913e-06,
+      "loss": 0.3143,
+      "step": 881
+    },
+    {
+      "epoch": 0.29449081803005006,
+      "grad_norm": 0.5736507645921805,
+      "learning_rate": 9.93215944118056e-06,
+      "loss": 0.3127,
+      "step": 882
+    },
+    {
+      "epoch": 0.2948247078464107,
+      "grad_norm": 0.6068061583925178,
+      "learning_rate": 9.931680252294363e-06,
+      "loss": 0.3112,
+      "step": 883
+    },
+    {
+      "epoch": 0.2951585976627713,
+      "grad_norm": 0.6226722350485845,
+      "learning_rate": 9.931199388635048e-06,
+      "loss": 0.3063,
+      "step": 884
+    },
+    {
+      "epoch": 0.29549248747913187,
+      "grad_norm": 0.5915478624612948,
+      "learning_rate": 9.930716850365919e-06,
+      "loss": 0.2995,
+      "step": 885
+    },
+    {
+      "epoch": 0.2958263772954925,
+      "grad_norm": 0.5862641333612119,
+      "learning_rate": 9.93023263765084e-06,
+      "loss": 0.2992,
+      "step": 886
+    },
+    {
+      "epoch": 0.2961602671118531,
+      "grad_norm": 0.6456116649503595,
+      "learning_rate": 9.92974675065425e-06,
+      "loss": 0.3023,
+      "step": 887
+    },
+    {
+      "epoch": 0.2964941569282137,
+      "grad_norm": 0.6122322708397328,
+      "learning_rate": 9.92925918954115e-06,
+      "loss": 0.3104,
+      "step": 888
+    },
+    {
+      "epoch": 0.2968280467445743,
+      "grad_norm": 0.6163031911027558,
+      "learning_rate": 9.928769954477117e-06,
+      "loss": 0.3144,
+      "step": 889
+    },
+    {
+      "epoch": 0.29716193656093487,
+      "grad_norm": 0.5967959336344377,
+      "learning_rate": 9.928279045628291e-06,
+      "loss": 0.3151,
+      "step": 890
+    },
+    {
+      "epoch": 0.2974958263772955,
+      "grad_norm": 0.6051518523078615,
+      "learning_rate": 9.927786463161382e-06,
+      "loss": 0.3012,
+      "step": 891
+    },
+    {
+      "epoch": 0.2978297161936561,
+      "grad_norm": 0.6140349683251903,
+      "learning_rate": 9.927292207243667e-06,
+      "loss": 0.3067,
+      "step": 892
+    },
+    {
+      "epoch": 0.2981636060100167,
+      "grad_norm": 0.5929846921078934,
+      "learning_rate": 9.926796278042997e-06,
+      "loss": 0.3135,
+      "step": 893
+    },
+    {
+      "epoch": 0.2984974958263773,
+      "grad_norm": 0.6267852171890413,
+      "learning_rate": 9.92629867572778e-06,
+      "loss": 0.2958,
+      "step": 894
+    },
+    {
+      "epoch": 0.2988313856427379,
+      "grad_norm": 0.6253263841617958,
+      "learning_rate": 9.925799400467008e-06,
+      "loss": 0.3106,
+      "step": 895
+    },
+    {
+      "epoch": 0.2991652754590985,
+      "grad_norm": 0.6525180216282289,
+      "learning_rate": 9.925298452430225e-06,
+      "loss": 0.3033,
+      "step": 896
+    },
+    {
+      "epoch": 0.2994991652754591,
+      "grad_norm": 0.6059360080835277,
+      "learning_rate": 9.924795831787555e-06,
+      "loss": 0.308,
+      "step": 897
+    },
+    {
+      "epoch": 0.2998330550918197,
+      "grad_norm": 0.6178477928798163,
+      "learning_rate": 9.924291538709683e-06,
+      "loss": 0.3069,
+      "step": 898
+    },
+    {
+      "epoch": 0.3001669449081803,
+      "grad_norm": 0.6073735011961882,
+      "learning_rate": 9.923785573367865e-06,
+      "loss": 0.3075,
+      "step": 899
+    },
+    {
+      "epoch": 0.3005008347245409,
+      "grad_norm": 0.6713402502566992,
+      "learning_rate": 9.923277935933925e-06,
+      "loss": 0.3125,
+      "step": 900
+    },
+    {
+      "epoch": 0.3008347245409015,
+      "grad_norm": 0.5788643114402922,
+      "learning_rate": 9.922768626580251e-06,
+      "loss": 0.3159,
+      "step": 901
+    },
+    {
+      "epoch": 0.3011686143572621,
+      "grad_norm": 0.59678330467515,
+      "learning_rate": 9.922257645479803e-06,
+      "loss": 0.2965,
+      "step": 902
+    },
+    {
+      "epoch": 0.30150250417362273,
+      "grad_norm": 0.6147051127111449,
+      "learning_rate": 9.92174499280611e-06,
+      "loss": 0.3102,
+      "step": 903
+    },
+    {
+      "epoch": 0.3018363939899833,
+      "grad_norm": 0.5799504911300849,
+      "learning_rate": 9.921230668733265e-06,
+      "loss": 0.317,
+      "step": 904
+    },
+    {
+      "epoch": 0.3021702838063439,
+      "grad_norm": 0.6210064699522896,
+      "learning_rate": 9.920714673435931e-06,
+      "loss": 0.3179,
+      "step": 905
+    },
+    {
+      "epoch": 0.3025041736227045,
+      "grad_norm": 0.6230884070271024,
+      "learning_rate": 9.920197007089334e-06,
+      "loss": 0.3064,
+      "step": 906
+    },
+    {
+      "epoch": 0.3028380634390651,
+      "grad_norm": 0.6191385332016315,
+      "learning_rate": 9.919677669869272e-06,
+      "loss": 0.3083,
+      "step": 907
+    },
+    {
+      "epoch": 0.30317195325542573,
+      "grad_norm": 0.5695686485326757,
+      "learning_rate": 9.919156661952113e-06,
+      "loss": 0.3118,
+      "step": 908
+    },
+    {
+      "epoch": 0.3035058430717863,
+      "grad_norm": 0.6490134948567529,
+      "learning_rate": 9.918633983514782e-06,
+      "loss": 0.3217,
+      "step": 909
+    },
+    {
+      "epoch": 0.3038397328881469,
+      "grad_norm": 0.6459794079652185,
+      "learning_rate": 9.918109634734783e-06,
+      "loss": 0.3191,
+      "step": 910
+    },
+    {
+      "epoch": 0.30417362270450754,
+      "grad_norm": 0.6151562602535916,
+      "learning_rate": 9.917583615790178e-06,
+      "loss": 0.3001,
+      "step": 911
+    },
+    {
+      "epoch": 0.3045075125208681,
+      "grad_norm": 0.5811633381035104,
+      "learning_rate": 9.917055926859606e-06,
+      "loss": 0.3017,
+      "step": 912
+    },
+    {
+      "epoch": 0.30484140233722873,
+      "grad_norm": 0.6351629035976917,
+      "learning_rate": 9.916526568122262e-06,
+      "loss": 0.3125,
+      "step": 913
+    },
+    {
+      "epoch": 0.3051752921535893,
+      "grad_norm": 0.6401909517786118,
+      "learning_rate": 9.915995539757918e-06,
+      "loss": 0.2976,
+      "step": 914
+    },
+    {
+      "epoch": 0.3055091819699499,
+      "grad_norm": 0.5853229069783724,
+      "learning_rate": 9.915462841946905e-06,
+      "loss": 0.2885,
+      "step": 915
+    },
+    {
+      "epoch": 0.30584307178631054,
+      "grad_norm": 0.6096945401975741,
+      "learning_rate": 9.914928474870125e-06,
+      "loss": 0.3139,
+      "step": 916
+    },
+    {
+      "epoch": 0.3061769616026711,
+      "grad_norm": 0.6458656151809331,
+      "learning_rate": 9.914392438709049e-06,
+      "loss": 0.3121,
+      "step": 917
+    },
+    {
+      "epoch": 0.3065108514190317,
+      "grad_norm": 0.6237791675616119,
+      "learning_rate": 9.913854733645708e-06,
+      "loss": 0.3108,
+      "step": 918
+    },
+    {
+      "epoch": 0.3068447412353923,
+      "grad_norm": 0.5951657364915529,
+      "learning_rate": 9.913315359862705e-06,
+      "loss": 0.3125,
+      "step": 919
+    },
+    {
+      "epoch": 0.3071786310517529,
+      "grad_norm": 0.6362705943692695,
+      "learning_rate": 9.912774317543212e-06,
+      "loss": 0.3015,
+      "step": 920
+    },
+    {
+      "epoch": 0.30751252086811354,
+      "grad_norm": 0.6360727681813466,
+      "learning_rate": 9.912231606870961e-06,
+      "loss": 0.3158,
+      "step": 921
+    },
+    {
+      "epoch": 0.3078464106844741,
+      "grad_norm": 0.5908010791554803,
+      "learning_rate": 9.911687228030253e-06,
+      "loss": 0.3016,
+      "step": 922
+    },
+    {
+      "epoch": 0.3081803005008347,
+      "grad_norm": 0.610241864878574,
+      "learning_rate": 9.91114118120596e-06,
+      "loss": 0.2883,
+      "step": 923
+    },
+    {
+      "epoch": 0.30851419031719535,
+      "grad_norm": 0.6014963278613095,
+      "learning_rate": 9.910593466583512e-06,
+      "loss": 0.3046,
+      "step": 924
+    },
+    {
+      "epoch": 0.3088480801335559,
+      "grad_norm": 0.6102154820179821,
+      "learning_rate": 9.910044084348913e-06,
+      "loss": 0.3025,
+      "step": 925
+    },
+    {
+      "epoch": 0.30918196994991654,
+      "grad_norm": 0.6217136091368356,
+      "learning_rate": 9.90949303468873e-06,
+      "loss": 0.3134,
+      "step": 926
+    },
+    {
+      "epoch": 0.3095158597662771,
+      "grad_norm": 0.6269550246575194,
+      "learning_rate": 9.908940317790096e-06,
+      "loss": 0.3164,
+      "step": 927
+    },
+    {
+      "epoch": 0.3098497495826377,
+      "grad_norm": 0.5942409385000056,
+      "learning_rate": 9.908385933840711e-06,
+      "loss": 0.3166,
+      "step": 928
+    },
+    {
+      "epoch": 0.31018363939899835,
+      "grad_norm": 0.6104189024798128,
+      "learning_rate": 9.907829883028842e-06,
+      "loss": 0.311,
+      "step": 929
+    },
+    {
+      "epoch": 0.3105175292153589,
+      "grad_norm": 0.6184123736632998,
+      "learning_rate": 9.907272165543317e-06,
+      "loss": 0.3099,
+      "step": 930
+    },
+    {
+      "epoch": 0.31085141903171953,
+      "grad_norm": 0.5995027379027167,
+      "learning_rate": 9.90671278157354e-06,
+      "loss": 0.3142,
+      "step": 931
+    },
+    {
+      "epoch": 0.31118530884808016,
+      "grad_norm": 0.59716576290827,
+      "learning_rate": 9.906151731309472e-06,
+      "loss": 0.3003,
+      "step": 932
+    },
+    {
+      "epoch": 0.3115191986644407,
+      "grad_norm": 0.6502434823089829,
+      "learning_rate": 9.905589014941641e-06,
+      "loss": 0.3065,
+      "step": 933
+    },
+    {
+      "epoch": 0.31185308848080134,
+      "grad_norm": 0.6045199407656775,
+      "learning_rate": 9.905024632661146e-06,
+      "loss": 0.3069,
+      "step": 934
+    },
+    {
+      "epoch": 0.3121869782971619,
+      "grad_norm": 0.6256990705341531,
+      "learning_rate": 9.904458584659645e-06,
+      "loss": 0.3076,
+      "step": 935
+    },
+    {
+      "epoch": 0.31252086811352253,
+      "grad_norm": 0.6293499304182937,
+      "learning_rate": 9.903890871129367e-06,
+      "loss": 0.3065,
+      "step": 936
+    },
+    {
+      "epoch": 0.31285475792988315,
+      "grad_norm": 0.5808596840302044,
+      "learning_rate": 9.903321492263106e-06,
+      "loss": 0.3006,
+      "step": 937
+    },
+    {
+      "epoch": 0.3131886477462437,
+      "grad_norm": 0.6396977042643922,
+      "learning_rate": 9.902750448254214e-06,
+      "loss": 0.302,
+      "step": 938
+    },
+    {
+      "epoch": 0.31352253756260434,
+      "grad_norm": 0.5989466937223189,
+      "learning_rate": 9.902177739296621e-06,
+      "loss": 0.3163,
+      "step": 939
+    },
+    {
+      "epoch": 0.31385642737896496,
+      "grad_norm": 0.6069586803818303,
+      "learning_rate": 9.901603365584816e-06,
+      "loss": 0.3062,
+      "step": 940
+    },
+    {
+      "epoch": 0.31419031719532553,
+      "grad_norm": 0.6177855712554505,
+      "learning_rate": 9.901027327313847e-06,
+      "loss": 0.3,
+      "step": 941
+    },
+    {
+      "epoch": 0.31452420701168615,
+      "grad_norm": 0.6554672668856966,
+      "learning_rate": 9.90044962467934e-06,
+      "loss": 0.3176,
+      "step": 942
+    },
+    {
+      "epoch": 0.3148580968280467,
+      "grad_norm": 0.5402208871075836,
+      "learning_rate": 9.89987025787748e-06,
+      "loss": 0.2924,
+      "step": 943
+    },
+    {
+      "epoch": 0.31519198664440734,
+      "grad_norm": 0.571132194967242,
+      "learning_rate": 9.899289227105012e-06,
+      "loss": 0.2992,
+      "step": 944
+    },
+    {
+      "epoch": 0.31552587646076796,
+      "grad_norm": 0.5966294568997814,
+      "learning_rate": 9.898706532559252e-06,
+      "loss": 0.2996,
+      "step": 945
+    },
+    {
+      "epoch": 0.31585976627712853,
+      "grad_norm": 0.6081699473817996,
+      "learning_rate": 9.898122174438083e-06,
+      "loss": 0.3088,
+      "step": 946
+    },
+    {
+      "epoch": 0.31619365609348915,
+      "grad_norm": 0.5817352874365662,
+      "learning_rate": 9.897536152939949e-06,
+      "loss": 0.3073,
+      "step": 947
+    },
+    {
+      "epoch": 0.3165275459098498,
+      "grad_norm": 0.5999553478474449,
+      "learning_rate": 9.89694846826386e-06,
+      "loss": 0.3155,
+      "step": 948
+    },
+    {
+      "epoch": 0.31686143572621034,
+      "grad_norm": 0.6084259353248475,
+      "learning_rate": 9.896359120609389e-06,
+      "loss": 0.2969,
+      "step": 949
+    },
+    {
+      "epoch": 0.31719532554257096,
+      "grad_norm": 0.5938293023600885,
+      "learning_rate": 9.895768110176677e-06,
+      "loss": 0.3159,
+      "step": 950
+    },
+    {
+      "epoch": 0.3175292153589315,
+      "grad_norm": 0.5737885007109333,
+      "learning_rate": 9.89517543716643e-06,
+      "loss": 0.2983,
+      "step": 951
+    },
+    {
+      "epoch": 0.31786310517529215,
+      "grad_norm": 0.6296551074060512,
+      "learning_rate": 9.894581101779912e-06,
+      "loss": 0.3174,
+      "step": 952
+    },
+    {
+      "epoch": 0.31819699499165277,
+      "grad_norm": 0.6137153267475868,
+      "learning_rate": 9.893985104218959e-06,
+      "loss": 0.3158,
+      "step": 953
+    },
+    {
+      "epoch": 0.31853088480801334,
+      "grad_norm": 0.5849209543270956,
+      "learning_rate": 9.893387444685969e-06,
+      "loss": 0.3081,
+      "step": 954
+    },
+    {
+      "epoch": 0.31886477462437396,
+      "grad_norm": 0.6583531819595155,
+      "learning_rate": 9.892788123383901e-06,
+      "loss": 0.3024,
+      "step": 955
+    },
+    {
+      "epoch": 0.3191986644407346,
+      "grad_norm": 0.5706517543342288,
+      "learning_rate": 9.892187140516285e-06,
+      "loss": 0.3003,
+      "step": 956
+    },
+    {
+      "epoch": 0.31953255425709515,
+      "grad_norm": 0.6104732390995953,
+      "learning_rate": 9.891584496287212e-06,
+      "loss": 0.2981,
+      "step": 957
+    },
+    {
+      "epoch": 0.31986644407345577,
+      "grad_norm": 0.5964014540452306,
+      "learning_rate": 9.890980190901336e-06,
+      "loss": 0.3184,
+      "step": 958
+    },
+    {
+      "epoch": 0.32020033388981634,
+      "grad_norm": 0.5652487742638719,
+      "learning_rate": 9.890374224563872e-06,
+      "loss": 0.2983,
+      "step": 959
+    },
+    {
+      "epoch": 0.32053422370617696,
+      "grad_norm": 0.5667040545786675,
+      "learning_rate": 9.889766597480609e-06,
+      "loss": 0.3073,
+      "step": 960
+    },
+    {
+      "epoch": 0.3208681135225376,
+      "grad_norm": 0.5926877128756307,
+      "learning_rate": 9.88915730985789e-06,
+      "loss": 0.2955,
+      "step": 961
+    },
+    {
+      "epoch": 0.32120200333889815,
+      "grad_norm": 0.6032070607653965,
+      "learning_rate": 9.888546361902629e-06,
+      "loss": 0.3146,
+      "step": 962
+    },
+    {
+      "epoch": 0.32153589315525877,
+      "grad_norm": 0.5891570249179235,
+      "learning_rate": 9.887933753822297e-06,
+      "loss": 0.307,
+      "step": 963
+    },
+    {
+      "epoch": 0.3218697829716194,
+      "grad_norm": 0.6206763084746624,
+      "learning_rate": 9.887319485824935e-06,
+      "loss": 0.3051,
+      "step": 964
+    },
+    {
+      "epoch": 0.32220367278797996,
+      "grad_norm": 0.6031729038023127,
+      "learning_rate": 9.886703558119146e-06,
+      "loss": 0.3085,
+      "step": 965
+    },
+    {
+      "epoch": 0.3225375626043406,
+      "grad_norm": 0.6063193994969583,
+      "learning_rate": 9.886085970914095e-06,
+      "loss": 0.3113,
+      "step": 966
+    },
+    {
+      "epoch": 0.32287145242070114,
+      "grad_norm": 0.5735807125646658,
+      "learning_rate": 9.88546672441951e-06,
+      "loss": 0.2974,
+      "step": 967
+    },
+    {
+      "epoch": 0.32320534223706177,
+      "grad_norm": 0.5842576339100178,
+      "learning_rate": 9.884845818845685e-06,
+      "loss": 0.3031,
+      "step": 968
+    },
+    {
+      "epoch": 0.3235392320534224,
+      "grad_norm": 0.5935202457136572,
+      "learning_rate": 9.884223254403478e-06,
+      "loss": 0.3031,
+      "step": 969
+    },
+    {
+      "epoch": 0.32387312186978295,
+      "grad_norm": 0.5832450555235476,
+      "learning_rate": 9.883599031304305e-06,
+      "loss": 0.311,
+      "step": 970
+    },
+    {
+      "epoch": 0.3242070116861436,
+      "grad_norm": 0.6493755926163098,
+      "learning_rate": 9.882973149760151e-06,
+      "loss": 0.3107,
+      "step": 971
+    },
+    {
+      "epoch": 0.3245409015025042,
+      "grad_norm": 0.6309993230563669,
+      "learning_rate": 9.882345609983564e-06,
+      "loss": 0.3072,
+      "step": 972
+    },
+    {
+      "epoch": 0.32487479131886476,
+      "grad_norm": 0.6170278513360982,
+      "learning_rate": 9.88171641218765e-06,
+      "loss": 0.3248,
+      "step": 973
+    },
+    {
+      "epoch": 0.3252086811352254,
+      "grad_norm": 0.5924626597108921,
+      "learning_rate": 9.881085556586083e-06,
+      "loss": 0.3176,
+      "step": 974
+    },
+    {
+      "epoch": 0.32554257095158595,
+      "grad_norm": 0.66628977514577,
+      "learning_rate": 9.880453043393098e-06,
+      "loss": 0.3084,
+      "step": 975
+    },
+    {
+      "epoch": 0.3258764607679466,
+      "grad_norm": 0.6018416963542437,
+      "learning_rate": 9.879818872823492e-06,
+      "loss": 0.301,
+      "step": 976
+    },
+    {
+      "epoch": 0.3262103505843072,
+      "grad_norm": 0.6093400107023318,
+      "learning_rate": 9.879183045092628e-06,
+      "loss": 0.3009,
+      "step": 977
+    },
+    {
+      "epoch": 0.32654424040066776,
+      "grad_norm": 0.6087121208400584,
+      "learning_rate": 9.87854556041643e-06,
+      "loss": 0.3027,
+      "step": 978
+    },
+    {
+      "epoch": 0.3268781302170284,
+      "grad_norm": 0.5927360841427374,
+      "learning_rate": 9.877906419011384e-06,
+      "loss": 0.3109,
+      "step": 979
+    },
+    {
+      "epoch": 0.327212020033389,
+      "grad_norm": 0.5952475918533987,
+      "learning_rate": 9.877265621094537e-06,
+      "loss": 0.3027,
+      "step": 980
+    },
+    {
+      "epoch": 0.3275459098497496,
+      "grad_norm": 0.6426509744141223,
+      "learning_rate": 9.876623166883503e-06,
+      "loss": 0.3058,
+      "step": 981
+    },
+    {
+      "epoch": 0.3278797996661102,
+      "grad_norm": 0.5993740708234978,
+      "learning_rate": 9.875979056596457e-06,
+      "loss": 0.292,
+      "step": 982
+    },
+    {
+      "epoch": 0.32821368948247076,
+      "grad_norm": 0.5673862938992573,
+      "learning_rate": 9.875333290452131e-06,
+      "loss": 0.3025,
+      "step": 983
+    },
+    {
+      "epoch": 0.3285475792988314,
+      "grad_norm": 0.5607157383434699,
+      "learning_rate": 9.874685868669829e-06,
+      "loss": 0.3106,
+      "step": 984
+    },
+    {
+      "epoch": 0.328881469115192,
+      "grad_norm": 0.5979939795966953,
+      "learning_rate": 9.87403679146941e-06,
+      "loss": 0.3069,
+      "step": 985
+    },
+    {
+      "epoch": 0.32921535893155257,
+      "grad_norm": 0.6163053429658633,
+      "learning_rate": 9.873386059071294e-06,
+      "loss": 0.3116,
+      "step": 986
+    },
+    {
+      "epoch": 0.3295492487479132,
+      "grad_norm": 0.6182536458208553,
+      "learning_rate": 9.872733671696473e-06,
+      "loss": 0.3065,
+      "step": 987
+    },
+    {
+      "epoch": 0.3298831385642738,
+      "grad_norm": 0.6053682505168259,
+      "learning_rate": 9.872079629566492e-06,
+      "loss": 0.3225,
+      "step": 988
+    },
+    {
+      "epoch": 0.3302170283806344,
+      "grad_norm": 0.644942484213265,
+      "learning_rate": 9.871423932903456e-06,
+      "loss": 0.284,
+      "step": 989
+    },
+    {
+      "epoch": 0.330550918196995,
+      "grad_norm": 0.6122141580523129,
+      "learning_rate": 9.87076658193004e-06,
+      "loss": 0.3028,
+      "step": 990
+    },
+    {
+      "epoch": 0.33088480801335557,
+      "grad_norm": 0.5854182723409429,
+      "learning_rate": 9.870107576869478e-06,
+      "loss": 0.2993,
+      "step": 991
+    },
+    {
+      "epoch": 0.3312186978297162,
+      "grad_norm": 0.5889405115990973,
+      "learning_rate": 9.869446917945561e-06,
+      "loss": 0.3088,
+      "step": 992
+    },
+    {
+      "epoch": 0.3315525876460768,
+      "grad_norm": 0.6826612595504571,
+      "learning_rate": 9.868784605382649e-06,
+      "loss": 0.3062,
+      "step": 993
+    },
+    {
+      "epoch": 0.3318864774624374,
+      "grad_norm": 0.6122370509716591,
+      "learning_rate": 9.868120639405659e-06,
+      "loss": 0.3096,
+      "step": 994
+    },
+    {
+      "epoch": 0.332220367278798,
+      "grad_norm": 0.5777543017973802,
+      "learning_rate": 9.86745502024007e-06,
+      "loss": 0.3101,
+      "step": 995
+    },
+    {
+      "epoch": 0.3325542570951586,
+      "grad_norm": 0.6631884585142126,
+      "learning_rate": 9.86678774811192e-06,
+      "loss": 0.3054,
+      "step": 996
+    },
+    {
+      "epoch": 0.3328881469115192,
+      "grad_norm": 0.6019560209038383,
+      "learning_rate": 9.866118823247816e-06,
+      "loss": 0.3116,
+      "step": 997
+    },
+    {
+      "epoch": 0.3332220367278798,
+      "grad_norm": 0.5917756371672592,
+      "learning_rate": 9.865448245874918e-06,
+      "loss": 0.3055,
+      "step": 998
+    },
+    {
+      "epoch": 0.3335559265442404,
+      "grad_norm": 0.5983332690822838,
+      "learning_rate": 9.864776016220952e-06,
+      "loss": 0.3058,
+      "step": 999
+    },
+    {
+      "epoch": 0.333889816360601,
+      "grad_norm": 0.5656163190148125,
+      "learning_rate": 9.864102134514205e-06,
+      "loss": 0.2976,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3342237061769616,
+      "grad_norm": 0.6283918178915275,
+      "learning_rate": 9.86342660098352e-06,
+      "loss": 0.3186,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3345575959933222,
+      "grad_norm": 0.6015389637030385,
+      "learning_rate": 9.86274941585831e-06,
+      "loss": 0.2944,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3348914858096828,
+      "grad_norm": 0.6188552889823554,
+      "learning_rate": 9.862070579368538e-06,
+      "loss": 0.3074,
+      "step": 1003
+    },
+    {
+      "epoch": 0.33522537562604343,
+      "grad_norm": 0.6194194461599768,
+      "learning_rate": 9.861390091744738e-06,
+      "loss": 0.2995,
+      "step": 1004
+    },
+    {
+      "epoch": 0.335559265442404,
+      "grad_norm": 0.6038757787893098,
+      "learning_rate": 9.860707953217996e-06,
+      "loss": 0.2926,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3358931552587646,
+      "grad_norm": 0.6066622385174689,
+      "learning_rate": 9.860024164019968e-06,
+      "loss": 0.3093,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3362270450751252,
+      "grad_norm": 0.5825662330292989,
+      "learning_rate": 9.859338724382859e-06,
+      "loss": 0.3076,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3365609348914858,
+      "grad_norm": 0.5818964884599969,
+      "learning_rate": 9.858651634539444e-06,
+      "loss": 0.304,
+      "step": 1008
+    },
+    {
+      "epoch": 0.33689482470784643,
+      "grad_norm": 0.5855949018409367,
+      "learning_rate": 9.857962894723056e-06,
+      "loss": 0.3093,
+      "step": 1009
+    },
+    {
+      "epoch": 0.337228714524207,
+      "grad_norm": 0.5968136759666197,
+      "learning_rate": 9.857272505167585e-06,
+      "loss": 0.3011,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3375626043405676,
+      "grad_norm": 0.5946203981819529,
+      "learning_rate": 9.856580466107487e-06,
+      "loss": 0.3122,
+      "step": 1011
+    },
+    {
+      "epoch": 0.33789649415692824,
+      "grad_norm": 0.5837319047682951,
+      "learning_rate": 9.855886777777772e-06,
+      "loss": 0.3075,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3382303839732888,
+      "grad_norm": 0.618805619921558,
+      "learning_rate": 9.855191440414014e-06,
+      "loss": 0.3091,
+      "step": 1013
+    },
+    {
+      "epoch": 0.33856427378964943,
+      "grad_norm": 0.5797138534509052,
+      "learning_rate": 9.854494454252345e-06,
+      "loss": 0.3035,
+      "step": 1014
+    },
+    {
+      "epoch": 0.33889816360601,
+      "grad_norm": 0.6000042687824859,
+      "learning_rate": 9.85379581952946e-06,
+      "loss": 0.3112,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3392320534223706,
+      "grad_norm": 0.6165247710435977,
+      "learning_rate": 9.85309553648261e-06,
+      "loss": 0.3128,
+      "step": 1016
+    },
+    {
+      "epoch": 0.33956594323873124,
+      "grad_norm": 0.5979065194634363,
+      "learning_rate": 9.85239360534961e-06,
+      "loss": 0.3071,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3398998330550918,
+      "grad_norm": 0.5857311476841967,
+      "learning_rate": 9.851690026368829e-06,
+      "loss": 0.2891,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3402337228714524,
+      "grad_norm": 0.6238906364696216,
+      "learning_rate": 9.850984799779199e-06,
+      "loss": 0.313,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34056761268781305,
+      "grad_norm": 0.59824988042071,
+      "learning_rate": 9.850277925820215e-06,
+      "loss": 0.3104,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3409015025041736,
+      "grad_norm": 0.6099320377748151,
+      "learning_rate": 9.849569404731922e-06,
+      "loss": 0.3145,
+      "step": 1021
+    },
+    {
+      "epoch": 0.34123539232053424,
+      "grad_norm": 0.5712706624189405,
+      "learning_rate": 9.848859236754936e-06,
+      "loss": 0.2991,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3415692821368948,
+      "grad_norm": 0.6482246287705586,
+      "learning_rate": 9.848147422130421e-06,
+      "loss": 0.319,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3419031719532554,
+      "grad_norm": 0.595417982429453,
+      "learning_rate": 9.84743396110011e-06,
+      "loss": 0.3021,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34223706176961605,
+      "grad_norm": 0.6351060617497407,
+      "learning_rate": 9.846718853906289e-06,
+      "loss": 0.3103,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3425709515859766,
+      "grad_norm": 0.5875936620212611,
+      "learning_rate": 9.846002100791802e-06,
+      "loss": 0.3033,
+      "step": 1026
+    },
+    {
+      "epoch": 0.34290484140233723,
+      "grad_norm": 0.5956756929993205,
+      "learning_rate": 9.84528370200006e-06,
+      "loss": 0.31,
+      "step": 1027
+    },
+    {
+      "epoch": 0.34323873121869786,
+      "grad_norm": 0.5979736952564252,
+      "learning_rate": 9.844563657775024e-06,
+      "loss": 0.3128,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3435726210350584,
+      "grad_norm": 0.6206316607305022,
+      "learning_rate": 9.843841968361217e-06,
+      "loss": 0.3199,
+      "step": 1029
+    },
+    {
+      "epoch": 0.34390651085141904,
+      "grad_norm": 0.5706712699213329,
+      "learning_rate": 9.843118634003724e-06,
+      "loss": 0.3257,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3442404006677796,
+      "grad_norm": 0.5686645885449831,
+      "learning_rate": 9.84239365494818e-06,
+      "loss": 0.3124,
+      "step": 1031
+    },
+    {
+      "epoch": 0.34457429048414023,
+      "grad_norm": 0.6229197814000341,
+      "learning_rate": 9.841667031440791e-06,
+      "loss": 0.307,
+      "step": 1032
+    },
+    {
+      "epoch": 0.34490818030050086,
+      "grad_norm": 0.6142235517521829,
+      "learning_rate": 9.840938763728312e-06,
+      "loss": 0.3207,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3452420701168614,
+      "grad_norm": 0.5867651759196412,
+      "learning_rate": 9.840208852058055e-06,
+      "loss": 0.3044,
+      "step": 1034
+    },
+    {
+      "epoch": 0.34557595993322204,
+      "grad_norm": 0.5683021403087946,
+      "learning_rate": 9.8394772966779e-06,
+      "loss": 0.2968,
+      "step": 1035
+    },
+    {
+      "epoch": 0.34590984974958267,
+      "grad_norm": 0.6173354919390572,
+      "learning_rate": 9.838744097836276e-06,
+      "loss": 0.3156,
+      "step": 1036
+    },
+    {
+      "epoch": 0.34624373956594323,
+      "grad_norm": 0.5642232485981289,
+      "learning_rate": 9.838009255782176e-06,
+      "loss": 0.3009,
+      "step": 1037
+    },
+    {
+      "epoch": 0.34657762938230385,
+      "grad_norm": 0.5903395802542488,
+      "learning_rate": 9.837272770765143e-06,
+      "loss": 0.2948,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3469115191986644,
+      "grad_norm": 0.6090387429533164,
+      "learning_rate": 9.836534643035292e-06,
+      "loss": 0.2989,
+      "step": 1039
+    },
+    {
+      "epoch": 0.34724540901502504,
+      "grad_norm": 0.6113717622585859,
+      "learning_rate": 9.835794872843281e-06,
+      "loss": 0.3159,
+      "step": 1040
+    },
+    {
+      "epoch": 0.34757929883138566,
+      "grad_norm": 0.5939071341355235,
+      "learning_rate": 9.835053460440334e-06,
+      "loss": 0.3041,
+      "step": 1041
+    },
+    {
+      "epoch": 0.34791318864774623,
+      "grad_norm": 0.609172802036514,
+      "learning_rate": 9.83431040607823e-06,
+      "loss": 0.296,
+      "step": 1042
+    },
+    {
+      "epoch": 0.34824707846410685,
+      "grad_norm": 0.6001241786440387,
+      "learning_rate": 9.833565710009308e-06,
+      "loss": 0.3061,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3485809682804674,
+      "grad_norm": 0.6080910571289091,
+      "learning_rate": 9.832819372486463e-06,
+      "loss": 0.2928,
+      "step": 1044
+    },
+    {
+      "epoch": 0.34891485809682804,
+      "grad_norm": 0.634908749880765,
+      "learning_rate": 9.832071393763145e-06,
+      "loss": 0.3183,
+      "step": 1045
+    },
+    {
+      "epoch": 0.34924874791318866,
+      "grad_norm": 0.573386788493356,
+      "learning_rate": 9.831321774093366e-06,
+      "loss": 0.2956,
+      "step": 1046
+    },
+    {
+      "epoch": 0.34958263772954923,
+      "grad_norm": 0.5960598307962216,
+      "learning_rate": 9.830570513731691e-06,
+      "loss": 0.3083,
+      "step": 1047
+    },
+    {
+      "epoch": 0.34991652754590985,
+      "grad_norm": 0.5943094623751214,
+      "learning_rate": 9.829817612933248e-06,
+      "loss": 0.3103,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35025041736227047,
+      "grad_norm": 0.5715508042021707,
+      "learning_rate": 9.829063071953715e-06,
+      "loss": 0.3163,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35058430717863104,
+      "grad_norm": 0.5876561399047894,
+      "learning_rate": 9.82830689104933e-06,
+      "loss": 0.2995,
+      "step": 1050
+    },
+    {
+      "epoch": 0.35091819699499166,
+      "grad_norm": 0.5704930141012101,
+      "learning_rate": 9.82754907047689e-06,
+      "loss": 0.3183,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3512520868113522,
+      "grad_norm": 0.5767198543952382,
+      "learning_rate": 9.826789610493744e-06,
+      "loss": 0.3085,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35158597662771285,
+      "grad_norm": 0.5771700590681086,
+      "learning_rate": 9.826028511357804e-06,
+      "loss": 0.2936,
+      "step": 1053
+    },
+    {
+      "epoch": 0.35191986644407347,
+      "grad_norm": 0.611939219827137,
+      "learning_rate": 9.825265773327533e-06,
+      "loss": 0.3127,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35225375626043404,
+      "grad_norm": 0.5801918168715917,
+      "learning_rate": 9.824501396661955e-06,
+      "loss": 0.3041,
+      "step": 1055
+    },
+    {
+      "epoch": 0.35258764607679466,
+      "grad_norm": 0.644357330728477,
+      "learning_rate": 9.823735381620645e-06,
+      "loss": 0.3028,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3529215358931553,
+      "grad_norm": 0.6104603722103434,
+      "learning_rate": 9.822967728463741e-06,
+      "loss": 0.3018,
+      "step": 1057
+    },
+    {
+      "epoch": 0.35325542570951585,
+      "grad_norm": 0.6009481811099752,
+      "learning_rate": 9.822198437451933e-06,
+      "loss": 0.3141,
+      "step": 1058
+    },
+    {
+      "epoch": 0.35358931552587647,
+      "grad_norm": 0.5889008153853024,
+      "learning_rate": 9.821427508846469e-06,
+      "loss": 0.3186,
+      "step": 1059
+    },
+    {
+      "epoch": 0.35392320534223703,
+      "grad_norm": 0.6228432208920752,
+      "learning_rate": 9.820654942909152e-06,
+      "loss": 0.3051,
+      "step": 1060
+    },
+    {
+      "epoch": 0.35425709515859766,
+      "grad_norm": 0.6118791777147035,
+      "learning_rate": 9.819880739902339e-06,
+      "loss": 0.3104,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3545909849749583,
+      "grad_norm": 0.5757261935100461,
+      "learning_rate": 9.819104900088948e-06,
+      "loss": 0.3131,
+      "step": 1062
+    },
+    {
+      "epoch": 0.35492487479131885,
+      "grad_norm": 0.5786866634972722,
+      "learning_rate": 9.81832742373245e-06,
+      "loss": 0.3069,
+      "step": 1063
+    },
+    {
+      "epoch": 0.35525876460767947,
+      "grad_norm": 0.6231114294698069,
+      "learning_rate": 9.81754831109687e-06,
+      "loss": 0.3074,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3555926544240401,
+      "grad_norm": 0.5866992745278014,
+      "learning_rate": 9.816767562446794e-06,
+      "loss": 0.2977,
+      "step": 1065
+    },
+    {
+      "epoch": 0.35592654424040066,
+      "grad_norm": 0.5948982616848886,
+      "learning_rate": 9.815985178047357e-06,
+      "loss": 0.3046,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3562604340567613,
+      "grad_norm": 0.5624019518473241,
+      "learning_rate": 9.815201158164253e-06,
+      "loss": 0.2929,
+      "step": 1067
+    },
+    {
+      "epoch": 0.35659432387312184,
+      "grad_norm": 0.5718543748818863,
+      "learning_rate": 9.814415503063735e-06,
+      "loss": 0.3025,
+      "step": 1068
+    },
+    {
+      "epoch": 0.35692821368948247,
+      "grad_norm": 0.6179360024921368,
+      "learning_rate": 9.813628213012603e-06,
+      "loss": 0.3022,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3572621035058431,
+      "grad_norm": 0.5704438426814514,
+      "learning_rate": 9.812839288278218e-06,
+      "loss": 0.2925,
+      "step": 1070
+    },
+    {
+      "epoch": 0.35759599332220365,
+      "grad_norm": 0.5704787302858472,
+      "learning_rate": 9.812048729128494e-06,
+      "loss": 0.3006,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3579298831385643,
+      "grad_norm": 0.6043999670824531,
+      "learning_rate": 9.811256535831904e-06,
+      "loss": 0.2923,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3582637729549249,
+      "grad_norm": 0.5672599478799503,
+      "learning_rate": 9.810462708657467e-06,
+      "loss": 0.3075,
+      "step": 1073
+    },
+    {
+      "epoch": 0.35859766277128546,
+      "grad_norm": 0.5745967129549346,
+      "learning_rate": 9.809667247874768e-06,
+      "loss": 0.3038,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3589315525876461,
+      "grad_norm": 0.5375744223995321,
+      "learning_rate": 9.80887015375394e-06,
+      "loss": 0.299,
+      "step": 1075
+    },
+    {
+      "epoch": 0.35926544240400665,
+      "grad_norm": 0.5926958834811291,
+      "learning_rate": 9.808071426565671e-06,
+      "loss": 0.3117,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3595993322203673,
+      "grad_norm": 0.5946850871403864,
+      "learning_rate": 9.807271066581203e-06,
+      "loss": 0.3031,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3599332220367279,
+      "grad_norm": 0.580690658545827,
+      "learning_rate": 9.806469074072339e-06,
+      "loss": 0.3016,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36026711185308846,
+      "grad_norm": 0.5934494278971795,
+      "learning_rate": 9.805665449311428e-06,
+      "loss": 0.3007,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3606010016694491,
+      "grad_norm": 0.5751288605897378,
+      "learning_rate": 9.804860192571376e-06,
+      "loss": 0.2992,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3609348914858097,
+      "grad_norm": 0.5612150333261724,
+      "learning_rate": 9.804053304125646e-06,
+      "loss": 0.2998,
+      "step": 1081
+    },
+    {
+      "epoch": 0.36126878130217027,
+      "grad_norm": 0.6446230181337367,
+      "learning_rate": 9.803244784248253e-06,
+      "loss": 0.3149,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3616026711185309,
+      "grad_norm": 0.578803487316165,
+      "learning_rate": 9.802434633213767e-06,
+      "loss": 0.3081,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36193656093489146,
+      "grad_norm": 0.5676971839485909,
+      "learning_rate": 9.801622851297309e-06,
+      "loss": 0.3022,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3622704507512521,
+      "grad_norm": 0.5657589892902519,
+      "learning_rate": 9.800809438774557e-06,
+      "loss": 0.2946,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3626043405676127,
+      "grad_norm": 0.557032899433067,
+      "learning_rate": 9.79999439592174e-06,
+      "loss": 0.2995,
+      "step": 1086
+    },
+    {
+      "epoch": 0.36293823038397327,
+      "grad_norm": 0.5988354592779646,
+      "learning_rate": 9.799177723015646e-06,
+      "loss": 0.3113,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3632721202003339,
+      "grad_norm": 0.5417892685274236,
+      "learning_rate": 9.798359420333609e-06,
+      "loss": 0.3072,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3636060100166945,
+      "grad_norm": 0.5903894462779202,
+      "learning_rate": 9.797539488153525e-06,
+      "loss": 0.3142,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3639398998330551,
+      "grad_norm": 0.5666238825733495,
+      "learning_rate": 9.796717926753832e-06,
+      "loss": 0.2915,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3642737896494157,
+      "grad_norm": 0.5743241729513006,
+      "learning_rate": 9.795894736413535e-06,
+      "loss": 0.2881,
+      "step": 1091
+    },
+    {
+      "epoch": 0.36460767946577627,
+      "grad_norm": 0.6032988009827134,
+      "learning_rate": 9.79506991741218e-06,
+      "loss": 0.2998,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3649415692821369,
+      "grad_norm": 0.5452234493601275,
+      "learning_rate": 9.794243470029875e-06,
+      "loss": 0.2933,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3652754590984975,
+      "grad_norm": 0.5478450514511412,
+      "learning_rate": 9.793415394547274e-06,
+      "loss": 0.2994,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3656093489148581,
+      "grad_norm": 0.658858751643103,
+      "learning_rate": 9.792585691245591e-06,
+      "loss": 0.3038,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3659432387312187,
+      "grad_norm": 0.6064694964760153,
+      "learning_rate": 9.791754360406584e-06,
+      "loss": 0.3107,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3662771285475793,
+      "grad_norm": 0.5956847582700077,
+      "learning_rate": 9.790921402312573e-06,
+      "loss": 0.3155,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3666110183639399,
+      "grad_norm": 0.5518422694322338,
+      "learning_rate": 9.790086817246425e-06,
+      "loss": 0.2884,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3669449081803005,
+      "grad_norm": 0.598263939591925,
+      "learning_rate": 9.789250605491558e-06,
+      "loss": 0.3036,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3672787979966611,
+      "grad_norm": 0.5900605663463003,
+      "learning_rate": 9.78841276733195e-06,
+      "loss": 0.2921,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3676126878130217,
+      "grad_norm": 0.5803864203805427,
+      "learning_rate": 9.787573303052123e-06,
+      "loss": 0.3006,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3679465776293823,
+      "grad_norm": 0.5773282321566462,
+      "learning_rate": 9.786732212937156e-06,
+      "loss": 0.3129,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3682804674457429,
+      "grad_norm": 0.5630935778276642,
+      "learning_rate": 9.785889497272678e-06,
+      "loss": 0.3033,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3686143572621035,
+      "grad_norm": 0.5746057950770693,
+      "learning_rate": 9.785045156344872e-06,
+      "loss": 0.3086,
+      "step": 1104
+    },
+    {
+      "epoch": 0.36894824707846413,
+      "grad_norm": 0.6111182652943348,
+      "learning_rate": 9.784199190440473e-06,
+      "loss": 0.3041,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3692821368948247,
+      "grad_norm": 0.5667081077429498,
+      "learning_rate": 9.783351599846762e-06,
+      "loss": 0.3088,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3696160267111853,
+      "grad_norm": 0.5851441550287735,
+      "learning_rate": 9.782502384851582e-06,
+      "loss": 0.3031,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3699499165275459,
+      "grad_norm": 0.5796777261115196,
+      "learning_rate": 9.781651545743319e-06,
+      "loss": 0.3099,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3702838063439065,
+      "grad_norm": 0.5968084439337071,
+      "learning_rate": 9.780799082810912e-06,
+      "loss": 0.3012,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37061769616026713,
+      "grad_norm": 0.6194858954293063,
+      "learning_rate": 9.779944996343858e-06,
+      "loss": 0.3144,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3709515859766277,
+      "grad_norm": 0.5822287161201267,
+      "learning_rate": 9.779089286632194e-06,
+      "loss": 0.3022,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3712854757929883,
+      "grad_norm": 0.5629870789422393,
+      "learning_rate": 9.778231953966519e-06,
+      "loss": 0.2939,
+      "step": 1112
+    },
+    {
+      "epoch": 0.37161936560934894,
+      "grad_norm": 0.5656540722923922,
+      "learning_rate": 9.77737299863798e-06,
+      "loss": 0.2939,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3719532554257095,
+      "grad_norm": 0.5515238361692589,
+      "learning_rate": 9.77651242093827e-06,
+      "loss": 0.2953,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3722871452420701,
+      "grad_norm": 0.5826321561136557,
+      "learning_rate": 9.775650221159635e-06,
+      "loss": 0.3056,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3726210350584307,
+      "grad_norm": 0.5571815608366419,
+      "learning_rate": 9.774786399594881e-06,
+      "loss": 0.2924,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3729549248747913,
+      "grad_norm": 0.6070754196943753,
+      "learning_rate": 9.773920956537351e-06,
+      "loss": 0.3052,
+      "step": 1117
+    },
+    {
+      "epoch": 0.37328881469115194,
+      "grad_norm": 0.555373867039528,
+      "learning_rate": 9.773053892280949e-06,
+      "loss": 0.2981,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3736227045075125,
+      "grad_norm": 0.5748439833206649,
+      "learning_rate": 9.772185207120122e-06,
+      "loss": 0.3104,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3739565943238731,
+      "grad_norm": 0.5915991013084175,
+      "learning_rate": 9.771314901349873e-06,
+      "loss": 0.3019,
+      "step": 1120
+    },
+    {
+      "epoch": 0.37429048414023375,
+      "grad_norm": 0.5824381297649524,
+      "learning_rate": 9.770442975265753e-06,
+      "loss": 0.308,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3746243739565943,
+      "grad_norm": 0.6153266756254893,
+      "learning_rate": 9.769569429163863e-06,
+      "loss": 0.2826,
+      "step": 1122
+    },
+    {
+      "epoch": 0.37495826377295494,
+      "grad_norm": 0.5639029683560274,
+      "learning_rate": 9.768694263340856e-06,
+      "loss": 0.3079,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3752921535893155,
+      "grad_norm": 0.5903801409366664,
+      "learning_rate": 9.767817478093933e-06,
+      "loss": 0.2975,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3756260434056761,
+      "grad_norm": 0.5644747459962814,
+      "learning_rate": 9.766939073720846e-06,
+      "loss": 0.2984,
+      "step": 1125
+    },
+    {
+      "epoch": 0.37595993322203675,
+      "grad_norm": 0.5873693976649459,
+      "learning_rate": 9.766059050519897e-06,
+      "loss": 0.2976,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3762938230383973,
+      "grad_norm": 0.6042052612731528,
+      "learning_rate": 9.765177408789935e-06,
+      "loss": 0.3056,
+      "step": 1127
+    },
+    {
+      "epoch": 0.37662771285475793,
+      "grad_norm": 0.5621265813316134,
+      "learning_rate": 9.764294148830363e-06,
+      "loss": 0.3031,
+      "step": 1128
+    },
+    {
+      "epoch": 0.37696160267111856,
+      "grad_norm": 0.5529008693243657,
+      "learning_rate": 9.763409270941132e-06,
+      "loss": 0.3025,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3772954924874791,
+      "grad_norm": 0.6522462274014118,
+      "learning_rate": 9.762522775422741e-06,
+      "loss": 0.3044,
+      "step": 1130
+    },
+    {
+      "epoch": 0.37762938230383974,
+      "grad_norm": 0.599875803229418,
+      "learning_rate": 9.76163466257624e-06,
+      "loss": 0.3059,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3779632721202003,
+      "grad_norm": 0.5672813749128086,
+      "learning_rate": 9.760744932703225e-06,
+      "loss": 0.3091,
+      "step": 1132
+    },
+    {
+      "epoch": 0.37829716193656093,
+      "grad_norm": 0.6097182030192447,
+      "learning_rate": 9.759853586105848e-06,
+      "loss": 0.3072,
+      "step": 1133
+    },
+    {
+      "epoch": 0.37863105175292155,
+      "grad_norm": 0.6431794445636085,
+      "learning_rate": 9.758960623086802e-06,
+      "loss": 0.3085,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3789649415692821,
+      "grad_norm": 0.5804240962871872,
+      "learning_rate": 9.758066043949335e-06,
+      "loss": 0.3069,
+      "step": 1135
+    },
+    {
+      "epoch": 0.37929883138564274,
+      "grad_norm": 0.568618635738777,
+      "learning_rate": 9.75716984899724e-06,
+      "loss": 0.3028,
+      "step": 1136
+    },
+    {
+      "epoch": 0.37963272120200336,
+      "grad_norm": 0.5513998128980915,
+      "learning_rate": 9.75627203853486e-06,
+      "loss": 0.3031,
+      "step": 1137
+    },
+    {
+      "epoch": 0.37996661101836393,
+      "grad_norm": 0.5883358110738967,
+      "learning_rate": 9.755372612867087e-06,
+      "loss": 0.3026,
+      "step": 1138
+    },
+    {
+      "epoch": 0.38030050083472455,
+      "grad_norm": 0.5935585650403364,
+      "learning_rate": 9.754471572299363e-06,
+      "loss": 0.3047,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3806343906510851,
+      "grad_norm": 0.5442289157737286,
+      "learning_rate": 9.753568917137675e-06,
+      "loss": 0.2923,
+      "step": 1140
+    },
+    {
+      "epoch": 0.38096828046744574,
+      "grad_norm": 0.6117897146578629,
+      "learning_rate": 9.752664647688562e-06,
+      "loss": 0.3084,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38130217028380636,
+      "grad_norm": 0.5634492676030709,
+      "learning_rate": 9.751758764259104e-06,
+      "loss": 0.297,
+      "step": 1142
+    },
+    {
+      "epoch": 0.38163606010016693,
+      "grad_norm": 0.5786835148858668,
+      "learning_rate": 9.75085126715694e-06,
+      "loss": 0.3033,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38196994991652755,
+      "grad_norm": 0.6098753564044375,
+      "learning_rate": 9.749942156690251e-06,
+      "loss": 0.2985,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3823038397328882,
+      "grad_norm": 0.5620452228851406,
+      "learning_rate": 9.749031433167761e-06,
+      "loss": 0.299,
+      "step": 1145
+    },
+    {
+      "epoch": 0.38263772954924874,
+      "grad_norm": 0.5790603941863683,
+      "learning_rate": 9.748119096898754e-06,
+      "loss": 0.3094,
+      "step": 1146
+    },
+    {
+      "epoch": 0.38297161936560936,
+      "grad_norm": 0.5771622054534739,
+      "learning_rate": 9.747205148193048e-06,
+      "loss": 0.2967,
+      "step": 1147
+    },
+    {
+      "epoch": 0.38330550918196993,
+      "grad_norm": 0.6547609634138749,
+      "learning_rate": 9.746289587361021e-06,
+      "loss": 0.3,
+      "step": 1148
+    },
+    {
+      "epoch": 0.38363939899833055,
+      "grad_norm": 0.6035510481894482,
+      "learning_rate": 9.74537241471359e-06,
+      "loss": 0.302,
+      "step": 1149
+    },
+    {
+      "epoch": 0.38397328881469117,
+      "grad_norm": 0.5922293517496449,
+      "learning_rate": 9.74445363056222e-06,
+      "loss": 0.3025,
+      "step": 1150
+    },
+    {
+      "epoch": 0.38430717863105174,
+      "grad_norm": 0.5906026852627928,
+      "learning_rate": 9.743533235218931e-06,
+      "loss": 0.2935,
+      "step": 1151
+    },
+    {
+      "epoch": 0.38464106844741236,
+      "grad_norm": 0.5501304982391326,
+      "learning_rate": 9.74261122899628e-06,
+      "loss": 0.2933,
+      "step": 1152
+    },
+    {
+      "epoch": 0.384974958263773,
+      "grad_norm": 0.5450497431582663,
+      "learning_rate": 9.741687612207375e-06,
+      "loss": 0.2961,
+      "step": 1153
+    },
+    {
+      "epoch": 0.38530884808013355,
+      "grad_norm": 0.5722983375084485,
+      "learning_rate": 9.740762385165874e-06,
+      "loss": 0.3082,
+      "step": 1154
+    },
+    {
+      "epoch": 0.38564273789649417,
+      "grad_norm": 0.5386792281863209,
+      "learning_rate": 9.739835548185978e-06,
+      "loss": 0.3197,
+      "step": 1155
+    },
+    {
+      "epoch": 0.38597662771285474,
+      "grad_norm": 0.594810967437067,
+      "learning_rate": 9.738907101582439e-06,
+      "loss": 0.3069,
+      "step": 1156
+    },
+    {
+      "epoch": 0.38631051752921536,
+      "grad_norm": 0.566850425549363,
+      "learning_rate": 9.737977045670549e-06,
+      "loss": 0.2889,
+      "step": 1157
+    },
+    {
+      "epoch": 0.386644407345576,
+      "grad_norm": 0.5906395245345933,
+      "learning_rate": 9.73704538076615e-06,
+      "loss": 0.3092,
+      "step": 1158
+    },
+    {
+      "epoch": 0.38697829716193655,
+      "grad_norm": 0.5884592783863184,
+      "learning_rate": 9.736112107185634e-06,
+      "loss": 0.301,
+      "step": 1159
+    },
+    {
+      "epoch": 0.38731218697829717,
+      "grad_norm": 0.5625745058654097,
+      "learning_rate": 9.73517722524593e-06,
+      "loss": 0.3046,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3876460767946578,
+      "grad_norm": 0.5982507223357283,
+      "learning_rate": 9.734240735264526e-06,
+      "loss": 0.3114,
+      "step": 1161
+    },
+    {
+      "epoch": 0.38797996661101836,
+      "grad_norm": 0.5609443343064003,
+      "learning_rate": 9.733302637559447e-06,
+      "loss": 0.291,
+      "step": 1162
+    },
+    {
+      "epoch": 0.388313856427379,
+      "grad_norm": 0.5660082718921481,
+      "learning_rate": 9.732362932449264e-06,
+      "loss": 0.3022,
+      "step": 1163
+    },
+    {
+      "epoch": 0.38864774624373954,
+      "grad_norm": 0.5594243768353703,
+      "learning_rate": 9.731421620253097e-06,
+      "loss": 0.2928,
+      "step": 1164
+    },
+    {
+      "epoch": 0.38898163606010017,
+      "grad_norm": 0.5617511973419447,
+      "learning_rate": 9.730478701290611e-06,
+      "loss": 0.3038,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3893155258764608,
+      "grad_norm": 0.5764939179770542,
+      "learning_rate": 9.729534175882016e-06,
+      "loss": 0.2952,
+      "step": 1166
+    },
+    {
+      "epoch": 0.38964941569282135,
+      "grad_norm": 0.5319977344870817,
+      "learning_rate": 9.72858804434807e-06,
+      "loss": 0.2961,
+      "step": 1167
+    },
+    {
+      "epoch": 0.389983305509182,
+      "grad_norm": 0.5515799410842278,
+      "learning_rate": 9.72764030701007e-06,
+      "loss": 0.3028,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3903171953255426,
+      "grad_norm": 0.5658137343509809,
+      "learning_rate": 9.726690964189866e-06,
+      "loss": 0.2971,
+      "step": 1169
+    },
+    {
+      "epoch": 0.39065108514190316,
+      "grad_norm": 0.551276289082696,
+      "learning_rate": 9.725740016209851e-06,
+      "loss": 0.2924,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3909849749582638,
+      "grad_norm": 0.5690741135728032,
+      "learning_rate": 9.72478746339296e-06,
+      "loss": 0.302,
+      "step": 1171
+    },
+    {
+      "epoch": 0.39131886477462435,
+      "grad_norm": 0.5714897120319088,
+      "learning_rate": 9.723833306062674e-06,
+      "loss": 0.3062,
+      "step": 1172
+    },
+    {
+      "epoch": 0.391652754590985,
+      "grad_norm": 0.5504016748354256,
+      "learning_rate": 9.722877544543023e-06,
+      "loss": 0.2968,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3919866444073456,
+      "grad_norm": 0.6188547793768102,
+      "learning_rate": 9.721920179158575e-06,
+      "loss": 0.3014,
+      "step": 1174
+    },
+    {
+      "epoch": 0.39232053422370616,
+      "grad_norm": 0.5548869008856522,
+      "learning_rate": 9.72096121023445e-06,
+      "loss": 0.3022,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3926544240400668,
+      "grad_norm": 0.6112598316360413,
+      "learning_rate": 9.720000638096306e-06,
+      "loss": 0.2982,
+      "step": 1176
+    },
+    {
+      "epoch": 0.39298831385642735,
+      "grad_norm": 0.5471057755061605,
+      "learning_rate": 9.71903846307035e-06,
+      "loss": 0.3089,
+      "step": 1177
+    },
+    {
+      "epoch": 0.393322203672788,
+      "grad_norm": 0.5259069397440171,
+      "learning_rate": 9.71807468548333e-06,
+      "loss": 0.3007,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3936560934891486,
+      "grad_norm": 0.5580448498965032,
+      "learning_rate": 9.717109305662542e-06,
+      "loss": 0.2968,
+      "step": 1179
+    },
+    {
+      "epoch": 0.39398998330550916,
+      "grad_norm": 0.5447788072705338,
+      "learning_rate": 9.716142323935823e-06,
+      "loss": 0.3004,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3943238731218698,
+      "grad_norm": 0.6206311893574655,
+      "learning_rate": 9.715173740631554e-06,
+      "loss": 0.2991,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3946577629382304,
+      "grad_norm": 0.5747782182416635,
+      "learning_rate": 9.714203556078662e-06,
+      "loss": 0.3006,
+      "step": 1182
+    },
+    {
+      "epoch": 0.39499165275459097,
+      "grad_norm": 0.5767494643186402,
+      "learning_rate": 9.713231770606616e-06,
+      "loss": 0.2985,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3953255425709516,
+      "grad_norm": 0.5664874370824636,
+      "learning_rate": 9.712258384545432e-06,
+      "loss": 0.2857,
+      "step": 1184
+    },
+    {
+      "epoch": 0.39565943238731216,
+      "grad_norm": 0.5757569265337157,
+      "learning_rate": 9.711283398225664e-06,
+      "loss": 0.3097,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3959933222036728,
+      "grad_norm": 0.5811148182482306,
+      "learning_rate": 9.710306811978413e-06,
+      "loss": 0.2923,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3963272120200334,
+      "grad_norm": 0.5917927025601647,
+      "learning_rate": 9.709328626135322e-06,
+      "loss": 0.312,
+      "step": 1187
+    },
+    {
+      "epoch": 0.39666110183639397,
+      "grad_norm": 0.5661538060309003,
+      "learning_rate": 9.70834884102858e-06,
+      "loss": 0.3049,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3969949916527546,
+      "grad_norm": 0.5966296401637186,
+      "learning_rate": 9.707367456990913e-06,
+      "loss": 0.3104,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3973288814691152,
+      "grad_norm": 0.5825750832113478,
+      "learning_rate": 9.7063844743556e-06,
+      "loss": 0.2971,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3976627712854758,
+      "grad_norm": 0.606245976200284,
+      "learning_rate": 9.705399893456449e-06,
+      "loss": 0.2998,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3979966611018364,
+      "grad_norm": 0.5662559780721085,
+      "learning_rate": 9.704413714627825e-06,
+      "loss": 0.3016,
+      "step": 1192
+    },
+    {
+      "epoch": 0.39833055091819697,
+      "grad_norm": 0.5645942085565784,
+      "learning_rate": 9.703425938204627e-06,
+      "loss": 0.3017,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3986644407345576,
+      "grad_norm": 0.5668697330028396,
+      "learning_rate": 9.7024365645223e-06,
+      "loss": 0.2915,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3989983305509182,
+      "grad_norm": 0.5840210118327054,
+      "learning_rate": 9.701445593916827e-06,
+      "loss": 0.281,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3993322203672788,
+      "grad_norm": 0.5607300966855532,
+      "learning_rate": 9.70045302672474e-06,
+      "loss": 0.2947,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3996661101836394,
+      "grad_norm": 0.5683912970656497,
+      "learning_rate": 9.699458863283106e-06,
+      "loss": 0.2938,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.593175520443169,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.3033,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4003338898163606,
+      "grad_norm": 0.5809798489011785,
+      "learning_rate": 9.697465749002202e-06,
+      "loss": 0.2957,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4006677796327212,
+      "grad_norm": 0.5555976397647713,
+      "learning_rate": 9.69646679883978e-06,
+      "loss": 0.3071,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4010016694490818,
+      "grad_norm": 0.5252054712633584,
+      "learning_rate": 9.695466253781514e-06,
+      "loss": 0.2855,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4013355592654424,
+      "grad_norm": 0.5881406863457443,
+      "learning_rate": 9.694464114167185e-06,
+      "loss": 0.2968,
+      "step": 1202
+    },
+    {
+      "epoch": 0.401669449081803,
+      "grad_norm": 0.569741808596393,
+      "learning_rate": 9.693460380337118e-06,
+      "loss": 0.3052,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4020033388981636,
+      "grad_norm": 0.5882553603259709,
+      "learning_rate": 9.69245505263217e-06,
+      "loss": 0.3052,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4023372287145242,
+      "grad_norm": 0.541706633128766,
+      "learning_rate": 9.69144813139375e-06,
+      "loss": 0.283,
+      "step": 1205
+    },
+    {
+      "epoch": 0.40267111853088483,
+      "grad_norm": 0.5728532450327456,
+      "learning_rate": 9.6904396169638e-06,
+      "loss": 0.3043,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4030050083472454,
+      "grad_norm": 0.5428552975600305,
+      "learning_rate": 9.689429509684808e-06,
+      "loss": 0.2977,
+      "step": 1207
+    },
+    {
+      "epoch": 0.403338898163606,
+      "grad_norm": 0.5639154052736126,
+      "learning_rate": 9.688417809899799e-06,
+      "loss": 0.3011,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4036727879799666,
+      "grad_norm": 0.5613499296949342,
+      "learning_rate": 9.687404517952345e-06,
+      "loss": 0.3005,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4040066777963272,
+      "grad_norm": 0.5591965248202864,
+      "learning_rate": 9.68638963418655e-06,
+      "loss": 0.2977,
+      "step": 1210
+    },
+    {
+      "epoch": 0.40434056761268783,
+      "grad_norm": 0.5567837257706638,
+      "learning_rate": 9.685373158947067e-06,
+      "loss": 0.2844,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4046744574290484,
+      "grad_norm": 0.557847952525362,
+      "learning_rate": 9.684355092579085e-06,
+      "loss": 0.295,
+      "step": 1212
+    },
+    {
+      "epoch": 0.405008347245409,
+      "grad_norm": 0.5987189672064394,
+      "learning_rate": 9.683335435428332e-06,
+      "loss": 0.2987,
+      "step": 1213
+    },
+    {
+      "epoch": 0.40534223706176964,
+      "grad_norm": 0.5501973046615829,
+      "learning_rate": 9.68231418784108e-06,
+      "loss": 0.297,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4056761268781302,
+      "grad_norm": 0.6128014743351172,
+      "learning_rate": 9.681291350164139e-06,
+      "loss": 0.3089,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4060100166944908,
+      "grad_norm": 0.5253966936312215,
+      "learning_rate": 9.680266922744862e-06,
+      "loss": 0.3011,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4063439065108514,
+      "grad_norm": 0.6100031097407531,
+      "learning_rate": 9.679240905931132e-06,
+      "loss": 0.3113,
+      "step": 1217
+    },
+    {
+      "epoch": 0.406677796327212,
+      "grad_norm": 0.5660780616647275,
+      "learning_rate": 9.678213300071387e-06,
+      "loss": 0.2998,
+      "step": 1218
+    },
+    {
+      "epoch": 0.40701168614357264,
+      "grad_norm": 0.581236602317156,
+      "learning_rate": 9.677184105514592e-06,
+      "loss": 0.3036,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4073455759599332,
+      "grad_norm": 0.5469502963275812,
+      "learning_rate": 9.676153322610259e-06,
+      "loss": 0.2914,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4076794657762938,
+      "grad_norm": 0.584597158712472,
+      "learning_rate": 9.675120951708435e-06,
+      "loss": 0.316,
+      "step": 1221
+    },
+    {
+      "epoch": 0.40801335559265445,
+      "grad_norm": 0.6070607019056828,
+      "learning_rate": 9.674086993159706e-06,
+      "loss": 0.3027,
+      "step": 1222
+    },
+    {
+      "epoch": 0.408347245409015,
+      "grad_norm": 0.5614838912138467,
+      "learning_rate": 9.673051447315202e-06,
+      "loss": 0.3035,
+      "step": 1223
+    },
+    {
+      "epoch": 0.40868113522537564,
+      "grad_norm": 0.5666885332661947,
+      "learning_rate": 9.672014314526588e-06,
+      "loss": 0.2977,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4090150250417362,
+      "grad_norm": 0.5618162016148968,
+      "learning_rate": 9.670975595146068e-06,
+      "loss": 0.3017,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4093489148580968,
+      "grad_norm": 0.5473686217890248,
+      "learning_rate": 9.66993528952639e-06,
+      "loss": 0.3055,
+      "step": 1226
+    },
+    {
+      "epoch": 0.40968280467445745,
+      "grad_norm": 0.550348155967605,
+      "learning_rate": 9.668893398020828e-06,
+      "loss": 0.3056,
+      "step": 1227
+    },
+    {
+      "epoch": 0.410016694490818,
+      "grad_norm": 0.5346949025583592,
+      "learning_rate": 9.66784992098321e-06,
+      "loss": 0.3014,
+      "step": 1228
+    },
+    {
+      "epoch": 0.41035058430717863,
+      "grad_norm": 0.5630153111596908,
+      "learning_rate": 9.666804858767894e-06,
+      "loss": 0.2921,
+      "step": 1229
+    },
+    {
+      "epoch": 0.41068447412353926,
+      "grad_norm": 0.5991164927849593,
+      "learning_rate": 9.665758211729776e-06,
+      "loss": 0.3107,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4110183639398998,
+      "grad_norm": 0.5814847664852429,
+      "learning_rate": 9.664709980224295e-06,
+      "loss": 0.3013,
+      "step": 1231
+    },
+    {
+      "epoch": 0.41135225375626044,
+      "grad_norm": 0.5650093010882212,
+      "learning_rate": 9.66366016460742e-06,
+      "loss": 0.2932,
+      "step": 1232
+    },
+    {
+      "epoch": 0.411686143572621,
+      "grad_norm": 0.5979786913139735,
+      "learning_rate": 9.662608765235668e-06,
+      "loss": 0.3106,
+      "step": 1233
+    },
+    {
+      "epoch": 0.41202003338898163,
+      "grad_norm": 0.5554214288201806,
+      "learning_rate": 9.661555782466087e-06,
+      "loss": 0.2928,
+      "step": 1234
+    },
+    {
+      "epoch": 0.41235392320534225,
+      "grad_norm": 0.5701732998095413,
+      "learning_rate": 9.660501216656264e-06,
+      "loss": 0.3056,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4126878130217028,
+      "grad_norm": 0.5573029476919751,
+      "learning_rate": 9.659445068164323e-06,
+      "loss": 0.2903,
+      "step": 1236
+    },
+    {
+      "epoch": 0.41302170283806344,
+      "grad_norm": 0.5889975941435514,
+      "learning_rate": 9.658387337348928e-06,
+      "loss": 0.2976,
+      "step": 1237
+    },
+    {
+      "epoch": 0.41335559265442406,
+      "grad_norm": 0.5723551242704946,
+      "learning_rate": 9.65732802456928e-06,
+      "loss": 0.3072,
+      "step": 1238
+    },
+    {
+      "epoch": 0.41368948247078463,
+      "grad_norm": 0.5828668474597671,
+      "learning_rate": 9.656267130185116e-06,
+      "loss": 0.303,
+      "step": 1239
+    },
+    {
+      "epoch": 0.41402337228714525,
+      "grad_norm": 0.7075537544610832,
+      "learning_rate": 9.655204654556707e-06,
+      "loss": 0.3051,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4143572621035058,
+      "grad_norm": 0.5743186541258293,
+      "learning_rate": 9.654140598044865e-06,
+      "loss": 0.2935,
+      "step": 1241
+    },
+    {
+      "epoch": 0.41469115191986644,
+      "grad_norm": 0.568839420977327,
+      "learning_rate": 9.653074961010944e-06,
+      "loss": 0.289,
+      "step": 1242
+    },
+    {
+      "epoch": 0.41502504173622706,
+      "grad_norm": 0.617709893370013,
+      "learning_rate": 9.65200774381682e-06,
+      "loss": 0.3104,
+      "step": 1243
+    },
+    {
+      "epoch": 0.41535893155258763,
+      "grad_norm": 0.6280675534696337,
+      "learning_rate": 9.650938946824921e-06,
+      "loss": 0.3135,
+      "step": 1244
+    },
+    {
+      "epoch": 0.41569282136894825,
+      "grad_norm": 0.5364413619553433,
+      "learning_rate": 9.649868570398204e-06,
+      "loss": 0.2992,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4160267111853089,
+      "grad_norm": 0.6237763046331954,
+      "learning_rate": 9.64879661490016e-06,
+      "loss": 0.302,
+      "step": 1246
+    },
+    {
+      "epoch": 0.41636060100166944,
+      "grad_norm": 0.5834314702641802,
+      "learning_rate": 9.647723080694822e-06,
+      "loss": 0.298,
+      "step": 1247
+    },
+    {
+      "epoch": 0.41669449081803006,
+      "grad_norm": 0.5412388233820598,
+      "learning_rate": 9.646647968146757e-06,
+      "loss": 0.287,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4170283806343906,
+      "grad_norm": 0.6005954330795679,
+      "learning_rate": 9.645571277621066e-06,
+      "loss": 0.2969,
+      "step": 1249
+    },
+    {
+      "epoch": 0.41736227045075125,
+      "grad_norm": 0.5830053015230998,
+      "learning_rate": 9.644493009483389e-06,
+      "loss": 0.2994,
+      "step": 1250
+    },
+    {
+      "epoch": 0.41769616026711187,
+      "grad_norm": 0.5572865140379246,
+      "learning_rate": 9.6434131640999e-06,
+      "loss": 0.2947,
+      "step": 1251
+    },
+    {
+      "epoch": 0.41803005008347244,
+      "grad_norm": 0.5823831299478428,
+      "learning_rate": 9.642331741837307e-06,
+      "loss": 0.2994,
+      "step": 1252
+    },
+    {
+      "epoch": 0.41836393989983306,
+      "grad_norm": 0.5924444296012723,
+      "learning_rate": 9.641248743062858e-06,
+      "loss": 0.3192,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4186978297161937,
+      "grad_norm": 0.5879496944040656,
+      "learning_rate": 9.640164168144332e-06,
+      "loss": 0.3047,
+      "step": 1254
+    },
+    {
+      "epoch": 0.41903171953255425,
+      "grad_norm": 0.5459748085760789,
+      "learning_rate": 9.639078017450048e-06,
+      "loss": 0.3064,
+      "step": 1255
+    },
+    {
+      "epoch": 0.41936560934891487,
+      "grad_norm": 0.5897904290846084,
+      "learning_rate": 9.637990291348853e-06,
+      "loss": 0.302,
+      "step": 1256
+    },
+    {
+      "epoch": 0.41969949916527544,
+      "grad_norm": 0.5745010751583389,
+      "learning_rate": 9.636900990210136e-06,
+      "loss": 0.2979,
+      "step": 1257
+    },
+    {
+      "epoch": 0.42003338898163606,
+      "grad_norm": 0.5804436284674882,
+      "learning_rate": 9.635810114403818e-06,
+      "loss": 0.3077,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4203672787979967,
+      "grad_norm": 0.5404595609004945,
+      "learning_rate": 9.634717664300352e-06,
+      "loss": 0.2925,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42070116861435725,
+      "grad_norm": 0.6041655939613235,
+      "learning_rate": 9.633623640270733e-06,
+      "loss": 0.2981,
+      "step": 1260
+    },
+    {
+      "epoch": 0.42103505843071787,
+      "grad_norm": 0.5730454305910874,
+      "learning_rate": 9.632528042686482e-06,
+      "loss": 0.3025,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4213689482470785,
+      "grad_norm": 0.5716928014672171,
+      "learning_rate": 9.631430871919659e-06,
+      "loss": 0.3103,
+      "step": 1262
+    },
+    {
+      "epoch": 0.42170283806343906,
+      "grad_norm": 0.565556985934863,
+      "learning_rate": 9.630332128342858e-06,
+      "loss": 0.3043,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4220367278797997,
+      "grad_norm": 0.6245400943735304,
+      "learning_rate": 9.629231812329207e-06,
+      "loss": 0.3077,
+      "step": 1264
+    },
+    {
+      "epoch": 0.42237061769616024,
+      "grad_norm": 0.5488887984697077,
+      "learning_rate": 9.628129924252368e-06,
+      "loss": 0.2981,
+      "step": 1265
+    },
+    {
+      "epoch": 0.42270450751252087,
+      "grad_norm": 0.6343982338968454,
+      "learning_rate": 9.627026464486535e-06,
+      "loss": 0.3001,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4230383973288815,
+      "grad_norm": 0.5645734523628672,
+      "learning_rate": 9.62592143340644e-06,
+      "loss": 0.3012,
+      "step": 1267
+    },
+    {
+      "epoch": 0.42337228714524205,
+      "grad_norm": 0.5580973265758677,
+      "learning_rate": 9.624814831387342e-06,
+      "loss": 0.3059,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4237061769616027,
+      "grad_norm": 0.5581075234052267,
+      "learning_rate": 9.62370665880504e-06,
+      "loss": 0.3056,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4240400667779633,
+      "grad_norm": 0.5737819759473839,
+      "learning_rate": 9.622596916035862e-06,
+      "loss": 0.3018,
+      "step": 1270
+    },
+    {
+      "epoch": 0.42437395659432386,
+      "grad_norm": 0.5549969940023873,
+      "learning_rate": 9.621485603456673e-06,
+      "loss": 0.3034,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4247078464106845,
+      "grad_norm": 0.5620845457261101,
+      "learning_rate": 9.620372721444868e-06,
+      "loss": 0.3093,
+      "step": 1272
+    },
+    {
+      "epoch": 0.42504173622704505,
+      "grad_norm": 0.5784702826291268,
+      "learning_rate": 9.619258270378375e-06,
+      "loss": 0.3027,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4253756260434057,
+      "grad_norm": 0.7134211688470319,
+      "learning_rate": 9.618142250635658e-06,
+      "loss": 0.3067,
+      "step": 1274
+    },
+    {
+      "epoch": 0.4257095158597663,
+      "grad_norm": 0.5449386973265349,
+      "learning_rate": 9.61702466259571e-06,
+      "loss": 0.3031,
+      "step": 1275
+    },
+    {
+      "epoch": 0.42604340567612686,
+      "grad_norm": 0.554233201192273,
+      "learning_rate": 9.615905506638056e-06,
+      "loss": 0.2932,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4263772954924875,
+      "grad_norm": 0.6246747008377045,
+      "learning_rate": 9.614784783142759e-06,
+      "loss": 0.2982,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4267111853088481,
+      "grad_norm": 0.5851902059370244,
+      "learning_rate": 9.613662492490409e-06,
+      "loss": 0.2996,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4270450751252087,
+      "grad_norm": 0.5893049471534945,
+      "learning_rate": 9.612538635062134e-06,
+      "loss": 0.2882,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4273789649415693,
+      "grad_norm": 0.5719008057046002,
+      "learning_rate": 9.611413211239585e-06,
+      "loss": 0.2917,
+      "step": 1280
+    },
+    {
+      "epoch": 0.42771285475792986,
+      "grad_norm": 0.5667301390812207,
+      "learning_rate": 9.610286221404951e-06,
+      "loss": 0.3101,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4280467445742905,
+      "grad_norm": 0.5623995673985254,
+      "learning_rate": 9.609157665940956e-06,
+      "loss": 0.3014,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4283806343906511,
+      "grad_norm": 0.5239835230037186,
+      "learning_rate": 9.608027545230847e-06,
+      "loss": 0.2938,
+      "step": 1283
+    },
+    {
+      "epoch": 0.42871452420701167,
+      "grad_norm": 0.5758149956786311,
+      "learning_rate": 9.606895859658408e-06,
+      "loss": 0.2993,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4290484140233723,
+      "grad_norm": 0.5721406583735558,
+      "learning_rate": 9.605762609607955e-06,
+      "loss": 0.3123,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4293823038397329,
+      "grad_norm": 0.542042193348531,
+      "learning_rate": 9.604627795464335e-06,
+      "loss": 0.2998,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4297161936560935,
+      "grad_norm": 0.5657077503451297,
+      "learning_rate": 9.60349141761292e-06,
+      "loss": 0.2966,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4300500834724541,
+      "grad_norm": 0.5741907895223441,
+      "learning_rate": 9.602353476439624e-06,
+      "loss": 0.3068,
+      "step": 1288
+    },
+    {
+      "epoch": 0.43038397328881467,
+      "grad_norm": 0.559257747982772,
+      "learning_rate": 9.601213972330883e-06,
+      "loss": 0.3127,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4307178631051753,
+      "grad_norm": 0.5780243605299029,
+      "learning_rate": 9.600072905673665e-06,
+      "loss": 0.2991,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4310517529215359,
+      "grad_norm": 0.6256044026630685,
+      "learning_rate": 9.598930276855474e-06,
+      "loss": 0.3048,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4313856427378965,
+      "grad_norm": 0.5617531490859377,
+      "learning_rate": 9.597786086264339e-06,
+      "loss": 0.2975,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4317195325542571,
+      "grad_norm": 0.5869002265233533,
+      "learning_rate": 9.596640334288822e-06,
+      "loss": 0.3079,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4320534223706177,
+      "grad_norm": 0.608292701584339,
+      "learning_rate": 9.595493021318012e-06,
+      "loss": 0.3007,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4323873121869783,
+      "grad_norm": 0.5404179205376484,
+      "learning_rate": 9.594344147741534e-06,
+      "loss": 0.2941,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4327212020033389,
+      "grad_norm": 0.5382353920526677,
+      "learning_rate": 9.593193713949537e-06,
+      "loss": 0.2957,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4330550918196995,
+      "grad_norm": 0.551629096127472,
+      "learning_rate": 9.592041720332703e-06,
+      "loss": 0.3001,
+      "step": 1297
+    },
+    {
+      "epoch": 0.4333889816360601,
+      "grad_norm": 0.5743223439079059,
+      "learning_rate": 9.590888167282245e-06,
+      "loss": 0.3032,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4337228714524207,
+      "grad_norm": 0.5376380887010805,
+      "learning_rate": 9.5897330551899e-06,
+      "loss": 0.2993,
+      "step": 1299
+    },
+    {
+      "epoch": 0.4340567612687813,
+      "grad_norm": 0.5490774397927959,
+      "learning_rate": 9.58857638444794e-06,
+      "loss": 0.3067,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4343906510851419,
+      "grad_norm": 0.5700154418629376,
+      "learning_rate": 9.587418155449167e-06,
+      "loss": 0.3044,
+      "step": 1301
+    },
+    {
+      "epoch": 0.43472454090150253,
+      "grad_norm": 0.5590528066174716,
+      "learning_rate": 9.586258368586906e-06,
+      "loss": 0.3044,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4350584307178631,
+      "grad_norm": 0.5551953537595498,
+      "learning_rate": 9.585097024255013e-06,
+      "loss": 0.3014,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4353923205342237,
+      "grad_norm": 0.5331378970133521,
+      "learning_rate": 9.58393412284788e-06,
+      "loss": 0.2917,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4357262103505843,
+      "grad_norm": 0.5524107589296284,
+      "learning_rate": 9.582769664760418e-06,
+      "loss": 0.3052,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4360601001669449,
+      "grad_norm": 0.5494466487842073,
+      "learning_rate": 9.581603650388073e-06,
+      "loss": 0.2933,
+      "step": 1306
+    },
+    {
+      "epoch": 0.43639398998330553,
+      "grad_norm": 0.5761279478717288,
+      "learning_rate": 9.580436080126819e-06,
+      "loss": 0.3056,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4367278797996661,
+      "grad_norm": 0.5937855044819362,
+      "learning_rate": 9.579266954373152e-06,
+      "loss": 0.3025,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4370617696160267,
+      "grad_norm": 0.5989128591630327,
+      "learning_rate": 9.578096273524104e-06,
+      "loss": 0.3101,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4373956594323873,
+      "grad_norm": 0.5686861743195103,
+      "learning_rate": 9.576924037977233e-06,
+      "loss": 0.2994,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4377295492487479,
+      "grad_norm": 0.5504071421515085,
+      "learning_rate": 9.575750248130623e-06,
+      "loss": 0.3018,
+      "step": 1311
+    },
+    {
+      "epoch": 0.43806343906510853,
+      "grad_norm": 0.5548753015861687,
+      "learning_rate": 9.574574904382887e-06,
+      "loss": 0.2894,
+      "step": 1312
+    },
+    {
+      "epoch": 0.4383973288814691,
+      "grad_norm": 0.558550827916228,
+      "learning_rate": 9.573398007133166e-06,
+      "loss": 0.2851,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4387312186978297,
+      "grad_norm": 0.570948109919578,
+      "learning_rate": 9.572219556781126e-06,
+      "loss": 0.3028,
+      "step": 1314
+    },
+    {
+      "epoch": 0.43906510851419034,
+      "grad_norm": 0.5804691231498674,
+      "learning_rate": 9.571039553726966e-06,
+      "loss": 0.3065,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4393989983305509,
+      "grad_norm": 0.5496100908049057,
+      "learning_rate": 9.569857998371408e-06,
+      "loss": 0.2982,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4397328881469115,
+      "grad_norm": 0.5523423161397331,
+      "learning_rate": 9.5686748911157e-06,
+      "loss": 0.2938,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4400667779632721,
+      "grad_norm": 0.5343506437471376,
+      "learning_rate": 9.567490232361623e-06,
+      "loss": 0.3087,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4404006677796327,
+      "grad_norm": 0.5576364951232331,
+      "learning_rate": 9.566304022511477e-06,
+      "loss": 0.2948,
+      "step": 1319
+    },
+    {
+      "epoch": 0.44073455759599334,
+      "grad_norm": 0.5718789226831753,
+      "learning_rate": 9.565116261968095e-06,
+      "loss": 0.2981,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4410684474123539,
+      "grad_norm": 0.5622947149713721,
+      "learning_rate": 9.563926951134834e-06,
+      "loss": 0.294,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4414023372287145,
+      "grad_norm": 0.5459770322486117,
+      "learning_rate": 9.562736090415577e-06,
+      "loss": 0.3049,
+      "step": 1322
+    },
+    {
+      "epoch": 0.44173622704507515,
+      "grad_norm": 0.5573728906480746,
+      "learning_rate": 9.561543680214735e-06,
+      "loss": 0.2975,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4420701168614357,
+      "grad_norm": 0.6272429485302781,
+      "learning_rate": 9.560349720937246e-06,
+      "loss": 0.2928,
+      "step": 1324
+    },
+    {
+      "epoch": 0.44240400667779634,
+      "grad_norm": 0.5583991154548642,
+      "learning_rate": 9.559154212988566e-06,
+      "loss": 0.2956,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4427378964941569,
+      "grad_norm": 0.6249378166218553,
+      "learning_rate": 9.55795715677469e-06,
+      "loss": 0.2852,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4430717863105175,
+      "grad_norm": 0.573671699851079,
+      "learning_rate": 9.556758552702132e-06,
+      "loss": 0.3066,
+      "step": 1327
+    },
+    {
+      "epoch": 0.44340567612687815,
+      "grad_norm": 0.5923548581470902,
+      "learning_rate": 9.555558401177927e-06,
+      "loss": 0.3171,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4437395659432387,
+      "grad_norm": 0.5872062699369104,
+      "learning_rate": 9.554356702609645e-06,
+      "loss": 0.309,
+      "step": 1329
+    },
+    {
+      "epoch": 0.44407345575959933,
+      "grad_norm": 0.5673281623775791,
+      "learning_rate": 9.553153457405373e-06,
+      "loss": 0.3036,
+      "step": 1330
+    },
+    {
+      "epoch": 0.44440734557595996,
+      "grad_norm": 0.5497738713252757,
+      "learning_rate": 9.551948665973728e-06,
+      "loss": 0.3003,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4447412353923205,
+      "grad_norm": 0.5281814153446281,
+      "learning_rate": 9.550742328723852e-06,
+      "loss": 0.298,
+      "step": 1332
+    },
+    {
+      "epoch": 0.44507512520868114,
+      "grad_norm": 0.5691089068273326,
+      "learning_rate": 9.54953444606541e-06,
+      "loss": 0.2853,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4454090150250417,
+      "grad_norm": 0.5516820634590399,
+      "learning_rate": 9.548325018408592e-06,
+      "loss": 0.2907,
+      "step": 1334
+    },
+    {
+      "epoch": 0.44574290484140233,
+      "grad_norm": 0.5364336578592238,
+      "learning_rate": 9.547114046164116e-06,
+      "loss": 0.2916,
+      "step": 1335
+    },
+    {
+      "epoch": 0.44607679465776295,
+      "grad_norm": 0.5130783412776536,
+      "learning_rate": 9.54590152974322e-06,
+      "loss": 0.2918,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4464106844741235,
+      "grad_norm": 0.5674020177519787,
+      "learning_rate": 9.544687469557667e-06,
+      "loss": 0.306,
+      "step": 1337
+    },
+    {
+      "epoch": 0.44674457429048414,
+      "grad_norm": 0.605836654659955,
+      "learning_rate": 9.543471866019748e-06,
+      "loss": 0.3076,
+      "step": 1338
+    },
+    {
+      "epoch": 0.44707846410684476,
+      "grad_norm": 0.5318128601252771,
+      "learning_rate": 9.542254719542274e-06,
+      "loss": 0.2987,
+      "step": 1339
+    },
+    {
+      "epoch": 0.44741235392320533,
+      "grad_norm": 0.5557555768306986,
+      "learning_rate": 9.541036030538584e-06,
+      "loss": 0.3045,
+      "step": 1340
+    },
+    {
+      "epoch": 0.44774624373956595,
+      "grad_norm": 0.5640647639387519,
+      "learning_rate": 9.539815799422536e-06,
+      "loss": 0.3061,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4480801335559265,
+      "grad_norm": 0.5656319799419318,
+      "learning_rate": 9.538594026608514e-06,
+      "loss": 0.2945,
+      "step": 1342
+    },
+    {
+      "epoch": 0.44841402337228714,
+      "grad_norm": 0.5568924267317756,
+      "learning_rate": 9.537370712511427e-06,
+      "loss": 0.3031,
+      "step": 1343
+    },
+    {
+      "epoch": 0.44874791318864776,
+      "grad_norm": 0.5569971162469466,
+      "learning_rate": 9.536145857546708e-06,
+      "loss": 0.2948,
+      "step": 1344
+    },
+    {
+      "epoch": 0.44908180300500833,
+      "grad_norm": 0.5477993481453296,
+      "learning_rate": 9.534919462130308e-06,
+      "loss": 0.2849,
+      "step": 1345
+    },
+    {
+      "epoch": 0.44941569282136895,
+      "grad_norm": 0.6021876430627953,
+      "learning_rate": 9.533691526678705e-06,
+      "loss": 0.2983,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4497495826377296,
+      "grad_norm": 0.5443745945949616,
+      "learning_rate": 9.532462051608901e-06,
+      "loss": 0.3068,
+      "step": 1347
+    },
+    {
+      "epoch": 0.45008347245409014,
+      "grad_norm": 0.5304694667199327,
+      "learning_rate": 9.53123103733842e-06,
+      "loss": 0.2873,
+      "step": 1348
+    },
+    {
+      "epoch": 0.45041736227045076,
+      "grad_norm": 0.5662724670595308,
+      "learning_rate": 9.529998484285306e-06,
+      "loss": 0.2887,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4507512520868113,
+      "grad_norm": 0.5886294045223402,
+      "learning_rate": 9.528764392868129e-06,
+      "loss": 0.2937,
+      "step": 1350
+    },
+    {
+      "epoch": 0.45108514190317195,
+      "grad_norm": 0.5248334253565741,
+      "learning_rate": 9.527528763505978e-06,
+      "loss": 0.2949,
+      "step": 1351
+    },
+    {
+      "epoch": 0.45141903171953257,
+      "grad_norm": 0.6163812803758598,
+      "learning_rate": 9.526291596618469e-06,
+      "loss": 0.3088,
+      "step": 1352
+    },
+    {
+      "epoch": 0.45175292153589314,
+      "grad_norm": 0.5669793514188297,
+      "learning_rate": 9.525052892625734e-06,
+      "loss": 0.3024,
+      "step": 1353
+    },
+    {
+      "epoch": 0.45208681135225376,
+      "grad_norm": 0.5386448739879618,
+      "learning_rate": 9.523812651948437e-06,
+      "loss": 0.2954,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4524207011686144,
+      "grad_norm": 0.5328902497363873,
+      "learning_rate": 9.52257087500775e-06,
+      "loss": 0.293,
+      "step": 1355
+    },
+    {
+      "epoch": 0.45275459098497495,
+      "grad_norm": 0.5459334123952501,
+      "learning_rate": 9.521327562225378e-06,
+      "loss": 0.2976,
+      "step": 1356
+    },
+    {
+      "epoch": 0.45308848080133557,
+      "grad_norm": 0.5849425555186517,
+      "learning_rate": 9.520082714023544e-06,
+      "loss": 0.292,
+      "step": 1357
+    },
+    {
+      "epoch": 0.45342237061769614,
+      "grad_norm": 0.5203239033835653,
+      "learning_rate": 9.51883633082499e-06,
+      "loss": 0.2974,
+      "step": 1358
+    },
+    {
+      "epoch": 0.45375626043405676,
+      "grad_norm": 0.5280612414037584,
+      "learning_rate": 9.517588413052983e-06,
+      "loss": 0.2971,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4540901502504174,
+      "grad_norm": 0.5858622883914416,
+      "learning_rate": 9.516338961131305e-06,
+      "loss": 0.2896,
+      "step": 1360
+    },
+    {
+      "epoch": 0.45442404006677795,
+      "grad_norm": 0.5484589407411365,
+      "learning_rate": 9.515087975484271e-06,
+      "loss": 0.285,
+      "step": 1361
+    },
+    {
+      "epoch": 0.45475792988313857,
+      "grad_norm": 0.5413935012904321,
+      "learning_rate": 9.513835456536704e-06,
+      "loss": 0.3029,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4550918196994992,
+      "grad_norm": 0.5541503314785775,
+      "learning_rate": 9.512581404713954e-06,
+      "loss": 0.3035,
+      "step": 1363
+    },
+    {
+      "epoch": 0.45542570951585976,
+      "grad_norm": 0.6036660928692797,
+      "learning_rate": 9.51132582044189e-06,
+      "loss": 0.3052,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4557595993322204,
+      "grad_norm": 0.5684425016454165,
+      "learning_rate": 9.510068704146905e-06,
+      "loss": 0.2996,
+      "step": 1365
+    },
+    {
+      "epoch": 0.45609348914858094,
+      "grad_norm": 0.517271189151883,
+      "learning_rate": 9.508810056255903e-06,
+      "loss": 0.2931,
+      "step": 1366
+    },
+    {
+      "epoch": 0.45642737896494157,
+      "grad_norm": 0.5764197256319434,
+      "learning_rate": 9.50754987719632e-06,
+      "loss": 0.3089,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4567612687813022,
+      "grad_norm": 0.5414189805089954,
+      "learning_rate": 9.506288167396106e-06,
+      "loss": 0.3003,
+      "step": 1368
+    },
+    {
+      "epoch": 0.45709515859766275,
+      "grad_norm": 0.5550848345811172,
+      "learning_rate": 9.505024927283729e-06,
+      "loss": 0.3014,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4574290484140234,
+      "grad_norm": 0.5456419969295169,
+      "learning_rate": 9.503760157288178e-06,
+      "loss": 0.2913,
+      "step": 1370
+    },
+    {
+      "epoch": 0.457762938230384,
+      "grad_norm": 0.5810747936959497,
+      "learning_rate": 9.502493857838963e-06,
+      "loss": 0.3025,
+      "step": 1371
+    },
+    {
+      "epoch": 0.45809682804674456,
+      "grad_norm": 0.5620376717268318,
+      "learning_rate": 9.501226029366115e-06,
+      "loss": 0.2934,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4584307178631052,
+      "grad_norm": 0.5546393266287942,
+      "learning_rate": 9.49995667230018e-06,
+      "loss": 0.2998,
+      "step": 1373
+    },
+    {
+      "epoch": 0.45876460767946575,
+      "grad_norm": 0.5635756025984424,
+      "learning_rate": 9.498685787072225e-06,
+      "loss": 0.3151,
+      "step": 1374
+    },
+    {
+      "epoch": 0.4590984974958264,
+      "grad_norm": 0.5598216356883206,
+      "learning_rate": 9.497413374113838e-06,
+      "loss": 0.2947,
+      "step": 1375
+    },
+    {
+      "epoch": 0.459432387312187,
+      "grad_norm": 0.5518913778856231,
+      "learning_rate": 9.49613943385712e-06,
+      "loss": 0.2959,
+      "step": 1376
+    },
+    {
+      "epoch": 0.45976627712854756,
+      "grad_norm": 0.5448155589962211,
+      "learning_rate": 9.494863966734698e-06,
+      "loss": 0.2981,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4601001669449082,
+      "grad_norm": 0.5346821160932808,
+      "learning_rate": 9.493586973179714e-06,
+      "loss": 0.304,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4604340567612688,
+      "grad_norm": 0.5454141792507852,
+      "learning_rate": 9.492308453625828e-06,
+      "loss": 0.2973,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4607679465776294,
+      "grad_norm": 0.6133756392291456,
+      "learning_rate": 9.491028408507215e-06,
+      "loss": 0.3091,
+      "step": 1380
+    },
+    {
+      "epoch": 0.46110183639399,
+      "grad_norm": 0.532718871189078,
+      "learning_rate": 9.489746838258578e-06,
+      "loss": 0.2998,
+      "step": 1381
+    },
+    {
+      "epoch": 0.46143572621035056,
+      "grad_norm": 0.5477618975712433,
+      "learning_rate": 9.488463743315126e-06,
+      "loss": 0.305,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4617696160267112,
+      "grad_norm": 0.4990205531828872,
+      "learning_rate": 9.487179124112596e-06,
+      "loss": 0.3014,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4621035058430718,
+      "grad_norm": 0.5451583359187276,
+      "learning_rate": 9.485892981087234e-06,
+      "loss": 0.3015,
+      "step": 1384
+    },
+    {
+      "epoch": 0.46243739565943237,
+      "grad_norm": 0.5551059512930452,
+      "learning_rate": 9.48460531467581e-06,
+      "loss": 0.2957,
+      "step": 1385
+    },
+    {
+      "epoch": 0.462771285475793,
+      "grad_norm": 0.542799794123925,
+      "learning_rate": 9.483316125315607e-06,
+      "loss": 0.3033,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4631051752921536,
+      "grad_norm": 0.5519994735464386,
+      "learning_rate": 9.48202541344443e-06,
+      "loss": 0.2886,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4634390651085142,
+      "grad_norm": 0.5794325203946948,
+      "learning_rate": 9.480733179500596e-06,
+      "loss": 0.297,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4637729549248748,
+      "grad_norm": 0.5733784441861524,
+      "learning_rate": 9.47943942392294e-06,
+      "loss": 0.3023,
+      "step": 1389
+    },
+    {
+      "epoch": 0.46410684474123537,
+      "grad_norm": 0.5402507513830642,
+      "learning_rate": 9.478144147150818e-06,
+      "loss": 0.3149,
+      "step": 1390
+    },
+    {
+      "epoch": 0.464440734557596,
+      "grad_norm": 0.56760382300041,
+      "learning_rate": 9.476847349624097e-06,
+      "loss": 0.3052,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4647746243739566,
+      "grad_norm": 0.5896455952749481,
+      "learning_rate": 9.475549031783161e-06,
+      "loss": 0.303,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4651085141903172,
+      "grad_norm": 0.5768560799032607,
+      "learning_rate": 9.474249194068915e-06,
+      "loss": 0.2882,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4654424040066778,
+      "grad_norm": 0.5481479216924592,
+      "learning_rate": 9.472947836922776e-06,
+      "loss": 0.3031,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4657762938230384,
+      "grad_norm": 0.5326018464341632,
+      "learning_rate": 9.47164496078668e-06,
+      "loss": 0.297,
+      "step": 1395
+    },
+    {
+      "epoch": 0.466110183639399,
+      "grad_norm": 0.560543058016086,
+      "learning_rate": 9.470340566103074e-06,
+      "loss": 0.2841,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4664440734557596,
+      "grad_norm": 0.5731737984162757,
+      "learning_rate": 9.469034653314927e-06,
+      "loss": 0.2925,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4667779632721202,
+      "grad_norm": 0.5605049683117135,
+      "learning_rate": 9.467727222865715e-06,
+      "loss": 0.319,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4671118530884808,
+      "grad_norm": 0.5258747552992491,
+      "learning_rate": 9.46641827519944e-06,
+      "loss": 0.304,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4674457429048414,
+      "grad_norm": 0.6015650957004121,
+      "learning_rate": 9.46510781076061e-06,
+      "loss": 0.2938,
+      "step": 1400
+    },
+    {
+      "epoch": 0.467779632721202,
+      "grad_norm": 0.5847123887257757,
+      "learning_rate": 9.463795829994255e-06,
+      "loss": 0.3047,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4681135225375626,
+      "grad_norm": 0.5378096517614142,
+      "learning_rate": 9.462482333345917e-06,
+      "loss": 0.2895,
+      "step": 1402
+    },
+    {
+      "epoch": 0.46844741235392323,
+      "grad_norm": 0.5890846447963135,
+      "learning_rate": 9.46116732126165e-06,
+      "loss": 0.3022,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4687813021702838,
+      "grad_norm": 0.6261517592572303,
+      "learning_rate": 9.459850794188026e-06,
+      "loss": 0.3069,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4691151919866444,
+      "grad_norm": 0.5612624478230952,
+      "learning_rate": 9.458532752572131e-06,
+      "loss": 0.2983,
+      "step": 1405
+    },
+    {
+      "epoch": 0.469449081803005,
+      "grad_norm": 0.6028943819294715,
+      "learning_rate": 9.457213196861567e-06,
+      "loss": 0.2998,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4697829716193656,
+      "grad_norm": 0.6508303617252174,
+      "learning_rate": 9.455892127504447e-06,
+      "loss": 0.2947,
+      "step": 1407
+    },
+    {
+      "epoch": 0.47011686143572623,
+      "grad_norm": 0.5333491669360131,
+      "learning_rate": 9.454569544949397e-06,
+      "loss": 0.2847,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4704507512520868,
+      "grad_norm": 0.5507938936453806,
+      "learning_rate": 9.453245449645562e-06,
+      "loss": 0.292,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4707846410684474,
+      "grad_norm": 0.5773571847833764,
+      "learning_rate": 9.451919842042599e-06,
+      "loss": 0.2977,
+      "step": 1410
+    },
+    {
+      "epoch": 0.47111853088480804,
+      "grad_norm": 0.5524945094057461,
+      "learning_rate": 9.450592722590675e-06,
+      "loss": 0.291,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4714524207011686,
+      "grad_norm": 0.6193524245388031,
+      "learning_rate": 9.449264091740473e-06,
+      "loss": 0.2963,
+      "step": 1412
+    },
+    {
+      "epoch": 0.47178631051752923,
+      "grad_norm": 0.5458455306310108,
+      "learning_rate": 9.447933949943188e-06,
+      "loss": 0.293,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4721202003338898,
+      "grad_norm": 0.526940944329889,
+      "learning_rate": 9.446602297650533e-06,
+      "loss": 0.293,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4724540901502504,
+      "grad_norm": 0.5311787315493047,
+      "learning_rate": 9.445269135314729e-06,
+      "loss": 0.3012,
+      "step": 1415
+    },
+    {
+      "epoch": 0.47278797996661104,
+      "grad_norm": 0.5214055936709477,
+      "learning_rate": 9.443934463388509e-06,
+      "loss": 0.2907,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4731218697829716,
+      "grad_norm": 0.51173393055626,
+      "learning_rate": 9.442598282325124e-06,
+      "loss": 0.2958,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4734557595993322,
+      "grad_norm": 0.5527398576625742,
+      "learning_rate": 9.44126059257833e-06,
+      "loss": 0.2964,
+      "step": 1418
+    },
+    {
+      "epoch": 0.47378964941569285,
+      "grad_norm": 0.5352706998416287,
+      "learning_rate": 9.439921394602403e-06,
+      "loss": 0.3038,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4741235392320534,
+      "grad_norm": 0.5541021236720325,
+      "learning_rate": 9.438580688852126e-06,
+      "loss": 0.2981,
+      "step": 1420
+    },
+    {
+      "epoch": 0.47445742904841404,
+      "grad_norm": 0.5435582311879625,
+      "learning_rate": 9.437238475782795e-06,
+      "loss": 0.2872,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4747913188647746,
+      "grad_norm": 0.5272899303003794,
+      "learning_rate": 9.43589475585022e-06,
+      "loss": 0.3068,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4751252086811352,
+      "grad_norm": 0.524918846048107,
+      "learning_rate": 9.434549529510724e-06,
+      "loss": 0.292,
+      "step": 1423
+    },
+    {
+      "epoch": 0.47545909849749585,
+      "grad_norm": 0.5121431980672432,
+      "learning_rate": 9.433202797221132e-06,
+      "loss": 0.2926,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4757929883138564,
+      "grad_norm": 0.5344770518466105,
+      "learning_rate": 9.431854559438795e-06,
+      "loss": 0.3049,
+      "step": 1425
+    },
+    {
+      "epoch": 0.47612687813021703,
+      "grad_norm": 0.5804702330075423,
+      "learning_rate": 9.430504816621562e-06,
+      "loss": 0.3029,
+      "step": 1426
+    },
+    {
+      "epoch": 0.47646076794657766,
+      "grad_norm": 0.5305934879647007,
+      "learning_rate": 9.4291535692278e-06,
+      "loss": 0.2868,
+      "step": 1427
+    },
+    {
+      "epoch": 0.4767946577629382,
+      "grad_norm": 0.5530597686592379,
+      "learning_rate": 9.427800817716388e-06,
+      "loss": 0.285,
+      "step": 1428
+    },
+    {
+      "epoch": 0.47712854757929885,
+      "grad_norm": 0.5152759687740402,
+      "learning_rate": 9.426446562546711e-06,
+      "loss": 0.28,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4774624373956594,
+      "grad_norm": 0.6419121578536966,
+      "learning_rate": 9.425090804178667e-06,
+      "loss": 0.308,
+      "step": 1430
+    },
+    {
+      "epoch": 0.47779632721202003,
+      "grad_norm": 0.6063838189867258,
+      "learning_rate": 9.423733543072666e-06,
+      "loss": 0.3084,
+      "step": 1431
+    },
+    {
+      "epoch": 0.47813021702838066,
+      "grad_norm": 0.5442204699404515,
+      "learning_rate": 9.422374779689625e-06,
+      "loss": 0.2945,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4784641068447412,
+      "grad_norm": 0.5307055588102764,
+      "learning_rate": 9.421014514490977e-06,
+      "loss": 0.2955,
+      "step": 1433
+    },
+    {
+      "epoch": 0.47879799666110184,
+      "grad_norm": 0.5713003849150992,
+      "learning_rate": 9.419652747938655e-06,
+      "loss": 0.3063,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4791318864774624,
+      "grad_norm": 0.5594318336108197,
+      "learning_rate": 9.418289480495114e-06,
+      "loss": 0.29,
+      "step": 1435
+    },
+    {
+      "epoch": 0.47946577629382303,
+      "grad_norm": 0.5695739994063159,
+      "learning_rate": 9.416924712623305e-06,
+      "loss": 0.3012,
+      "step": 1436
+    },
+    {
+      "epoch": 0.47979966611018365,
+      "grad_norm": 0.5373820158920897,
+      "learning_rate": 9.415558444786706e-06,
+      "loss": 0.2969,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4801335559265442,
+      "grad_norm": 0.5572099584038135,
+      "learning_rate": 9.414190677449285e-06,
+      "loss": 0.297,
+      "step": 1438
+    },
+    {
+      "epoch": 0.48046744574290484,
+      "grad_norm": 0.5519338347000604,
+      "learning_rate": 9.412821411075535e-06,
+      "loss": 0.2989,
+      "step": 1439
+    },
+    {
+      "epoch": 0.48080133555926546,
+      "grad_norm": 0.5531644878014962,
+      "learning_rate": 9.411450646130448e-06,
+      "loss": 0.2946,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48113522537562603,
+      "grad_norm": 0.5218642492850009,
+      "learning_rate": 9.410078383079532e-06,
+      "loss": 0.2908,
+      "step": 1441
+    },
+    {
+      "epoch": 0.48146911519198665,
+      "grad_norm": 0.5641176144664156,
+      "learning_rate": 9.4087046223888e-06,
+      "loss": 0.3033,
+      "step": 1442
+    },
+    {
+      "epoch": 0.4818030050083472,
+      "grad_norm": 0.5579977143132894,
+      "learning_rate": 9.40732936452477e-06,
+      "loss": 0.29,
+      "step": 1443
+    },
+    {
+      "epoch": 0.48213689482470784,
+      "grad_norm": 0.5486538369661588,
+      "learning_rate": 9.405952609954478e-06,
+      "loss": 0.2883,
+      "step": 1444
+    },
+    {
+      "epoch": 0.48247078464106846,
+      "grad_norm": 0.5454924804611442,
+      "learning_rate": 9.40457435914546e-06,
+      "loss": 0.305,
+      "step": 1445
+    },
+    {
+      "epoch": 0.48280467445742903,
+      "grad_norm": 0.5539441561446176,
+      "learning_rate": 9.40319461256576e-06,
+      "loss": 0.3013,
+      "step": 1446
+    },
+    {
+      "epoch": 0.48313856427378965,
+      "grad_norm": 0.5590383641801764,
+      "learning_rate": 9.401813370683938e-06,
+      "loss": 0.3,
+      "step": 1447
+    },
+    {
+      "epoch": 0.48347245409015027,
+      "grad_norm": 0.5117274585639341,
+      "learning_rate": 9.400430633969054e-06,
+      "loss": 0.2871,
+      "step": 1448
+    },
+    {
+      "epoch": 0.48380634390651084,
+      "grad_norm": 0.5406368507669954,
+      "learning_rate": 9.39904640289068e-06,
+      "loss": 0.2907,
+      "step": 1449
+    },
+    {
+      "epoch": 0.48414023372287146,
+      "grad_norm": 0.5654575997609788,
+      "learning_rate": 9.39766067791889e-06,
+      "loss": 0.2908,
+      "step": 1450
+    },
+    {
+      "epoch": 0.484474123539232,
+      "grad_norm": 0.5660913412181767,
+      "learning_rate": 9.396273459524272e-06,
+      "loss": 0.3042,
+      "step": 1451
+    },
+    {
+      "epoch": 0.48480801335559265,
+      "grad_norm": 0.5829773935970148,
+      "learning_rate": 9.394884748177917e-06,
+      "loss": 0.3082,
+      "step": 1452
+    },
+    {
+      "epoch": 0.48514190317195327,
+      "grad_norm": 0.5015907140583401,
+      "learning_rate": 9.393494544351425e-06,
+      "loss": 0.2844,
+      "step": 1453
+    },
+    {
+      "epoch": 0.48547579298831384,
+      "grad_norm": 0.5555208305169961,
+      "learning_rate": 9.392102848516901e-06,
+      "loss": 0.2823,
+      "step": 1454
+    },
+    {
+      "epoch": 0.48580968280467446,
+      "grad_norm": 0.5666585834234532,
+      "learning_rate": 9.39070966114696e-06,
+      "loss": 0.2944,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4861435726210351,
+      "grad_norm": 0.5515349341953032,
+      "learning_rate": 9.389314982714718e-06,
+      "loss": 0.2946,
+      "step": 1456
+    },
+    {
+      "epoch": 0.48647746243739565,
+      "grad_norm": 0.5497180805577743,
+      "learning_rate": 9.387918813693805e-06,
+      "loss": 0.3012,
+      "step": 1457
+    },
+    {
+      "epoch": 0.48681135225375627,
+      "grad_norm": 0.528643958621244,
+      "learning_rate": 9.386521154558347e-06,
+      "loss": 0.3001,
+      "step": 1458
+    },
+    {
+      "epoch": 0.48714524207011684,
+      "grad_norm": 0.566234102528402,
+      "learning_rate": 9.385122005782986e-06,
+      "loss": 0.2963,
+      "step": 1459
+    },
+    {
+      "epoch": 0.48747913188647746,
+      "grad_norm": 0.5694481616265379,
+      "learning_rate": 9.383721367842867e-06,
+      "loss": 0.3005,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4878130217028381,
+      "grad_norm": 0.5665379354204643,
+      "learning_rate": 9.382319241213633e-06,
+      "loss": 0.2987,
+      "step": 1461
+    },
+    {
+      "epoch": 0.48814691151919865,
+      "grad_norm": 0.5471049595952897,
+      "learning_rate": 9.380915626371445e-06,
+      "loss": 0.2935,
+      "step": 1462
+    },
+    {
+      "epoch": 0.48848080133555927,
+      "grad_norm": 0.5805556650939894,
+      "learning_rate": 9.37951052379296e-06,
+      "loss": 0.2975,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4888146911519199,
+      "grad_norm": 0.5860643992684963,
+      "learning_rate": 9.378103933955345e-06,
+      "loss": 0.2979,
+      "step": 1464
+    },
+    {
+      "epoch": 0.48914858096828046,
+      "grad_norm": 0.5538190592281108,
+      "learning_rate": 9.37669585733627e-06,
+      "loss": 0.3098,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4894824707846411,
+      "grad_norm": 0.5900086838657749,
+      "learning_rate": 9.375286294413914e-06,
+      "loss": 0.3088,
+      "step": 1466
+    },
+    {
+      "epoch": 0.48981636060100164,
+      "grad_norm": 0.55252216123743,
+      "learning_rate": 9.373875245666953e-06,
+      "loss": 0.3106,
+      "step": 1467
+    },
+    {
+      "epoch": 0.49015025041736227,
+      "grad_norm": 0.5877819339105316,
+      "learning_rate": 9.37246271157457e-06,
+      "loss": 0.3118,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4904841402337229,
+      "grad_norm": 0.5245464242318695,
+      "learning_rate": 9.371048692616461e-06,
+      "loss": 0.3032,
+      "step": 1469
+    },
+    {
+      "epoch": 0.49081803005008345,
+      "grad_norm": 0.5465668423886672,
+      "learning_rate": 9.369633189272815e-06,
+      "loss": 0.2935,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4911519198664441,
+      "grad_norm": 0.557550344343343,
+      "learning_rate": 9.368216202024329e-06,
+      "loss": 0.3033,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4914858096828047,
+      "grad_norm": 0.545839965077155,
+      "learning_rate": 9.36679773135221e-06,
+      "loss": 0.3052,
+      "step": 1472
+    },
+    {
+      "epoch": 0.49181969949916526,
+      "grad_norm": 0.5571105869085422,
+      "learning_rate": 9.365377777738155e-06,
+      "loss": 0.3055,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4921535893155259,
+      "grad_norm": 0.5506834999774838,
+      "learning_rate": 9.36395634166438e-06,
+      "loss": 0.2944,
+      "step": 1474
+    },
+    {
+      "epoch": 0.49248747913188645,
+      "grad_norm": 0.5394380530286853,
+      "learning_rate": 9.362533423613592e-06,
+      "loss": 0.3005,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4928213689482471,
+      "grad_norm": 0.5417451900071981,
+      "learning_rate": 9.361109024069011e-06,
+      "loss": 0.2911,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4931552587646077,
+      "grad_norm": 0.558560989425386,
+      "learning_rate": 9.359683143514354e-06,
+      "loss": 0.3006,
+      "step": 1477
+    },
+    {
+      "epoch": 0.49348914858096826,
+      "grad_norm": 0.5753754629211304,
+      "learning_rate": 9.358255782433842e-06,
+      "loss": 0.2948,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4938230383973289,
+      "grad_norm": 0.5213992732907958,
+      "learning_rate": 9.3568269413122e-06,
+      "loss": 0.3026,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4941569282136895,
+      "grad_norm": 0.5846961069675506,
+      "learning_rate": 9.355396620634657e-06,
+      "loss": 0.2891,
+      "step": 1480
+    },
+    {
+      "epoch": 0.49449081803005007,
+      "grad_norm": 0.5993113167585559,
+      "learning_rate": 9.353964820886938e-06,
+      "loss": 0.3023,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4948247078464107,
+      "grad_norm": 0.5404784391815952,
+      "learning_rate": 9.35253154255528e-06,
+      "loss": 0.2869,
+      "step": 1482
+    },
+    {
+      "epoch": 0.49515859766277126,
+      "grad_norm": 0.5874292027278316,
+      "learning_rate": 9.351096786126413e-06,
+      "loss": 0.288,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4954924874791319,
+      "grad_norm": 0.600236587607361,
+      "learning_rate": 9.349660552087574e-06,
+      "loss": 0.2935,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4958263772954925,
+      "grad_norm": 0.5372732086569023,
+      "learning_rate": 9.3482228409265e-06,
+      "loss": 0.302,
+      "step": 1485
+    },
+    {
+      "epoch": 0.49616026711185307,
+      "grad_norm": 0.5842747691486533,
+      "learning_rate": 9.346783653131432e-06,
+      "loss": 0.304,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4964941569282137,
+      "grad_norm": 0.6216307333866692,
+      "learning_rate": 9.34534298919111e-06,
+      "loss": 0.309,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4968280467445743,
+      "grad_norm": 0.558871468745058,
+      "learning_rate": 9.343900849594777e-06,
+      "loss": 0.2978,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4971619365609349,
+      "grad_norm": 0.5217413648411765,
+      "learning_rate": 9.342457234832175e-06,
+      "loss": 0.3018,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4974958263772955,
+      "grad_norm": 0.6450756403205696,
+      "learning_rate": 9.341012145393546e-06,
+      "loss": 0.2896,
+      "step": 1490
+    },
+    {
+      "epoch": 0.49782971619365607,
+      "grad_norm": 0.5491513347129208,
+      "learning_rate": 9.339565581769639e-06,
+      "loss": 0.3,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4981636060100167,
+      "grad_norm": 0.5523551395775559,
+      "learning_rate": 9.338117544451697e-06,
+      "loss": 0.2943,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4984974958263773,
+      "grad_norm": 0.5553364576840443,
+      "learning_rate": 9.336668033931466e-06,
+      "loss": 0.299,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4988313856427379,
+      "grad_norm": 0.5391786466786052,
+      "learning_rate": 9.335217050701193e-06,
+      "loss": 0.3134,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4991652754590985,
+      "grad_norm": 0.5745244499989205,
+      "learning_rate": 9.333764595253625e-06,
+      "loss": 0.3071,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4994991652754591,
+      "grad_norm": 0.5217799472805608,
+      "learning_rate": 9.332310668082007e-06,
+      "loss": 0.2917,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4998330550918197,
+      "grad_norm": 0.5643071780093937,
+      "learning_rate": 9.330855269680085e-06,
+      "loss": 0.3067,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5001669449081803,
+      "grad_norm": 0.5252165447187881,
+      "learning_rate": 9.329398400542106e-06,
+      "loss": 0.3011,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5005008347245409,
+      "grad_norm": 0.5725770692873122,
+      "learning_rate": 9.327940061162816e-06,
+      "loss": 0.3051,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5008347245409015,
+      "grad_norm": 0.5583583237617101,
+      "learning_rate": 9.326480252037457e-06,
+      "loss": 0.2909,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5011686143572621,
+      "grad_norm": 0.5399773492308132,
+      "learning_rate": 9.325018973661776e-06,
+      "loss": 0.2974,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5015025041736227,
+      "grad_norm": 0.5080222378758982,
+      "learning_rate": 9.32355622653201e-06,
+      "loss": 0.2846,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5018363939899833,
+      "grad_norm": 0.5197322572208812,
+      "learning_rate": 9.322092011144907e-06,
+      "loss": 0.2924,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5021702838063439,
+      "grad_norm": 0.5266851796036764,
+      "learning_rate": 9.320626327997703e-06,
+      "loss": 0.2846,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5025041736227045,
+      "grad_norm": 0.537048027161128,
+      "learning_rate": 9.319159177588137e-06,
+      "loss": 0.2944,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5028380634390651,
+      "grad_norm": 0.5010810066642427,
+      "learning_rate": 9.317690560414445e-06,
+      "loss": 0.2912,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5031719532554257,
+      "grad_norm": 0.5446510370878416,
+      "learning_rate": 9.316220476975364e-06,
+      "loss": 0.3051,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5035058430717864,
+      "grad_norm": 0.5367129818278349,
+      "learning_rate": 9.314748927770126e-06,
+      "loss": 0.2934,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5038397328881469,
+      "grad_norm": 0.5202334951605865,
+      "learning_rate": 9.313275913298458e-06,
+      "loss": 0.2793,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5041736227045075,
+      "grad_norm": 0.5425059669322545,
+      "learning_rate": 9.311801434060594e-06,
+      "loss": 0.2841,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5045075125208681,
+      "grad_norm": 0.5485861263499934,
+      "learning_rate": 9.310325490557254e-06,
+      "loss": 0.3122,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5048414023372287,
+      "grad_norm": 0.5352189707383892,
+      "learning_rate": 9.308848083289666e-06,
+      "loss": 0.3,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5051752921535894,
+      "grad_norm": 0.5257644830619306,
+      "learning_rate": 9.307369212759547e-06,
+      "loss": 0.2945,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5055091819699499,
+      "grad_norm": 0.5487008810455634,
+      "learning_rate": 9.305888879469112e-06,
+      "loss": 0.3012,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5058430717863105,
+      "grad_norm": 0.5355154343267235,
+      "learning_rate": 9.30440708392108e-06,
+      "loss": 0.2911,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5061769616026711,
+      "grad_norm": 0.5431170992645884,
+      "learning_rate": 9.302923826618654e-06,
+      "loss": 0.2865,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5065108514190317,
+      "grad_norm": 0.5593447214101199,
+      "learning_rate": 9.301439108065546e-06,
+      "loss": 0.2925,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5068447412353924,
+      "grad_norm": 0.5635904094381008,
+      "learning_rate": 9.299952928765956e-06,
+      "loss": 0.2893,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5071786310517529,
+      "grad_norm": 0.5620534831137272,
+      "learning_rate": 9.298465289224585e-06,
+      "loss": 0.2912,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5075125208681135,
+      "grad_norm": 0.5662689437334766,
+      "learning_rate": 9.296976189946625e-06,
+      "loss": 0.3016,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5078464106844741,
+      "grad_norm": 0.5858398028942337,
+      "learning_rate": 9.295485631437768e-06,
+      "loss": 0.2954,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5081803005008347,
+      "grad_norm": 0.590341320624517,
+      "learning_rate": 9.293993614204202e-06,
+      "loss": 0.3084,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5085141903171954,
+      "grad_norm": 0.5309122357228887,
+      "learning_rate": 9.292500138752606e-06,
+      "loss": 0.2881,
+      "step": 1523
+    },
+    {
+      "epoch": 0.508848080133556,
+      "grad_norm": 0.5712680781508501,
+      "learning_rate": 9.291005205590153e-06,
+      "loss": 0.2937,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5091819699499165,
+      "grad_norm": 0.5216227659167598,
+      "learning_rate": 9.289508815224522e-06,
+      "loss": 0.2747,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5095158597662771,
+      "grad_norm": 0.5475199605027703,
+      "learning_rate": 9.288010968163874e-06,
+      "loss": 0.2877,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5098497495826377,
+      "grad_norm": 0.5973009167700832,
+      "learning_rate": 9.28651166491687e-06,
+      "loss": 0.2955,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5101836393989984,
+      "grad_norm": 0.539002505651918,
+      "learning_rate": 9.28501090599267e-06,
+      "loss": 0.2979,
+      "step": 1528
+    },
+    {
+      "epoch": 0.510517529215359,
+      "grad_norm": 0.5392262757899636,
+      "learning_rate": 9.283508691900922e-06,
+      "loss": 0.288,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5108514190317195,
+      "grad_norm": 0.5047808907035247,
+      "learning_rate": 9.282005023151767e-06,
+      "loss": 0.2881,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5111853088480801,
+      "grad_norm": 0.5590044530271029,
+      "learning_rate": 9.280499900255848e-06,
+      "loss": 0.308,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5115191986644407,
+      "grad_norm": 0.5260840364113362,
+      "learning_rate": 9.278993323724293e-06,
+      "loss": 0.2904,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5118530884808014,
+      "grad_norm": 0.5233147920945581,
+      "learning_rate": 9.27748529406873e-06,
+      "loss": 0.2994,
+      "step": 1533
+    },
+    {
+      "epoch": 0.512186978297162,
+      "grad_norm": 0.5099247864299015,
+      "learning_rate": 9.275975811801278e-06,
+      "loss": 0.2925,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5125208681135225,
+      "grad_norm": 0.5550493480831713,
+      "learning_rate": 9.274464877434548e-06,
+      "loss": 0.3087,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5128547579298831,
+      "grad_norm": 0.5536178460815789,
+      "learning_rate": 9.272952491481646e-06,
+      "loss": 0.3118,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5131886477462437,
+      "grad_norm": 0.542821652130094,
+      "learning_rate": 9.271438654456171e-06,
+      "loss": 0.2901,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5135225375626044,
+      "grad_norm": 0.5367812070339708,
+      "learning_rate": 9.269923366872213e-06,
+      "loss": 0.2798,
+      "step": 1538
+    },
+    {
+      "epoch": 0.513856427378965,
+      "grad_norm": 0.5505593526201684,
+      "learning_rate": 9.268406629244358e-06,
+      "loss": 0.2964,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5141903171953256,
+      "grad_norm": 0.5270120577692455,
+      "learning_rate": 9.266888442087681e-06,
+      "loss": 0.2939,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5145242070116861,
+      "grad_norm": 0.5329492251795381,
+      "learning_rate": 9.26536880591775e-06,
+      "loss": 0.2913,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5148580968280467,
+      "grad_norm": 0.5095761967534578,
+      "learning_rate": 9.263847721250624e-06,
+      "loss": 0.2845,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5151919866444074,
+      "grad_norm": 0.5092377963564793,
+      "learning_rate": 9.262325188602861e-06,
+      "loss": 0.2856,
+      "step": 1543
+    },
+    {
+      "epoch": 0.515525876460768,
+      "grad_norm": 0.515168119241561,
+      "learning_rate": 9.260801208491499e-06,
+      "loss": 0.3011,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5158597662771286,
+      "grad_norm": 0.5553773215164958,
+      "learning_rate": 9.259275781434079e-06,
+      "loss": 0.2887,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5161936560934891,
+      "grad_norm": 0.502974539660859,
+      "learning_rate": 9.257748907948623e-06,
+      "loss": 0.2976,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5165275459098497,
+      "grad_norm": 0.5382781838263281,
+      "learning_rate": 9.256220588553653e-06,
+      "loss": 0.3102,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5168614357262103,
+      "grad_norm": 0.5377500503515952,
+      "learning_rate": 9.254690823768178e-06,
+      "loss": 0.2956,
+      "step": 1548
+    },
+    {
+      "epoch": 0.517195325542571,
+      "grad_norm": 0.5528344552784087,
+      "learning_rate": 9.253159614111698e-06,
+      "loss": 0.2925,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5175292153589316,
+      "grad_norm": 0.5328919261228519,
+      "learning_rate": 9.251626960104201e-06,
+      "loss": 0.2932,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5178631051752921,
+      "grad_norm": 0.5202036274322469,
+      "learning_rate": 9.250092862266174e-06,
+      "loss": 0.2954,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5181969949916527,
+      "grad_norm": 0.5310715586202158,
+      "learning_rate": 9.248557321118583e-06,
+      "loss": 0.2884,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5185308848080133,
+      "grad_norm": 0.5631506619960539,
+      "learning_rate": 9.247020337182892e-06,
+      "loss": 0.3031,
+      "step": 1553
+    },
+    {
+      "epoch": 0.518864774624374,
+      "grad_norm": 0.5457316542484535,
+      "learning_rate": 9.245481910981056e-06,
+      "loss": 0.3017,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5191986644407346,
+      "grad_norm": 0.5245835677697122,
+      "learning_rate": 9.243942043035513e-06,
+      "loss": 0.2906,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5195325542570952,
+      "grad_norm": 0.5262475889076998,
+      "learning_rate": 9.242400733869193e-06,
+      "loss": 0.295,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5198664440734557,
+      "grad_norm": 0.561936432669135,
+      "learning_rate": 9.240857984005521e-06,
+      "loss": 0.3002,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5202003338898163,
+      "grad_norm": 0.5321551179489142,
+      "learning_rate": 9.239313793968402e-06,
+      "loss": 0.2833,
+      "step": 1558
+    },
+    {
+      "epoch": 0.520534223706177,
+      "grad_norm": 0.534961809615907,
+      "learning_rate": 9.237768164282239e-06,
+      "loss": 0.2993,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5208681135225376,
+      "grad_norm": 0.5554059928782842,
+      "learning_rate": 9.236221095471918e-06,
+      "loss": 0.2996,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5212020033388982,
+      "grad_norm": 0.5409175001875158,
+      "learning_rate": 9.234672588062813e-06,
+      "loss": 0.295,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5215358931552587,
+      "grad_norm": 0.5109485708794117,
+      "learning_rate": 9.233122642580796e-06,
+      "loss": 0.2972,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5218697829716193,
+      "grad_norm": 0.4942442249705856,
+      "learning_rate": 9.231571259552214e-06,
+      "loss": 0.2871,
+      "step": 1563
+    },
+    {
+      "epoch": 0.52220367278798,
+      "grad_norm": 0.5164828426726152,
+      "learning_rate": 9.230018439503913e-06,
+      "loss": 0.2901,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5225375626043406,
+      "grad_norm": 0.5108940193129852,
+      "learning_rate": 9.22846418296322e-06,
+      "loss": 0.2948,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5228714524207012,
+      "grad_norm": 0.5538778448187159,
+      "learning_rate": 9.226908490457952e-06,
+      "loss": 0.2805,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5232053422370617,
+      "grad_norm": 0.5496025696674337,
+      "learning_rate": 9.225351362516418e-06,
+      "loss": 0.2879,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5235392320534223,
+      "grad_norm": 0.5299733372571178,
+      "learning_rate": 9.223792799667403e-06,
+      "loss": 0.2897,
+      "step": 1568
+    },
+    {
+      "epoch": 0.523873121869783,
+      "grad_norm": 0.5524058588798325,
+      "learning_rate": 9.222232802440197e-06,
+      "loss": 0.2904,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5242070116861436,
+      "grad_norm": 0.538132792807304,
+      "learning_rate": 9.220671371364559e-06,
+      "loss": 0.2909,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5245409015025042,
+      "grad_norm": 0.5431348117889244,
+      "learning_rate": 9.219108506970747e-06,
+      "loss": 0.2957,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5248747913188647,
+      "grad_norm": 0.5297573825526544,
+      "learning_rate": 9.2175442097895e-06,
+      "loss": 0.3033,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5252086811352253,
+      "grad_norm": 0.5169348324085431,
+      "learning_rate": 9.215978480352043e-06,
+      "loss": 0.2884,
+      "step": 1573
+    },
+    {
+      "epoch": 0.525542570951586,
+      "grad_norm": 0.5341098667339907,
+      "learning_rate": 9.214411319190095e-06,
+      "loss": 0.2959,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5258764607679466,
+      "grad_norm": 0.5234158677346797,
+      "learning_rate": 9.212842726835854e-06,
+      "loss": 0.2757,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5262103505843072,
+      "grad_norm": 0.536611233354249,
+      "learning_rate": 9.211272703822003e-06,
+      "loss": 0.3029,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5265442404006678,
+      "grad_norm": 0.5029220259967321,
+      "learning_rate": 9.209701250681715e-06,
+      "loss": 0.2942,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5268781302170283,
+      "grad_norm": 0.5466838293157125,
+      "learning_rate": 9.20812836794865e-06,
+      "loss": 0.2886,
+      "step": 1578
+    },
+    {
+      "epoch": 0.527212020033389,
+      "grad_norm": 0.5674236384548486,
+      "learning_rate": 9.206554056156948e-06,
+      "loss": 0.2939,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5275459098497496,
+      "grad_norm": 0.527636978425116,
+      "learning_rate": 9.204978315841238e-06,
+      "loss": 0.2941,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5278797996661102,
+      "grad_norm": 0.5409962333373036,
+      "learning_rate": 9.203401147536633e-06,
+      "loss": 0.3055,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5282136894824708,
+      "grad_norm": 0.5525295183775997,
+      "learning_rate": 9.201822551778731e-06,
+      "loss": 0.3026,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5285475792988313,
+      "grad_norm": 0.5194061139025133,
+      "learning_rate": 9.200242529103615e-06,
+      "loss": 0.2846,
+      "step": 1583
+    },
+    {
+      "epoch": 0.528881469115192,
+      "grad_norm": 0.5539337744665075,
+      "learning_rate": 9.198661080047855e-06,
+      "loss": 0.2956,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5292153589315526,
+      "grad_norm": 0.5236501805660891,
+      "learning_rate": 9.1970782051485e-06,
+      "loss": 0.2828,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5295492487479132,
+      "grad_norm": 0.5328558151673959,
+      "learning_rate": 9.195493904943087e-06,
+      "loss": 0.3064,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5298831385642738,
+      "grad_norm": 0.5256373043501402,
+      "learning_rate": 9.193908179969635e-06,
+      "loss": 0.2852,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5302170283806343,
+      "grad_norm": 0.5504484713695529,
+      "learning_rate": 9.192321030766652e-06,
+      "loss": 0.292,
+      "step": 1588
+    },
+    {
+      "epoch": 0.530550918196995,
+      "grad_norm": 0.5060405940286877,
+      "learning_rate": 9.19073245787312e-06,
+      "loss": 0.2991,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5308848080133556,
+      "grad_norm": 0.5654823780547457,
+      "learning_rate": 9.189142461828512e-06,
+      "loss": 0.3006,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5312186978297162,
+      "grad_norm": 0.5184936028646476,
+      "learning_rate": 9.187551043172787e-06,
+      "loss": 0.3082,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5315525876460768,
+      "grad_norm": 0.5544507995070184,
+      "learning_rate": 9.185958202446375e-06,
+      "loss": 0.2943,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5318864774624374,
+      "grad_norm": 0.5795802077446741,
+      "learning_rate": 9.184363940190204e-06,
+      "loss": 0.3092,
+      "step": 1593
+    },
+    {
+      "epoch": 0.532220367278798,
+      "grad_norm": 0.5147343837190675,
+      "learning_rate": 9.182768256945671e-06,
+      "loss": 0.2874,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5325542570951586,
+      "grad_norm": 0.5331646228644329,
+      "learning_rate": 9.181171153254666e-06,
+      "loss": 0.3017,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5328881469115192,
+      "grad_norm": 0.5551884968084929,
+      "learning_rate": 9.179572629659554e-06,
+      "loss": 0.2918,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5332220367278798,
+      "grad_norm": 0.5239794442137014,
+      "learning_rate": 9.17797268670319e-06,
+      "loss": 0.2906,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5335559265442404,
+      "grad_norm": 0.5154492503762258,
+      "learning_rate": 9.1763713249289e-06,
+      "loss": 0.2931,
+      "step": 1598
+    },
+    {
+      "epoch": 0.533889816360601,
+      "grad_norm": 0.5228756914801239,
+      "learning_rate": 9.174768544880503e-06,
+      "loss": 0.2881,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5342237061769616,
+      "grad_norm": 0.5466533471008379,
+      "learning_rate": 9.173164347102293e-06,
+      "loss": 0.2764,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5345575959933222,
+      "grad_norm": 0.5309789745270768,
+      "learning_rate": 9.171558732139046e-06,
+      "loss": 0.2861,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5348914858096828,
+      "grad_norm": 0.5341166500973623,
+      "learning_rate": 9.169951700536024e-06,
+      "loss": 0.2844,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5352253756260434,
+      "grad_norm": 0.5308667343245688,
+      "learning_rate": 9.168343252838963e-06,
+      "loss": 0.2937,
+      "step": 1603
+    },
+    {
+      "epoch": 0.535559265442404,
+      "grad_norm": 0.5210220559741984,
+      "learning_rate": 9.166733389594087e-06,
+      "loss": 0.2803,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5358931552587646,
+      "grad_norm": 0.54434653778892,
+      "learning_rate": 9.165122111348095e-06,
+      "loss": 0.3058,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5362270450751252,
+      "grad_norm": 0.5357790685303268,
+      "learning_rate": 9.16350941864817e-06,
+      "loss": 0.291,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5365609348914858,
+      "grad_norm": 0.5550002729590755,
+      "learning_rate": 9.16189531204197e-06,
+      "loss": 0.3064,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5368948247078464,
+      "grad_norm": 0.5097476492175045,
+      "learning_rate": 9.160279792077645e-06,
+      "loss": 0.2907,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5372287145242071,
+      "grad_norm": 0.5152994331884875,
+      "learning_rate": 9.158662859303809e-06,
+      "loss": 0.3043,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5375626043405676,
+      "grad_norm": 0.5251750548612713,
+      "learning_rate": 9.15704451426957e-06,
+      "loss": 0.2968,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5378964941569282,
+      "grad_norm": 0.5387633783029494,
+      "learning_rate": 9.155424757524505e-06,
+      "loss": 0.3074,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5382303839732888,
+      "grad_norm": 0.5323011757010737,
+      "learning_rate": 9.153803589618681e-06,
+      "loss": 0.2969,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5385642737896494,
+      "grad_norm": 0.5200945093252465,
+      "learning_rate": 9.152181011102632e-06,
+      "loss": 0.2964,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5388981636060101,
+      "grad_norm": 0.5746564234323394,
+      "learning_rate": 9.150557022527382e-06,
+      "loss": 0.2996,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5392320534223706,
+      "grad_norm": 0.5687697374188031,
+      "learning_rate": 9.148931624444425e-06,
+      "loss": 0.2934,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5395659432387312,
+      "grad_norm": 0.5172789818842796,
+      "learning_rate": 9.147304817405741e-06,
+      "loss": 0.2864,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5398998330550918,
+      "grad_norm": 0.5473343420216503,
+      "learning_rate": 9.145676601963783e-06,
+      "loss": 0.2948,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5402337228714524,
+      "grad_norm": 0.5515702584471417,
+      "learning_rate": 9.144046978671486e-06,
+      "loss": 0.298,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5405676126878131,
+      "grad_norm": 0.508015713753843,
+      "learning_rate": 9.142415948082264e-06,
+      "loss": 0.2936,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5409015025041736,
+      "grad_norm": 0.5304674261243852,
+      "learning_rate": 9.14078351075e-06,
+      "loss": 0.2979,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5412353923205342,
+      "grad_norm": 0.4989068485008668,
+      "learning_rate": 9.139149667229067e-06,
+      "loss": 0.2866,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5415692821368948,
+      "grad_norm": 0.5065769483196985,
+      "learning_rate": 9.137514418074307e-06,
+      "loss": 0.2784,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5419031719532554,
+      "grad_norm": 0.5365197851088195,
+      "learning_rate": 9.135877763841046e-06,
+      "loss": 0.2958,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5422370617696161,
+      "grad_norm": 0.541181683200849,
+      "learning_rate": 9.134239705085077e-06,
+      "loss": 0.283,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5425709515859767,
+      "grad_norm": 0.5271195585881349,
+      "learning_rate": 9.132600242362682e-06,
+      "loss": 0.279,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5429048414023372,
+      "grad_norm": 0.5289506279029171,
+      "learning_rate": 9.13095937623061e-06,
+      "loss": 0.2943,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5432387312186978,
+      "grad_norm": 0.544662368207497,
+      "learning_rate": 9.129317107246091e-06,
+      "loss": 0.3049,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5435726210350584,
+      "grad_norm": 0.526296386386341,
+      "learning_rate": 9.127673435966837e-06,
+      "loss": 0.2872,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5439065108514191,
+      "grad_norm": 0.5293164228531773,
+      "learning_rate": 9.126028362951023e-06,
+      "loss": 0.2882,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5442404006677797,
+      "grad_norm": 0.5232302291797787,
+      "learning_rate": 9.124381888757309e-06,
+      "loss": 0.3024,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5445742904841402,
+      "grad_norm": 0.5141304222438831,
+      "learning_rate": 9.122734013944832e-06,
+      "loss": 0.2855,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5449081803005008,
+      "grad_norm": 0.526470732203583,
+      "learning_rate": 9.1210847390732e-06,
+      "loss": 0.2888,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5452420701168614,
+      "grad_norm": 0.5172187989501033,
+      "learning_rate": 9.119434064702496e-06,
+      "loss": 0.2933,
+      "step": 1633
+    },
+    {
+      "epoch": 0.545575959933222,
+      "grad_norm": 0.5302389405206853,
+      "learning_rate": 9.117781991393285e-06,
+      "loss": 0.2879,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5459098497495827,
+      "grad_norm": 0.47805114453584097,
+      "learning_rate": 9.1161285197066e-06,
+      "loss": 0.2907,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5462437395659432,
+      "grad_norm": 0.5284698485504173,
+      "learning_rate": 9.114473650203949e-06,
+      "loss": 0.3042,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5465776293823038,
+      "grad_norm": 0.5091192007711065,
+      "learning_rate": 9.112817383447323e-06,
+      "loss": 0.2948,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5469115191986644,
+      "grad_norm": 0.7159428293348,
+      "learning_rate": 9.111159719999176e-06,
+      "loss": 0.2981,
+      "step": 1638
+    },
+    {
+      "epoch": 0.547245409015025,
+      "grad_norm": 0.489983149437252,
+      "learning_rate": 9.109500660422448e-06,
+      "loss": 0.289,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5475792988313857,
+      "grad_norm": 0.549009961338341,
+      "learning_rate": 9.10784020528054e-06,
+      "loss": 0.3134,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5479131886477463,
+      "grad_norm": 0.5031833366871177,
+      "learning_rate": 9.106178355137338e-06,
+      "loss": 0.2922,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5482470784641068,
+      "grad_norm": 0.5251338568807088,
+      "learning_rate": 9.104515110557197e-06,
+      "loss": 0.2923,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5485809682804674,
+      "grad_norm": 0.5299790595155648,
+      "learning_rate": 9.102850472104945e-06,
+      "loss": 0.3054,
+      "step": 1643
+    },
+    {
+      "epoch": 0.548914858096828,
+      "grad_norm": 0.5278058924583283,
+      "learning_rate": 9.101184440345886e-06,
+      "loss": 0.2942,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5492487479131887,
+      "grad_norm": 0.515521464915241,
+      "learning_rate": 9.099517015845799e-06,
+      "loss": 0.2984,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5495826377295493,
+      "grad_norm": 0.5061959834199731,
+      "learning_rate": 9.097848199170928e-06,
+      "loss": 0.2964,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5499165275459098,
+      "grad_norm": 0.5159183339981025,
+      "learning_rate": 9.096177990887995e-06,
+      "loss": 0.3145,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5502504173622704,
+      "grad_norm": 0.5484183906963813,
+      "learning_rate": 9.094506391564196e-06,
+      "loss": 0.3004,
+      "step": 1648
+    },
+    {
+      "epoch": 0.550584307178631,
+      "grad_norm": 0.5171525651231725,
+      "learning_rate": 9.092833401767194e-06,
+      "loss": 0.2958,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5509181969949917,
+      "grad_norm": 0.5154261413716101,
+      "learning_rate": 9.091159022065134e-06,
+      "loss": 0.2841,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5512520868113523,
+      "grad_norm": 0.5004890725071637,
+      "learning_rate": 9.08948325302662e-06,
+      "loss": 0.2915,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5515859766277128,
+      "grad_norm": 0.5260905610791475,
+      "learning_rate": 9.08780609522074e-06,
+      "loss": 0.3029,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5519198664440734,
+      "grad_norm": 0.4972204290959787,
+      "learning_rate": 9.086127549217044e-06,
+      "loss": 0.2893,
+      "step": 1653
+    },
+    {
+      "epoch": 0.552253756260434,
+      "grad_norm": 0.5373730265380743,
+      "learning_rate": 9.08444761558556e-06,
+      "loss": 0.3022,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5525876460767947,
+      "grad_norm": 0.47859343650788894,
+      "learning_rate": 9.082766294896781e-06,
+      "loss": 0.2814,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5529215358931553,
+      "grad_norm": 0.5294868607329991,
+      "learning_rate": 9.08108358772168e-06,
+      "loss": 0.2948,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5532554257095159,
+      "grad_norm": 0.5177259238125063,
+      "learning_rate": 9.079399494631693e-06,
+      "loss": 0.2902,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5535893155258764,
+      "grad_norm": 0.5025529027940995,
+      "learning_rate": 9.077714016198729e-06,
+      "loss": 0.296,
+      "step": 1658
+    },
+    {
+      "epoch": 0.553923205342237,
+      "grad_norm": 0.5371096293305234,
+      "learning_rate": 9.07602715299517e-06,
+      "loss": 0.2993,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5542570951585977,
+      "grad_norm": 0.511533632965308,
+      "learning_rate": 9.074338905593862e-06,
+      "loss": 0.2934,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5545909849749583,
+      "grad_norm": 0.5330694969829437,
+      "learning_rate": 9.07264927456813e-06,
+      "loss": 0.2966,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5549248747913189,
+      "grad_norm": 0.49386015516734555,
+      "learning_rate": 9.070958260491761e-06,
+      "loss": 0.2841,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5552587646076794,
+      "grad_norm": 0.5164966332885456,
+      "learning_rate": 9.069265863939012e-06,
+      "loss": 0.3013,
+      "step": 1663
+    },
+    {
+      "epoch": 0.55559265442404,
+      "grad_norm": 0.5130514537237197,
+      "learning_rate": 9.067572085484618e-06,
+      "loss": 0.2924,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5559265442404007,
+      "grad_norm": 0.49909674315574354,
+      "learning_rate": 9.065876925703775e-06,
+      "loss": 0.2827,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5562604340567613,
+      "grad_norm": 0.5236057548908438,
+      "learning_rate": 9.064180385172148e-06,
+      "loss": 0.2995,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5565943238731219,
+      "grad_norm": 0.5162496213024012,
+      "learning_rate": 9.062482464465877e-06,
+      "loss": 0.2918,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5569282136894824,
+      "grad_norm": 0.5141191673890282,
+      "learning_rate": 9.060783164161567e-06,
+      "loss": 0.2941,
+      "step": 1668
+    },
+    {
+      "epoch": 0.557262103505843,
+      "grad_norm": 0.5048045554239935,
+      "learning_rate": 9.059082484836287e-06,
+      "loss": 0.2765,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5575959933222037,
+      "grad_norm": 0.5185270718843973,
+      "learning_rate": 9.057380427067584e-06,
+      "loss": 0.2955,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5579298831385643,
+      "grad_norm": 0.5268049766920692,
+      "learning_rate": 9.055676991433467e-06,
+      "loss": 0.2938,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5582637729549249,
+      "grad_norm": 0.5241039704631429,
+      "learning_rate": 9.05397217851241e-06,
+      "loss": 0.3095,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5585976627712855,
+      "grad_norm": 0.5137549714941346,
+      "learning_rate": 9.052265988883365e-06,
+      "loss": 0.2919,
+      "step": 1673
+    },
+    {
+      "epoch": 0.558931552587646,
+      "grad_norm": 0.49489609653303823,
+      "learning_rate": 9.050558423125742e-06,
+      "loss": 0.2826,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5592654424040067,
+      "grad_norm": 0.5306693113894287,
+      "learning_rate": 9.048849481819418e-06,
+      "loss": 0.287,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5595993322203673,
+      "grad_norm": 0.5250453339891665,
+      "learning_rate": 9.047139165544748e-06,
+      "loss": 0.2914,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5599332220367279,
+      "grad_norm": 0.5098557698474278,
+      "learning_rate": 9.04542747488254e-06,
+      "loss": 0.3089,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5602671118530885,
+      "grad_norm": 0.5243715338381985,
+      "learning_rate": 9.043714410414078e-06,
+      "loss": 0.294,
+      "step": 1678
+    },
+    {
+      "epoch": 0.560601001669449,
+      "grad_norm": 0.5424390695811396,
+      "learning_rate": 9.04199997272111e-06,
+      "loss": 0.2973,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5609348914858097,
+      "grad_norm": 0.5204986922051797,
+      "learning_rate": 9.04028416238585e-06,
+      "loss": 0.3049,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5612687813021703,
+      "grad_norm": 0.5173243807807231,
+      "learning_rate": 9.038566979990978e-06,
+      "loss": 0.293,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5616026711185309,
+      "grad_norm": 0.5426442248781633,
+      "learning_rate": 9.03684842611964e-06,
+      "loss": 0.2947,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5619365609348915,
+      "grad_norm": 0.5156448649685081,
+      "learning_rate": 9.03512850135545e-06,
+      "loss": 0.2881,
+      "step": 1683
+    },
+    {
+      "epoch": 0.562270450751252,
+      "grad_norm": 0.5158698222610877,
+      "learning_rate": 9.033407206282484e-06,
+      "loss": 0.2953,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5626043405676127,
+      "grad_norm": 0.47038150020955677,
+      "learning_rate": 9.031684541485285e-06,
+      "loss": 0.2794,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5629382303839733,
+      "grad_norm": 0.4854410597936983,
+      "learning_rate": 9.029960507548858e-06,
+      "loss": 0.2911,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5632721202003339,
+      "grad_norm": 0.5249167214214256,
+      "learning_rate": 9.02823510505868e-06,
+      "loss": 0.2904,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5636060100166945,
+      "grad_norm": 0.5261825029535506,
+      "learning_rate": 9.02650833460069e-06,
+      "loss": 0.2957,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5639398998330551,
+      "grad_norm": 0.5185675223951558,
+      "learning_rate": 9.024780196761287e-06,
+      "loss": 0.2934,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5642737896494157,
+      "grad_norm": 0.4903967928893247,
+      "learning_rate": 9.02305069212734e-06,
+      "loss": 0.2954,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5646076794657763,
+      "grad_norm": 0.5305183243288658,
+      "learning_rate": 9.021319821286176e-06,
+      "loss": 0.3006,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5649415692821369,
+      "grad_norm": 0.5487777506879479,
+      "learning_rate": 9.019587584825596e-06,
+      "loss": 0.2942,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5652754590984975,
+      "grad_norm": 0.5115990375204026,
+      "learning_rate": 9.01785398333385e-06,
+      "loss": 0.298,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5656093489148581,
+      "grad_norm": 0.48527211329659203,
+      "learning_rate": 9.01611901739967e-06,
+      "loss": 0.2898,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5659432387312187,
+      "grad_norm": 0.5064001740472893,
+      "learning_rate": 9.014382687612235e-06,
+      "loss": 0.2921,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5662771285475793,
+      "grad_norm": 0.513765081944085,
+      "learning_rate": 9.012644994561193e-06,
+      "loss": 0.2963,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5666110183639399,
+      "grad_norm": 0.5344073387802407,
+      "learning_rate": 9.01090593883666e-06,
+      "loss": 0.2812,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5669449081803005,
+      "grad_norm": 0.506638123310359,
+      "learning_rate": 9.009165521029206e-06,
+      "loss": 0.2981,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5672787979966611,
+      "grad_norm": 0.5129115661654238,
+      "learning_rate": 9.007423741729873e-06,
+      "loss": 0.2756,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5676126878130217,
+      "grad_norm": 0.5201515140717153,
+      "learning_rate": 9.005680601530154e-06,
+      "loss": 0.2968,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5679465776293823,
+      "grad_norm": 0.5065735858673918,
+      "learning_rate": 9.003936101022015e-06,
+      "loss": 0.2828,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5682804674457429,
+      "grad_norm": 0.5505144754408697,
+      "learning_rate": 9.002190240797878e-06,
+      "loss": 0.3016,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5686143572621035,
+      "grad_norm": 0.5225669753647894,
+      "learning_rate": 9.000443021450627e-06,
+      "loss": 0.2785,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5689482470784641,
+      "grad_norm": 0.5021288834057651,
+      "learning_rate": 8.998694443573612e-06,
+      "loss": 0.2887,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5692821368948247,
+      "grad_norm": 0.504828110054851,
+      "learning_rate": 8.996944507760636e-06,
+      "loss": 0.2986,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5696160267111853,
+      "grad_norm": 0.5458520853992889,
+      "learning_rate": 8.995193214605972e-06,
+      "loss": 0.294,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5699499165275459,
+      "grad_norm": 0.5357164501401809,
+      "learning_rate": 8.99344056470435e-06,
+      "loss": 0.2982,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5702838063439065,
+      "grad_norm": 0.5348366763882592,
+      "learning_rate": 8.991686558650959e-06,
+      "loss": 0.2921,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5706176961602671,
+      "grad_norm": 0.5634104054240997,
+      "learning_rate": 8.989931197041451e-06,
+      "loss": 0.29,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5709515859766278,
+      "grad_norm": 0.48516421279678495,
+      "learning_rate": 8.988174480471937e-06,
+      "loss": 0.293,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5712854757929883,
+      "grad_norm": 0.5407061831234591,
+      "learning_rate": 8.98641640953899e-06,
+      "loss": 0.2949,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5716193656093489,
+      "grad_norm": 0.5355607708982095,
+      "learning_rate": 8.984656984839643e-06,
+      "loss": 0.3048,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5719532554257095,
+      "grad_norm": 0.5680002445866251,
+      "learning_rate": 8.982896206971385e-06,
+      "loss": 0.3029,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5722871452420701,
+      "grad_norm": 0.5494812363396626,
+      "learning_rate": 8.981134076532169e-06,
+      "loss": 0.3002,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5726210350584308,
+      "grad_norm": 0.5515375355715046,
+      "learning_rate": 8.979370594120404e-06,
+      "loss": 0.3001,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5729549248747913,
+      "grad_norm": 0.5151897634815344,
+      "learning_rate": 8.97760576033496e-06,
+      "loss": 0.2998,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5732888146911519,
+      "grad_norm": 0.5404010699288314,
+      "learning_rate": 8.975839575775165e-06,
+      "loss": 0.307,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5736227045075125,
+      "grad_norm": 0.5180375261389119,
+      "learning_rate": 8.974072041040807e-06,
+      "loss": 0.2939,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5739565943238731,
+      "grad_norm": 0.5311837068275759,
+      "learning_rate": 8.972303156732131e-06,
+      "loss": 0.2977,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5742904841402338,
+      "grad_norm": 0.5312410384339422,
+      "learning_rate": 8.970532923449841e-06,
+      "loss": 0.2935,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5746243739565943,
+      "grad_norm": 0.520286205629817,
+      "learning_rate": 8.968761341795102e-06,
+      "loss": 0.2868,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5749582637729549,
+      "grad_norm": 0.532971282427156,
+      "learning_rate": 8.966988412369528e-06,
+      "loss": 0.2976,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5752921535893155,
+      "grad_norm": 0.5065064955542917,
+      "learning_rate": 8.965214135775202e-06,
+      "loss": 0.287,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5756260434056761,
+      "grad_norm": 0.516557923818658,
+      "learning_rate": 8.963438512614657e-06,
+      "loss": 0.2854,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5759599332220368,
+      "grad_norm": 0.5040912454125666,
+      "learning_rate": 8.961661543490884e-06,
+      "loss": 0.2926,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5762938230383974,
+      "grad_norm": 0.5534852677518909,
+      "learning_rate": 8.959883229007335e-06,
+      "loss": 0.3003,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5766277128547579,
+      "grad_norm": 0.5486410602300661,
+      "learning_rate": 8.958103569767915e-06,
+      "loss": 0.293,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5769616026711185,
+      "grad_norm": 0.5223405971334065,
+      "learning_rate": 8.956322566376989e-06,
+      "loss": 0.3022,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5772954924874791,
+      "grad_norm": 0.5155447991363801,
+      "learning_rate": 8.954540219439374e-06,
+      "loss": 0.2867,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5776293823038398,
+      "grad_norm": 0.5249093147026593,
+      "learning_rate": 8.952756529560346e-06,
+      "loss": 0.296,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5779632721202004,
+      "grad_norm": 0.5189031950071447,
+      "learning_rate": 8.95097149734564e-06,
+      "loss": 0.295,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5782971619365609,
+      "grad_norm": 0.48844797483753044,
+      "learning_rate": 8.94918512340144e-06,
+      "loss": 0.2952,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5786310517529215,
+      "grad_norm": 0.5569468625866528,
+      "learning_rate": 8.94739740833439e-06,
+      "loss": 0.2911,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5789649415692821,
+      "grad_norm": 0.5654101760216309,
+      "learning_rate": 8.945608352751592e-06,
+      "loss": 0.2893,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5792988313856428,
+      "grad_norm": 0.5251867504842971,
+      "learning_rate": 8.943817957260596e-06,
+      "loss": 0.2992,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5796327212020034,
+      "grad_norm": 0.5511538429038318,
+      "learning_rate": 8.942026222469414e-06,
+      "loss": 0.3043,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5799666110183639,
+      "grad_norm": 0.5126968513988768,
+      "learning_rate": 8.940233148986507e-06,
+      "loss": 0.284,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5803005008347245,
+      "grad_norm": 0.49149087290257065,
+      "learning_rate": 8.938438737420797e-06,
+      "loss": 0.2724,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5806343906510851,
+      "grad_norm": 0.5245156465060361,
+      "learning_rate": 8.936642988381652e-06,
+      "loss": 0.3012,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5809682804674458,
+      "grad_norm": 0.5525047384380505,
+      "learning_rate": 8.934845902478903e-06,
+      "loss": 0.2935,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5813021702838064,
+      "grad_norm": 0.5368830560370447,
+      "learning_rate": 8.933047480322832e-06,
+      "loss": 0.2898,
+      "step": 1741
+    },
+    {
+      "epoch": 0.581636060100167,
+      "grad_norm": 0.4974433339692518,
+      "learning_rate": 8.93124772252417e-06,
+      "loss": 0.2805,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5819699499165275,
+      "grad_norm": 0.5148789936314748,
+      "learning_rate": 8.929446629694109e-06,
+      "loss": 0.2988,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5823038397328881,
+      "grad_norm": 0.5321920531898152,
+      "learning_rate": 8.927644202444287e-06,
+      "loss": 0.2882,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5826377295492488,
+      "grad_norm": 0.49790121221594114,
+      "learning_rate": 8.925840441386803e-06,
+      "loss": 0.2976,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5829716193656094,
+      "grad_norm": 0.5125457751123195,
+      "learning_rate": 8.924035347134204e-06,
+      "loss": 0.2878,
+      "step": 1746
+    },
+    {
+      "epoch": 0.58330550918197,
+      "grad_norm": 0.4841771076330561,
+      "learning_rate": 8.922228920299489e-06,
+      "loss": 0.2938,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5836393989983305,
+      "grad_norm": 0.5223395214974591,
+      "learning_rate": 8.92042116149611e-06,
+      "loss": 0.2948,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5839732888146911,
+      "grad_norm": 0.5107814683830151,
+      "learning_rate": 8.918612071337978e-06,
+      "loss": 0.2851,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5843071786310517,
+      "grad_norm": 0.5332367895737296,
+      "learning_rate": 8.916801650439446e-06,
+      "loss": 0.2821,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5846410684474124,
+      "grad_norm": 0.512798093948217,
+      "learning_rate": 8.914989899415323e-06,
+      "loss": 0.2882,
+      "step": 1751
+    },
+    {
+      "epoch": 0.584974958263773,
+      "grad_norm": 0.5107525005858141,
+      "learning_rate": 8.913176818880874e-06,
+      "loss": 0.2938,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5853088480801335,
+      "grad_norm": 0.5109022806512434,
+      "learning_rate": 8.911362409451808e-06,
+      "loss": 0.2852,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5856427378964941,
+      "grad_norm": 0.5367696336492025,
+      "learning_rate": 8.909546671744294e-06,
+      "loss": 0.2954,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5859766277128547,
+      "grad_norm": 0.502377660549589,
+      "learning_rate": 8.907729606374941e-06,
+      "loss": 0.2792,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5863105175292154,
+      "grad_norm": 0.5087887267846765,
+      "learning_rate": 8.905911213960821e-06,
+      "loss": 0.2895,
+      "step": 1756
+    },
+    {
+      "epoch": 0.586644407345576,
+      "grad_norm": 0.54136640395992,
+      "learning_rate": 8.904091495119447e-06,
+      "loss": 0.2959,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5869782971619366,
+      "grad_norm": 0.5073074198390894,
+      "learning_rate": 8.902270450468785e-06,
+      "loss": 0.2924,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5873121869782971,
+      "grad_norm": 0.4917300174936157,
+      "learning_rate": 8.900448080627255e-06,
+      "loss": 0.2915,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5876460767946577,
+      "grad_norm": 0.5289954272284237,
+      "learning_rate": 8.898624386213724e-06,
+      "loss": 0.2988,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5879799666110184,
+      "grad_norm": 0.5654704441214344,
+      "learning_rate": 8.896799367847508e-06,
+      "loss": 0.3019,
+      "step": 1761
+    },
+    {
+      "epoch": 0.588313856427379,
+      "grad_norm": 0.5289561843885225,
+      "learning_rate": 8.894973026148376e-06,
+      "loss": 0.2932,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5886477462437396,
+      "grad_norm": 0.5084317032364842,
+      "learning_rate": 8.893145361736541e-06,
+      "loss": 0.2993,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5889816360601001,
+      "grad_norm": 0.5304472272858347,
+      "learning_rate": 8.891316375232673e-06,
+      "loss": 0.291,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5893155258764607,
+      "grad_norm": 0.5211266860120053,
+      "learning_rate": 8.88948606725788e-06,
+      "loss": 0.2809,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5896494156928214,
+      "grad_norm": 0.51422824743627,
+      "learning_rate": 8.88765443843373e-06,
+      "loss": 0.297,
+      "step": 1766
+    },
+    {
+      "epoch": 0.589983305509182,
+      "grad_norm": 0.5182502934308443,
+      "learning_rate": 8.885821489382235e-06,
+      "loss": 0.2876,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5903171953255426,
+      "grad_norm": 0.5204218902545781,
+      "learning_rate": 8.883987220725852e-06,
+      "loss": 0.3024,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5906510851419031,
+      "grad_norm": 0.5529796987083383,
+      "learning_rate": 8.88215163308749e-06,
+      "loss": 0.3039,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5909849749582637,
+      "grad_norm": 0.5222198151359614,
+      "learning_rate": 8.880314727090506e-06,
+      "loss": 0.2875,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5913188647746244,
+      "grad_norm": 0.5059392664311582,
+      "learning_rate": 8.878476503358703e-06,
+      "loss": 0.2962,
+      "step": 1771
+    },
+    {
+      "epoch": 0.591652754590985,
+      "grad_norm": 0.5161274771708435,
+      "learning_rate": 8.876636962516333e-06,
+      "loss": 0.2814,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5919866444073456,
+      "grad_norm": 0.525150910428456,
+      "learning_rate": 8.874796105188095e-06,
+      "loss": 0.3008,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5923205342237062,
+      "grad_norm": 0.5679427408582293,
+      "learning_rate": 8.872953931999132e-06,
+      "loss": 0.3032,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5926544240400667,
+      "grad_norm": 0.5411704725132049,
+      "learning_rate": 8.87111044357504e-06,
+      "loss": 0.3049,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5929883138564274,
+      "grad_norm": 0.5379118985038661,
+      "learning_rate": 8.869265640541854e-06,
+      "loss": 0.3021,
+      "step": 1776
+    },
+    {
+      "epoch": 0.593322203672788,
+      "grad_norm": 0.5205295613646171,
+      "learning_rate": 8.867419523526065e-06,
+      "loss": 0.2795,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5936560934891486,
+      "grad_norm": 0.5185239414626835,
+      "learning_rate": 8.8655720931546e-06,
+      "loss": 0.3079,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5939899833055092,
+      "grad_norm": 0.5142820612008608,
+      "learning_rate": 8.863723350054838e-06,
+      "loss": 0.2939,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5943238731218697,
+      "grad_norm": 0.5083756770306124,
+      "learning_rate": 8.861873294854605e-06,
+      "loss": 0.3034,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5946577629382304,
+      "grad_norm": 0.5104062361532635,
+      "learning_rate": 8.860021928182167e-06,
+      "loss": 0.2885,
+      "step": 1781
+    },
+    {
+      "epoch": 0.594991652754591,
+      "grad_norm": 0.5169798074727808,
+      "learning_rate": 8.85816925066624e-06,
+      "loss": 0.2794,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5953255425709516,
+      "grad_norm": 0.5195533518828972,
+      "learning_rate": 8.856315262935986e-06,
+      "loss": 0.2932,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5956594323873122,
+      "grad_norm": 0.5076827121351667,
+      "learning_rate": 8.854459965621005e-06,
+      "loss": 0.2847,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5959933222036727,
+      "grad_norm": 0.5716594549681856,
+      "learning_rate": 8.852603359351348e-06,
+      "loss": 0.2926,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5963272120200334,
+      "grad_norm": 0.5555688232249454,
+      "learning_rate": 8.850745444757512e-06,
+      "loss": 0.28,
+      "step": 1786
+    },
+    {
+      "epoch": 0.596661101836394,
+      "grad_norm": 0.528622530851528,
+      "learning_rate": 8.84888622247043e-06,
+      "loss": 0.2798,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5969949916527546,
+      "grad_norm": 0.5107260675428527,
+      "learning_rate": 8.84702569312149e-06,
+      "loss": 0.2966,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5973288814691152,
+      "grad_norm": 0.5311938176006115,
+      "learning_rate": 8.84516385734251e-06,
+      "loss": 0.2888,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5976627712854758,
+      "grad_norm": 0.5979177059030855,
+      "learning_rate": 8.84330071576577e-06,
+      "loss": 0.2952,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5979966611018364,
+      "grad_norm": 0.535547828911881,
+      "learning_rate": 8.841436269023978e-06,
+      "loss": 0.2813,
+      "step": 1791
+    },
+    {
+      "epoch": 0.598330550918197,
+      "grad_norm": 0.5248611045307728,
+      "learning_rate": 8.839570517750289e-06,
+      "loss": 0.2892,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5986644407345576,
+      "grad_norm": 0.581562803366519,
+      "learning_rate": 8.837703462578304e-06,
+      "loss": 0.305,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5989983305509182,
+      "grad_norm": 0.5369896509282042,
+      "learning_rate": 8.835835104142066e-06,
+      "loss": 0.3044,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5993322203672788,
+      "grad_norm": 0.5356168831666406,
+      "learning_rate": 8.83396544307606e-06,
+      "loss": 0.2796,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5996661101836394,
+      "grad_norm": 0.5109137849630921,
+      "learning_rate": 8.832094480015211e-06,
+      "loss": 0.2799,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5293506226444836,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.2892,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6003338898163606,
+      "grad_norm": 0.5102240014671815,
+      "learning_rate": 8.82834865045091e-06,
+      "loss": 0.2912,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6006677796327212,
+      "grad_norm": 0.5310260074727396,
+      "learning_rate": 8.82647378521952e-06,
+      "loss": 0.3007,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6010016694490818,
+      "grad_norm": 0.5365592311565591,
+      "learning_rate": 8.82459762053742e-06,
+      "loss": 0.297,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6013355592654424,
+      "grad_norm": 0.5466683701205476,
+      "learning_rate": 8.82272015704174e-06,
+      "loss": 0.2828,
+      "step": 1801
+    },
+    {
+      "epoch": 0.601669449081803,
+      "grad_norm": 0.49424688221409563,
+      "learning_rate": 8.820841395370061e-06,
+      "loss": 0.2899,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6020033388981636,
+      "grad_norm": 0.5047117870838401,
+      "learning_rate": 8.818961336160399e-06,
+      "loss": 0.291,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6023372287145242,
+      "grad_norm": 0.5047013391467766,
+      "learning_rate": 8.817079980051211e-06,
+      "loss": 0.2818,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6026711185308848,
+      "grad_norm": 0.5386786740411962,
+      "learning_rate": 8.815197327681399e-06,
+      "loss": 0.2966,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6030050083472455,
+      "grad_norm": 0.5371138430689052,
+      "learning_rate": 8.8133133796903e-06,
+      "loss": 0.2996,
+      "step": 1806
+    },
+    {
+      "epoch": 0.603338898163606,
+      "grad_norm": 0.530159768842618,
+      "learning_rate": 8.811428136717694e-06,
+      "loss": 0.3042,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6036727879799666,
+      "grad_norm": 0.5470227832806351,
+      "learning_rate": 8.8095415994038e-06,
+      "loss": 0.2911,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6040066777963272,
+      "grad_norm": 0.5269360935557499,
+      "learning_rate": 8.807653768389274e-06,
+      "loss": 0.2913,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6043405676126878,
+      "grad_norm": 0.5339457082975434,
+      "learning_rate": 8.805764644315216e-06,
+      "loss": 0.2869,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6046744574290485,
+      "grad_norm": 0.4970754833631806,
+      "learning_rate": 8.803874227823161e-06,
+      "loss": 0.2908,
+      "step": 1811
+    },
+    {
+      "epoch": 0.605008347245409,
+      "grad_norm": 0.5365353273932049,
+      "learning_rate": 8.801982519555087e-06,
+      "loss": 0.2977,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6053422370617696,
+      "grad_norm": 0.5362286611743993,
+      "learning_rate": 8.800089520153406e-06,
+      "loss": 0.2836,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6056761268781302,
+      "grad_norm": 0.5345171409452153,
+      "learning_rate": 8.798195230260973e-06,
+      "loss": 0.294,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6060100166944908,
+      "grad_norm": 0.5454866579335431,
+      "learning_rate": 8.796299650521076e-06,
+      "loss": 0.3037,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6063439065108515,
+      "grad_norm": 0.5121292768047362,
+      "learning_rate": 8.794402781577448e-06,
+      "loss": 0.3021,
+      "step": 1816
+    },
+    {
+      "epoch": 0.606677796327212,
+      "grad_norm": 0.5402293338470899,
+      "learning_rate": 8.792504624074253e-06,
+      "loss": 0.3033,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6070116861435726,
+      "grad_norm": 0.5450551549582728,
+      "learning_rate": 8.790605178656097e-06,
+      "loss": 0.2944,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6073455759599332,
+      "grad_norm": 0.5440695531214077,
+      "learning_rate": 8.788704445968019e-06,
+      "loss": 0.2979,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6076794657762938,
+      "grad_norm": 0.5375654372998255,
+      "learning_rate": 8.786802426655502e-06,
+      "loss": 0.2891,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6080133555926545,
+      "grad_norm": 0.48472643054943576,
+      "learning_rate": 8.78489912136446e-06,
+      "loss": 0.2721,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6083472454090151,
+      "grad_norm": 0.5032265700803503,
+      "learning_rate": 8.782994530741245e-06,
+      "loss": 0.2859,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6086811352253756,
+      "grad_norm": 0.48145987085204217,
+      "learning_rate": 8.781088655432648e-06,
+      "loss": 0.2892,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6090150250417362,
+      "grad_norm": 0.5290616588415672,
+      "learning_rate": 8.779181496085893e-06,
+      "loss": 0.2995,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6093489148580968,
+      "grad_norm": 0.4894334159396376,
+      "learning_rate": 8.77727305334864e-06,
+      "loss": 0.2795,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6096828046744575,
+      "grad_norm": 0.507068948999562,
+      "learning_rate": 8.77536332786899e-06,
+      "loss": 0.2916,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6100166944908181,
+      "grad_norm": 0.5200922388687084,
+      "learning_rate": 8.773452320295472e-06,
+      "loss": 0.283,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6103505843071786,
+      "grad_norm": 0.511424786425708,
+      "learning_rate": 8.771540031277058e-06,
+      "loss": 0.2917,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6106844741235392,
+      "grad_norm": 0.5119575521893944,
+      "learning_rate": 8.769626461463147e-06,
+      "loss": 0.2996,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6110183639398998,
+      "grad_norm": 0.5276205538982768,
+      "learning_rate": 8.767711611503584e-06,
+      "loss": 0.3003,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6113522537562605,
+      "grad_norm": 0.5060967640603768,
+      "learning_rate": 8.765795482048635e-06,
+      "loss": 0.2933,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6116861435726211,
+      "grad_norm": 0.5202742227370262,
+      "learning_rate": 8.763878073749013e-06,
+      "loss": 0.2854,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6120200333889816,
+      "grad_norm": 0.48653209099422906,
+      "learning_rate": 8.761959387255857e-06,
+      "loss": 0.2918,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6123539232053422,
+      "grad_norm": 0.5230760682104283,
+      "learning_rate": 8.760039423220741e-06,
+      "loss": 0.2896,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6126878130217028,
+      "grad_norm": 0.5303819277392806,
+      "learning_rate": 8.758118182295682e-06,
+      "loss": 0.2998,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6130217028380635,
+      "grad_norm": 0.5245688638704021,
+      "learning_rate": 8.756195665133115e-06,
+      "loss": 0.2946,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6133555926544241,
+      "grad_norm": 0.4972936957528265,
+      "learning_rate": 8.754271872385925e-06,
+      "loss": 0.2905,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6136894824707846,
+      "grad_norm": 0.5130144274892927,
+      "learning_rate": 8.752346804707417e-06,
+      "loss": 0.2933,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6140233722871452,
+      "grad_norm": 0.5299591618073501,
+      "learning_rate": 8.750420462751334e-06,
+      "loss": 0.2962,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6143572621035058,
+      "grad_norm": 0.5054161119089154,
+      "learning_rate": 8.748492847171853e-06,
+      "loss": 0.303,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6146911519198665,
+      "grad_norm": 0.527256376700174,
+      "learning_rate": 8.746563958623584e-06,
+      "loss": 0.2947,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6150250417362271,
+      "grad_norm": 0.49341356973779776,
+      "learning_rate": 8.744633797761564e-06,
+      "loss": 0.2794,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6153589315525877,
+      "grad_norm": 0.4938637171084416,
+      "learning_rate": 8.742702365241266e-06,
+      "loss": 0.2933,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6156928213689482,
+      "grad_norm": 0.518344540979955,
+      "learning_rate": 8.7407696617186e-06,
+      "loss": 0.2787,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6160267111853088,
+      "grad_norm": 0.5155582759551715,
+      "learning_rate": 8.738835687849896e-06,
+      "loss": 0.2932,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6163606010016695,
+      "grad_norm": 0.5008046346339879,
+      "learning_rate": 8.736900444291923e-06,
+      "loss": 0.2985,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6166944908180301,
+      "grad_norm": 0.5175093831885689,
+      "learning_rate": 8.73496393170188e-06,
+      "loss": 0.2928,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6170283806343907,
+      "grad_norm": 0.47036922442554935,
+      "learning_rate": 8.733026150737398e-06,
+      "loss": 0.2753,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6173622704507512,
+      "grad_norm": 0.5050043037256268,
+      "learning_rate": 8.731087102056536e-06,
+      "loss": 0.2913,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6176961602671118,
+      "grad_norm": 0.5054496269601205,
+      "learning_rate": 8.729146786317787e-06,
+      "loss": 0.2898,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6180300500834724,
+      "grad_norm": 0.5395732751887221,
+      "learning_rate": 8.727205204180067e-06,
+      "loss": 0.2993,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6183639398998331,
+      "grad_norm": 0.4809858268414416,
+      "learning_rate": 8.725262356302732e-06,
+      "loss": 0.2841,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6186978297161937,
+      "grad_norm": 0.49821875017336875,
+      "learning_rate": 8.723318243345562e-06,
+      "loss": 0.2964,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6190317195325542,
+      "grad_norm": 0.5052951812328256,
+      "learning_rate": 8.721372865968766e-06,
+      "loss": 0.2904,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6193656093489148,
+      "grad_norm": 0.4892270060845172,
+      "learning_rate": 8.71942622483299e-06,
+      "loss": 0.284,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6196994991652754,
+      "grad_norm": 0.5254182574798381,
+      "learning_rate": 8.717478320599295e-06,
+      "loss": 0.2878,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6200333889816361,
+      "grad_norm": 0.5052083252315408,
+      "learning_rate": 8.715529153929182e-06,
+      "loss": 0.2919,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6203672787979967,
+      "grad_norm": 0.5162412357430611,
+      "learning_rate": 8.713578725484582e-06,
+      "loss": 0.2842,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6207011686143573,
+      "grad_norm": 0.5299797482973189,
+      "learning_rate": 8.711627035927848e-06,
+      "loss": 0.2988,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6210350584307178,
+      "grad_norm": 0.5022352950949878,
+      "learning_rate": 8.70967408592176e-06,
+      "loss": 0.2945,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6213689482470784,
+      "grad_norm": 0.5431712024135832,
+      "learning_rate": 8.707719876129535e-06,
+      "loss": 0.2969,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6217028380634391,
+      "grad_norm": 0.4895101471474791,
+      "learning_rate": 8.70576440721481e-06,
+      "loss": 0.2999,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6220367278797997,
+      "grad_norm": 0.48125317554187685,
+      "learning_rate": 8.703807679841651e-06,
+      "loss": 0.2829,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6223706176961603,
+      "grad_norm": 0.49641511460574506,
+      "learning_rate": 8.701849694674555e-06,
+      "loss": 0.291,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6227045075125208,
+      "grad_norm": 0.49365498501789107,
+      "learning_rate": 8.699890452378442e-06,
+      "loss": 0.2864,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6230383973288814,
+      "grad_norm": 0.5243678597443715,
+      "learning_rate": 8.697929953618663e-06,
+      "loss": 0.292,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6233722871452421,
+      "grad_norm": 0.5137814513710532,
+      "learning_rate": 8.695968199060991e-06,
+      "loss": 0.2958,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6237061769616027,
+      "grad_norm": 0.48852181308348475,
+      "learning_rate": 8.694005189371627e-06,
+      "loss": 0.2839,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6240400667779633,
+      "grad_norm": 0.518500861130519,
+      "learning_rate": 8.6920409252172e-06,
+      "loss": 0.2974,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6243739565943238,
+      "grad_norm": 0.5060435543516724,
+      "learning_rate": 8.690075407264766e-06,
+      "loss": 0.295,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6247078464106844,
+      "grad_norm": 0.4908879550497152,
+      "learning_rate": 8.6881086361818e-06,
+      "loss": 0.2847,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6250417362270451,
+      "grad_norm": 0.471621444290821,
+      "learning_rate": 8.68614061263621e-06,
+      "loss": 0.2946,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6253756260434057,
+      "grad_norm": 0.5353044403677495,
+      "learning_rate": 8.68417133729633e-06,
+      "loss": 0.2897,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6257095158597663,
+      "grad_norm": 0.5337837840260427,
+      "learning_rate": 8.68220081083091e-06,
+      "loss": 0.2951,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6260434056761269,
+      "grad_norm": 0.5091429452053128,
+      "learning_rate": 8.680229033909134e-06,
+      "loss": 0.2709,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6263772954924874,
+      "grad_norm": 0.5039911800783088,
+      "learning_rate": 8.678256007200604e-06,
+      "loss": 0.2906,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6267111853088481,
+      "grad_norm": 0.5084372865172586,
+      "learning_rate": 8.676281731375355e-06,
+      "loss": 0.2962,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6270450751252087,
+      "grad_norm": 0.5525321823756856,
+      "learning_rate": 8.674306207103836e-06,
+      "loss": 0.2881,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6273789649415693,
+      "grad_norm": 0.530747681207317,
+      "learning_rate": 8.672329435056926e-06,
+      "loss": 0.306,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6277128547579299,
+      "grad_norm": 0.5364977888272203,
+      "learning_rate": 8.670351415905928e-06,
+      "loss": 0.3036,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6280467445742904,
+      "grad_norm": 0.5019105352169145,
+      "learning_rate": 8.668372150322566e-06,
+      "loss": 0.2898,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6283806343906511,
+      "grad_norm": 0.532755378111501,
+      "learning_rate": 8.66639163897899e-06,
+      "loss": 0.2813,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6287145242070117,
+      "grad_norm": 0.542942334762994,
+      "learning_rate": 8.664409882547769e-06,
+      "loss": 0.2969,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6290484140233723,
+      "grad_norm": 0.5538879142136782,
+      "learning_rate": 8.662426881701897e-06,
+      "loss": 0.2999,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6293823038397329,
+      "grad_norm": 0.5170692524928487,
+      "learning_rate": 8.660442637114794e-06,
+      "loss": 0.2927,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6297161936560934,
+      "grad_norm": 0.5249640264872658,
+      "learning_rate": 8.658457149460296e-06,
+      "loss": 0.292,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6300500834724541,
+      "grad_norm": 0.5089988864074777,
+      "learning_rate": 8.656470419412666e-06,
+      "loss": 0.2997,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6303839732888147,
+      "grad_norm": 0.506885282776522,
+      "learning_rate": 8.654482447646586e-06,
+      "loss": 0.2821,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6307178631051753,
+      "grad_norm": 0.4881261084848336,
+      "learning_rate": 8.652493234837162e-06,
+      "loss": 0.2933,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6310517529215359,
+      "grad_norm": 0.4856821216633955,
+      "learning_rate": 8.650502781659922e-06,
+      "loss": 0.2873,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6313856427378965,
+      "grad_norm": 0.5043792656292179,
+      "learning_rate": 8.64851108879081e-06,
+      "loss": 0.2984,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6317195325542571,
+      "grad_norm": 0.5037561481293364,
+      "learning_rate": 8.646518156906198e-06,
+      "loss": 0.2881,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6320534223706177,
+      "grad_norm": 0.49593719831140126,
+      "learning_rate": 8.64452398668287e-06,
+      "loss": 0.2908,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6323873121869783,
+      "grad_norm": 0.509604231027456,
+      "learning_rate": 8.642528578798044e-06,
+      "loss": 0.3047,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6327212020033389,
+      "grad_norm": 0.5055847496607455,
+      "learning_rate": 8.640531933929344e-06,
+      "loss": 0.2858,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6330550918196995,
+      "grad_norm": 0.503851484268156,
+      "learning_rate": 8.638534052754823e-06,
+      "loss": 0.2889,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6333889816360601,
+      "grad_norm": 0.49255338709336616,
+      "learning_rate": 8.636534935952949e-06,
+      "loss": 0.2943,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6337228714524207,
+      "grad_norm": 0.4841822851747714,
+      "learning_rate": 8.634534584202615e-06,
+      "loss": 0.2849,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6340567612687813,
+      "grad_norm": 0.502440708310003,
+      "learning_rate": 8.632532998183126e-06,
+      "loss": 0.2955,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6343906510851419,
+      "grad_norm": 0.5308464722250141,
+      "learning_rate": 8.630530178574212e-06,
+      "loss": 0.2942,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6347245409015025,
+      "grad_norm": 0.5338582811396704,
+      "learning_rate": 8.628526126056023e-06,
+      "loss": 0.2941,
+      "step": 1901
+    },
+    {
+      "epoch": 0.635058430717863,
+      "grad_norm": 0.49080803119686356,
+      "learning_rate": 8.626520841309121e-06,
+      "loss": 0.3014,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6353923205342237,
+      "grad_norm": 0.5022007050843098,
+      "learning_rate": 8.62451432501449e-06,
+      "loss": 0.3042,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6357262103505843,
+      "grad_norm": 0.5115412434319847,
+      "learning_rate": 8.622506577853538e-06,
+      "loss": 0.3058,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6360601001669449,
+      "grad_norm": 0.5167787898561084,
+      "learning_rate": 8.62049760050808e-06,
+      "loss": 0.2993,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6363939899833055,
+      "grad_norm": 0.49587745197481303,
+      "learning_rate": 8.618487393660355e-06,
+      "loss": 0.296,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6367278797996662,
+      "grad_norm": 0.5432290777777322,
+      "learning_rate": 8.616475957993021e-06,
+      "loss": 0.3016,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6370617696160267,
+      "grad_norm": 0.498476722286559,
+      "learning_rate": 8.61446329418915e-06,
+      "loss": 0.2931,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6373956594323873,
+      "grad_norm": 0.4862637032865321,
+      "learning_rate": 8.612449402932233e-06,
+      "loss": 0.2945,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6377295492487479,
+      "grad_norm": 0.5142978997896477,
+      "learning_rate": 8.610434284906175e-06,
+      "loss": 0.2969,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6380634390651085,
+      "grad_norm": 0.5297165027998227,
+      "learning_rate": 8.608417940795304e-06,
+      "loss": 0.3016,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6383973288814692,
+      "grad_norm": 0.48968090519881086,
+      "learning_rate": 8.606400371284356e-06,
+      "loss": 0.2875,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6387312186978297,
+      "grad_norm": 0.5211553259721099,
+      "learning_rate": 8.604381577058486e-06,
+      "loss": 0.2884,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6390651085141903,
+      "grad_norm": 0.5028698516504979,
+      "learning_rate": 8.602361558803272e-06,
+      "loss": 0.287,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6393989983305509,
+      "grad_norm": 0.5418532715604891,
+      "learning_rate": 8.600340317204697e-06,
+      "loss": 0.286,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6397328881469115,
+      "grad_norm": 0.5483962174004143,
+      "learning_rate": 8.598317852949163e-06,
+      "loss": 0.2831,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6400667779632722,
+      "grad_norm": 0.5188376378171756,
+      "learning_rate": 8.596294166723492e-06,
+      "loss": 0.2978,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6404006677796327,
+      "grad_norm": 0.5212714837023597,
+      "learning_rate": 8.594269259214918e-06,
+      "loss": 0.289,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6407345575959933,
+      "grad_norm": 0.5358149851719421,
+      "learning_rate": 8.592243131111085e-06,
+      "loss": 0.2967,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6410684474123539,
+      "grad_norm": 0.486638661125879,
+      "learning_rate": 8.590215783100058e-06,
+      "loss": 0.2802,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6414023372287145,
+      "grad_norm": 0.5113666191696726,
+      "learning_rate": 8.588187215870315e-06,
+      "loss": 0.2816,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6417362270450752,
+      "grad_norm": 0.525380034989698,
+      "learning_rate": 8.586157430110747e-06,
+      "loss": 0.2856,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6420701168614358,
+      "grad_norm": 0.5325347345808721,
+      "learning_rate": 8.584126426510655e-06,
+      "loss": 0.2965,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6424040066777963,
+      "grad_norm": 0.5054911505700107,
+      "learning_rate": 8.582094205759762e-06,
+      "loss": 0.2992,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6427378964941569,
+      "grad_norm": 0.5640102293681744,
+      "learning_rate": 8.580060768548194e-06,
+      "loss": 0.2935,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6430717863105175,
+      "grad_norm": 0.5607977593603072,
+      "learning_rate": 8.578026115566503e-06,
+      "loss": 0.2948,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6434056761268782,
+      "grad_norm": 0.5329289922509304,
+      "learning_rate": 8.57599024750564e-06,
+      "loss": 0.3003,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6437395659432388,
+      "grad_norm": 0.5114465208334875,
+      "learning_rate": 8.573953165056978e-06,
+      "loss": 0.2926,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6440734557595993,
+      "grad_norm": 0.5498007665168229,
+      "learning_rate": 8.571914868912303e-06,
+      "loss": 0.2777,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6444073455759599,
+      "grad_norm": 0.5450402898750754,
+      "learning_rate": 8.569875359763804e-06,
+      "loss": 0.2907,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6447412353923205,
+      "grad_norm": 0.47767693328033706,
+      "learning_rate": 8.56783463830409e-06,
+      "loss": 0.3024,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6450751252086812,
+      "grad_norm": 0.49748993876738457,
+      "learning_rate": 8.565792705226183e-06,
+      "loss": 0.2866,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6454090150250418,
+      "grad_norm": 0.4865645238557324,
+      "learning_rate": 8.563749561223507e-06,
+      "loss": 0.2954,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6457429048414023,
+      "grad_norm": 0.483935068551559,
+      "learning_rate": 8.561705206989907e-06,
+      "loss": 0.2958,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6460767946577629,
+      "grad_norm": 0.5097557128685111,
+      "learning_rate": 8.559659643219634e-06,
+      "loss": 0.2982,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6464106844741235,
+      "grad_norm": 0.5035568804011799,
+      "learning_rate": 8.55761287060735e-06,
+      "loss": 0.2909,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6467445742904842,
+      "grad_norm": 0.4947339497112875,
+      "learning_rate": 8.55556488984813e-06,
+      "loss": 0.2863,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6470784641068448,
+      "grad_norm": 0.5398050639570983,
+      "learning_rate": 8.553515701637458e-06,
+      "loss": 0.2882,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6474123539232054,
+      "grad_norm": 0.543228954353323,
+      "learning_rate": 8.551465306671225e-06,
+      "loss": 0.2955,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6477462437395659,
+      "grad_norm": 0.4711487089806967,
+      "learning_rate": 8.549413705645737e-06,
+      "loss": 0.281,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6480801335559265,
+      "grad_norm": 0.5410798383917854,
+      "learning_rate": 8.547360899257707e-06,
+      "loss": 0.2962,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6484140233722872,
+      "grad_norm": 0.5431707985191377,
+      "learning_rate": 8.545306888204255e-06,
+      "loss": 0.2896,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6487479131886478,
+      "grad_norm": 0.5218590786000455,
+      "learning_rate": 8.543251673182915e-06,
+      "loss": 0.2965,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6490818030050084,
+      "grad_norm": 0.5575584767548769,
+      "learning_rate": 8.541195254891628e-06,
+      "loss": 0.2945,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6494156928213689,
+      "grad_norm": 0.5262298770680346,
+      "learning_rate": 8.53913763402874e-06,
+      "loss": 0.2891,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6497495826377295,
+      "grad_norm": 0.4967601742178605,
+      "learning_rate": 8.537078811293012e-06,
+      "loss": 0.2871,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6500834724540901,
+      "grad_norm": 0.48696136740884116,
+      "learning_rate": 8.535018787383606e-06,
+      "loss": 0.2857,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6504173622704508,
+      "grad_norm": 0.521093066758217,
+      "learning_rate": 8.532957563000096e-06,
+      "loss": 0.282,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6507512520868114,
+      "grad_norm": 0.5074049165191769,
+      "learning_rate": 8.530895138842467e-06,
+      "loss": 0.2872,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6510851419031719,
+      "grad_norm": 0.5004127239427241,
+      "learning_rate": 8.528831515611105e-06,
+      "loss": 0.2976,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6514190317195325,
+      "grad_norm": 0.5045683531686321,
+      "learning_rate": 8.526766694006804e-06,
+      "loss": 0.2829,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6517529215358931,
+      "grad_norm": 0.5194778024378418,
+      "learning_rate": 8.524700674730769e-06,
+      "loss": 0.2835,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6520868113522538,
+      "grad_norm": 0.4772012250269805,
+      "learning_rate": 8.52263345848461e-06,
+      "loss": 0.2952,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6524207011686144,
+      "grad_norm": 0.5162044478404759,
+      "learning_rate": 8.52056504597034e-06,
+      "loss": 0.2966,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6527545909849749,
+      "grad_norm": 0.5181473574881149,
+      "learning_rate": 8.518495437890383e-06,
+      "loss": 0.2911,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6530884808013355,
+      "grad_norm": 0.49200659927205065,
+      "learning_rate": 8.516424634947569e-06,
+      "loss": 0.2953,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6534223706176961,
+      "grad_norm": 0.47338184429851277,
+      "learning_rate": 8.514352637845128e-06,
+      "loss": 0.2915,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6537562604340568,
+      "grad_norm": 0.5183791057331741,
+      "learning_rate": 8.512279447286704e-06,
+      "loss": 0.2994,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6540901502504174,
+      "grad_norm": 0.5054770613161101,
+      "learning_rate": 8.510205063976338e-06,
+      "loss": 0.3053,
+      "step": 1959
+    },
+    {
+      "epoch": 0.654424040066778,
+      "grad_norm": 0.5020218838468772,
+      "learning_rate": 8.50812948861848e-06,
+      "loss": 0.2882,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6547579298831385,
+      "grad_norm": 0.48531444929872264,
+      "learning_rate": 8.506052721917989e-06,
+      "loss": 0.2899,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6550918196994991,
+      "grad_norm": 0.4964831198306773,
+      "learning_rate": 8.50397476458012e-06,
+      "loss": 0.2928,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6554257095158598,
+      "grad_norm": 0.48496068260104547,
+      "learning_rate": 8.501895617310538e-06,
+      "loss": 0.296,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6557595993322204,
+      "grad_norm": 0.5164176661817028,
+      "learning_rate": 8.499815280815309e-06,
+      "loss": 0.2899,
+      "step": 1964
+    },
+    {
+      "epoch": 0.656093489148581,
+      "grad_norm": 0.5099698463474833,
+      "learning_rate": 8.497733755800906e-06,
+      "loss": 0.2884,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6564273789649415,
+      "grad_norm": 0.5251052743530261,
+      "learning_rate": 8.495651042974206e-06,
+      "loss": 0.2976,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6567612687813021,
+      "grad_norm": 0.5025863045346948,
+      "learning_rate": 8.493567143042485e-06,
+      "loss": 0.2879,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6570951585976628,
+      "grad_norm": 0.5009406715602157,
+      "learning_rate": 8.491482056713425e-06,
+      "loss": 0.2988,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6574290484140234,
+      "grad_norm": 0.49533340926386643,
+      "learning_rate": 8.48939578469511e-06,
+      "loss": 0.2876,
+      "step": 1969
+    },
+    {
+      "epoch": 0.657762938230384,
+      "grad_norm": 0.5127956865568035,
+      "learning_rate": 8.487308327696028e-06,
+      "loss": 0.2971,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6580968280467445,
+      "grad_norm": 0.5110099424084426,
+      "learning_rate": 8.48521968642507e-06,
+      "loss": 0.2865,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6584307178631051,
+      "grad_norm": 0.5104110113857787,
+      "learning_rate": 8.483129861591525e-06,
+      "loss": 0.2855,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6587646076794658,
+      "grad_norm": 0.5119419030253013,
+      "learning_rate": 8.481038853905091e-06,
+      "loss": 0.2918,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6590984974958264,
+      "grad_norm": 0.5101029362643744,
+      "learning_rate": 8.478946664075857e-06,
+      "loss": 0.2679,
+      "step": 1974
+    },
+    {
+      "epoch": 0.659432387312187,
+      "grad_norm": 0.47687789222065596,
+      "learning_rate": 8.476853292814324e-06,
+      "loss": 0.2794,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6597662771285476,
+      "grad_norm": 0.5134306871687518,
+      "learning_rate": 8.47475874083139e-06,
+      "loss": 0.292,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6601001669449081,
+      "grad_norm": 0.4742201083502798,
+      "learning_rate": 8.472663008838353e-06,
+      "loss": 0.2857,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6604340567612688,
+      "grad_norm": 0.4972320290837381,
+      "learning_rate": 8.470566097546915e-06,
+      "loss": 0.281,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6607679465776294,
+      "grad_norm": 0.5005733282433982,
+      "learning_rate": 8.468468007669173e-06,
+      "loss": 0.295,
+      "step": 1979
+    },
+    {
+      "epoch": 0.66110183639399,
+      "grad_norm": 0.5285596065010989,
+      "learning_rate": 8.466368739917628e-06,
+      "loss": 0.2903,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6614357262103506,
+      "grad_norm": 0.49596995617723105,
+      "learning_rate": 8.46426829500518e-06,
+      "loss": 0.2936,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6617696160267111,
+      "grad_norm": 0.5127864655827731,
+      "learning_rate": 8.462166673645134e-06,
+      "loss": 0.2836,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6621035058430718,
+      "grad_norm": 0.502643448749405,
+      "learning_rate": 8.460063876551186e-06,
+      "loss": 0.2935,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6624373956594324,
+      "grad_norm": 0.49498103807054733,
+      "learning_rate": 8.457959904437434e-06,
+      "loss": 0.2937,
+      "step": 1984
+    },
+    {
+      "epoch": 0.662771285475793,
+      "grad_norm": 0.4759933057650423,
+      "learning_rate": 8.455854758018377e-06,
+      "loss": 0.2784,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6631051752921536,
+      "grad_norm": 0.528450301081564,
+      "learning_rate": 8.453748438008912e-06,
+      "loss": 0.286,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6634390651085141,
+      "grad_norm": 0.5125478337001144,
+      "learning_rate": 8.451640945124336e-06,
+      "loss": 0.299,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6637729549248748,
+      "grad_norm": 0.5138354616854217,
+      "learning_rate": 8.449532280080341e-06,
+      "loss": 0.2886,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6641068447412354,
+      "grad_norm": 0.4919490519369325,
+      "learning_rate": 8.447422443593018e-06,
+      "loss": 0.2804,
+      "step": 1989
+    },
+    {
+      "epoch": 0.664440734557596,
+      "grad_norm": 0.5043922128189631,
+      "learning_rate": 8.445311436378859e-06,
+      "loss": 0.2909,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6647746243739566,
+      "grad_norm": 0.5040549776012629,
+      "learning_rate": 8.443199259154748e-06,
+      "loss": 0.2878,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6651085141903172,
+      "grad_norm": 0.5033767869644652,
+      "learning_rate": 8.441085912637973e-06,
+      "loss": 0.2799,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6654424040066778,
+      "grad_norm": 0.5142423951435381,
+      "learning_rate": 8.438971397546214e-06,
+      "loss": 0.2861,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6657762938230384,
+      "grad_norm": 0.5131111198603047,
+      "learning_rate": 8.436855714597548e-06,
+      "loss": 0.2896,
+      "step": 1994
+    },
+    {
+      "epoch": 0.666110183639399,
+      "grad_norm": 0.5244778468963095,
+      "learning_rate": 8.43473886451045e-06,
+      "loss": 0.2989,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6664440734557596,
+      "grad_norm": 0.506009382399752,
+      "learning_rate": 8.432620848003795e-06,
+      "loss": 0.2927,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6667779632721202,
+      "grad_norm": 0.4894685813976203,
+      "learning_rate": 8.430501665796846e-06,
+      "loss": 0.2897,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6671118530884808,
+      "grad_norm": 0.5069780378770419,
+      "learning_rate": 8.428381318609268e-06,
+      "loss": 0.2949,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6674457429048414,
+      "grad_norm": 0.5007637425841345,
+      "learning_rate": 8.426259807161123e-06,
+      "loss": 0.2911,
+      "step": 1999
+    },
+    {
+      "epoch": 0.667779632721202,
+      "grad_norm": 0.4729174638207238,
+      "learning_rate": 8.42413713217286e-06,
+      "loss": 0.2854,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6681135225375626,
+      "grad_norm": 0.5180191326209399,
+      "learning_rate": 8.422013294365332e-06,
+      "loss": 0.2861,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6684474123539232,
+      "grad_norm": 0.4843818355360035,
+      "learning_rate": 8.419888294459782e-06,
+      "loss": 0.2885,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6687813021702838,
+      "grad_norm": 0.487922822164398,
+      "learning_rate": 8.417762133177849e-06,
+      "loss": 0.2806,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6691151919866444,
+      "grad_norm": 0.49730436818369994,
+      "learning_rate": 8.415634811241568e-06,
+      "loss": 0.289,
+      "step": 2004
+    },
+    {
+      "epoch": 0.669449081803005,
+      "grad_norm": 0.4905582122415976,
+      "learning_rate": 8.413506329373364e-06,
+      "loss": 0.2841,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6697829716193656,
+      "grad_norm": 0.5156307822043974,
+      "learning_rate": 8.41137668829606e-06,
+      "loss": 0.297,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6701168614357262,
+      "grad_norm": 0.518370817642097,
+      "learning_rate": 8.40924588873287e-06,
+      "loss": 0.2991,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6704507512520869,
+      "grad_norm": 0.4932774628064574,
+      "learning_rate": 8.407113931407404e-06,
+      "loss": 0.2814,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6707846410684474,
+      "grad_norm": 0.5423207580669606,
+      "learning_rate": 8.404980817043661e-06,
+      "loss": 0.2956,
+      "step": 2009
+    },
+    {
+      "epoch": 0.671118530884808,
+      "grad_norm": 0.5199898224487058,
+      "learning_rate": 8.40284654636604e-06,
+      "loss": 0.2911,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6714524207011686,
+      "grad_norm": 0.48988883936086414,
+      "learning_rate": 8.400711120099325e-06,
+      "loss": 0.2937,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6717863105175292,
+      "grad_norm": 0.5054625858186238,
+      "learning_rate": 8.398574538968697e-06,
+      "loss": 0.2943,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6721202003338899,
+      "grad_norm": 0.4973052143825497,
+      "learning_rate": 8.396436803699728e-06,
+      "loss": 0.2847,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6724540901502504,
+      "grad_norm": 0.47979713458016765,
+      "learning_rate": 8.394297915018378e-06,
+      "loss": 0.2815,
+      "step": 2014
+    },
+    {
+      "epoch": 0.672787979966611,
+      "grad_norm": 0.4944434720216582,
+      "learning_rate": 8.39215787365101e-06,
+      "loss": 0.2839,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6731218697829716,
+      "grad_norm": 0.5200378752495103,
+      "learning_rate": 8.390016680324364e-06,
+      "loss": 0.2971,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6734557595993322,
+      "grad_norm": 0.5110816046307066,
+      "learning_rate": 8.387874335765582e-06,
+      "loss": 0.2933,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6737896494156929,
+      "grad_norm": 0.49857783774783154,
+      "learning_rate": 8.385730840702193e-06,
+      "loss": 0.2839,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6741235392320534,
+      "grad_norm": 0.4823185637120209,
+      "learning_rate": 8.383586195862114e-06,
+      "loss": 0.2865,
+      "step": 2019
+    },
+    {
+      "epoch": 0.674457429048414,
+      "grad_norm": 0.4824159681680813,
+      "learning_rate": 8.38144040197366e-06,
+      "loss": 0.2977,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6747913188647746,
+      "grad_norm": 0.5388834987025742,
+      "learning_rate": 8.379293459765527e-06,
+      "loss": 0.2911,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6751252086811352,
+      "grad_norm": 0.5312670826601585,
+      "learning_rate": 8.377145369966806e-06,
+      "loss": 0.294,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6754590984974959,
+      "grad_norm": 0.5163895264745563,
+      "learning_rate": 8.374996133306981e-06,
+      "loss": 0.2929,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6757929883138565,
+      "grad_norm": 0.4969010873564039,
+      "learning_rate": 8.372845750515917e-06,
+      "loss": 0.2877,
+      "step": 2024
+    },
+    {
+      "epoch": 0.676126878130217,
+      "grad_norm": 0.5135985950577571,
+      "learning_rate": 8.370694222323876e-06,
+      "loss": 0.3057,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6764607679465776,
+      "grad_norm": 0.5138428509596064,
+      "learning_rate": 8.368541549461504e-06,
+      "loss": 0.3007,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6767946577629382,
+      "grad_norm": 0.5105475978310682,
+      "learning_rate": 8.36638773265984e-06,
+      "loss": 0.3007,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6771285475792989,
+      "grad_norm": 0.48906005449901696,
+      "learning_rate": 8.364232772650306e-06,
+      "loss": 0.2913,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6774624373956595,
+      "grad_norm": 0.48767212373387114,
+      "learning_rate": 8.362076670164718e-06,
+      "loss": 0.2876,
+      "step": 2029
+    },
+    {
+      "epoch": 0.67779632721202,
+      "grad_norm": 0.4927561290623782,
+      "learning_rate": 8.359919425935276e-06,
+      "loss": 0.2882,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6781302170283806,
+      "grad_norm": 0.5364581049949967,
+      "learning_rate": 8.357761040694568e-06,
+      "loss": 0.3026,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6784641068447412,
+      "grad_norm": 0.5297653500533351,
+      "learning_rate": 8.35560151517557e-06,
+      "loss": 0.3031,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6787979966611019,
+      "grad_norm": 0.4953140793123126,
+      "learning_rate": 8.353440850111651e-06,
+      "loss": 0.2965,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6791318864774625,
+      "grad_norm": 0.5070953598402053,
+      "learning_rate": 8.351279046236556e-06,
+      "loss": 0.2911,
+      "step": 2034
+    },
+    {
+      "epoch": 0.679465776293823,
+      "grad_norm": 0.5300739255491553,
+      "learning_rate": 8.349116104284423e-06,
+      "loss": 0.292,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6797996661101836,
+      "grad_norm": 0.5128223256044724,
+      "learning_rate": 8.346952024989778e-06,
+      "loss": 0.2961,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6801335559265442,
+      "grad_norm": 0.4973571351428466,
+      "learning_rate": 8.34478680908753e-06,
+      "loss": 0.29,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6804674457429049,
+      "grad_norm": 0.465962123876017,
+      "learning_rate": 8.342620457312977e-06,
+      "loss": 0.2878,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6808013355592655,
+      "grad_norm": 0.5249504617230119,
+      "learning_rate": 8.340452970401798e-06,
+      "loss": 0.2888,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6811352253756261,
+      "grad_norm": 0.5563434242869134,
+      "learning_rate": 8.338284349090063e-06,
+      "loss": 0.2949,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6814691151919866,
+      "grad_norm": 0.48687956137430694,
+      "learning_rate": 8.33611459411422e-06,
+      "loss": 0.2705,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6818030050083472,
+      "grad_norm": 0.5210323782649583,
+      "learning_rate": 8.333943706211112e-06,
+      "loss": 0.2951,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6821368948247079,
+      "grad_norm": 0.5457855177436318,
+      "learning_rate": 8.331771686117959e-06,
+      "loss": 0.289,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6824707846410685,
+      "grad_norm": 0.5305062868082653,
+      "learning_rate": 8.329598534572368e-06,
+      "loss": 0.2912,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6828046744574291,
+      "grad_norm": 0.5131017493935651,
+      "learning_rate": 8.327424252312328e-06,
+      "loss": 0.2818,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6831385642737896,
+      "grad_norm": 0.5406699234522095,
+      "learning_rate": 8.325248840076219e-06,
+      "loss": 0.2945,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6834724540901502,
+      "grad_norm": 0.519927355585587,
+      "learning_rate": 8.323072298602795e-06,
+      "loss": 0.2801,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6838063439065108,
+      "grad_norm": 0.5056645862540814,
+      "learning_rate": 8.3208946286312e-06,
+      "loss": 0.2784,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6841402337228715,
+      "grad_norm": 0.47476859783897013,
+      "learning_rate": 8.318715830900959e-06,
+      "loss": 0.2946,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6844741235392321,
+      "grad_norm": 0.5454068404856154,
+      "learning_rate": 8.316535906151981e-06,
+      "loss": 0.2955,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6848080133555926,
+      "grad_norm": 0.49876494253609027,
+      "learning_rate": 8.314354855124559e-06,
+      "loss": 0.2949,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6851419031719532,
+      "grad_norm": 0.5158031672713368,
+      "learning_rate": 8.312172678559363e-06,
+      "loss": 0.2841,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6854757929883138,
+      "grad_norm": 0.515224537045618,
+      "learning_rate": 8.30998937719745e-06,
+      "loss": 0.2851,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6858096828046745,
+      "grad_norm": 0.5146059017238462,
+      "learning_rate": 8.307804951780258e-06,
+      "loss": 0.2836,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6861435726210351,
+      "grad_norm": 0.5098903662939623,
+      "learning_rate": 8.305619403049607e-06,
+      "loss": 0.2785,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6864774624373957,
+      "grad_norm": 0.5329895803564972,
+      "learning_rate": 8.3034327317477e-06,
+      "loss": 0.288,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6868113522537562,
+      "grad_norm": 0.5022613708784806,
+      "learning_rate": 8.301244938617117e-06,
+      "loss": 0.2864,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6871452420701168,
+      "grad_norm": 0.510783213743551,
+      "learning_rate": 8.29905602440082e-06,
+      "loss": 0.2922,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6874791318864775,
+      "grad_norm": 0.5011654131444976,
+      "learning_rate": 8.296865989842154e-06,
+      "loss": 0.2847,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6878130217028381,
+      "grad_norm": 0.5056035442795356,
+      "learning_rate": 8.294674835684846e-06,
+      "loss": 0.2924,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6881469115191987,
+      "grad_norm": 0.5260445836668335,
+      "learning_rate": 8.292482562672997e-06,
+      "loss": 0.2934,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6884808013355592,
+      "grad_norm": 0.5415758107073558,
+      "learning_rate": 8.290289171551093e-06,
+      "loss": 0.2961,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6888146911519198,
+      "grad_norm": 0.5346168642488242,
+      "learning_rate": 8.288094663064e-06,
+      "loss": 0.2784,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6891485809682805,
+      "grad_norm": 0.5353656513136282,
+      "learning_rate": 8.285899037956958e-06,
+      "loss": 0.2842,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6894824707846411,
+      "grad_norm": 0.5045995053000498,
+      "learning_rate": 8.283702296975593e-06,
+      "loss": 0.2926,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6898163606010017,
+      "grad_norm": 0.54386287963066,
+      "learning_rate": 8.281504440865905e-06,
+      "loss": 0.2918,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6901502504173622,
+      "grad_norm": 0.5171777443384836,
+      "learning_rate": 8.279305470374278e-06,
+      "loss": 0.2967,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6904841402337228,
+      "grad_norm": 0.5282553233667787,
+      "learning_rate": 8.277105386247466e-06,
+      "loss": 0.289,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6908180300500835,
+      "grad_norm": 0.534349711628389,
+      "learning_rate": 8.274904189232609e-06,
+      "loss": 0.2977,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6911519198664441,
+      "grad_norm": 0.5185465619322038,
+      "learning_rate": 8.272701880077221e-06,
+      "loss": 0.2971,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6914858096828047,
+      "grad_norm": 0.5115600217292389,
+      "learning_rate": 8.270498459529197e-06,
+      "loss": 0.2948,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6918196994991653,
+      "grad_norm": 0.5124389258624229,
+      "learning_rate": 8.268293928336805e-06,
+      "loss": 0.2775,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6921535893155258,
+      "grad_norm": 0.5348671947493329,
+      "learning_rate": 8.266088287248696e-06,
+      "loss": 0.2806,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6924874791318865,
+      "grad_norm": 0.5116725473529693,
+      "learning_rate": 8.263881537013888e-06,
+      "loss": 0.2937,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6928213689482471,
+      "grad_norm": 0.5197845811557033,
+      "learning_rate": 8.261673678381786e-06,
+      "loss": 0.3004,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6931552587646077,
+      "grad_norm": 0.4978751890192807,
+      "learning_rate": 8.259464712102168e-06,
+      "loss": 0.2936,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6934891485809683,
+      "grad_norm": 0.5353174463288588,
+      "learning_rate": 8.257254638925188e-06,
+      "loss": 0.2965,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6938230383973288,
+      "grad_norm": 0.49153169432936145,
+      "learning_rate": 8.25504345960137e-06,
+      "loss": 0.2885,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6941569282136895,
+      "grad_norm": 0.5155257298650643,
+      "learning_rate": 8.252831174881626e-06,
+      "loss": 0.2975,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6944908180300501,
+      "grad_norm": 0.5137324181188666,
+      "learning_rate": 8.250617785517232e-06,
+      "loss": 0.2912,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6948247078464107,
+      "grad_norm": 0.51352435267916,
+      "learning_rate": 8.248403292259845e-06,
+      "loss": 0.2835,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6951585976627713,
+      "grad_norm": 0.4984020463066979,
+      "learning_rate": 8.246187695861495e-06,
+      "loss": 0.2955,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6954924874791318,
+      "grad_norm": 0.4806138499085471,
+      "learning_rate": 8.243970997074589e-06,
+      "loss": 0.3018,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6958263772954925,
+      "grad_norm": 0.49930189561496896,
+      "learning_rate": 8.241753196651903e-06,
+      "loss": 0.2953,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6961602671118531,
+      "grad_norm": 0.5086384443192604,
+      "learning_rate": 8.239534295346593e-06,
+      "loss": 0.2965,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6964941569282137,
+      "grad_norm": 0.5298628893993458,
+      "learning_rate": 8.237314293912186e-06,
+      "loss": 0.2949,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6968280467445743,
+      "grad_norm": 0.4884481331539497,
+      "learning_rate": 8.235093193102583e-06,
+      "loss": 0.2825,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6971619365609348,
+      "grad_norm": 0.5298804831301784,
+      "learning_rate": 8.23287099367206e-06,
+      "loss": 0.3018,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6974958263772955,
+      "grad_norm": 0.5133086295124811,
+      "learning_rate": 8.23064769637526e-06,
+      "loss": 0.2857,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6978297161936561,
+      "grad_norm": 0.5140975199266322,
+      "learning_rate": 8.228423301967207e-06,
+      "loss": 0.2972,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6981636060100167,
+      "grad_norm": 0.49581325704287366,
+      "learning_rate": 8.226197811203294e-06,
+      "loss": 0.2956,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6984974958263773,
+      "grad_norm": 0.5190261449686957,
+      "learning_rate": 8.223971224839282e-06,
+      "loss": 0.2768,
+      "step": 2092
+    },
+    {
+      "epoch": 0.698831385642738,
+      "grad_norm": 0.522341724313072,
+      "learning_rate": 8.221743543631314e-06,
+      "loss": 0.2958,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6991652754590985,
+      "grad_norm": 0.5108225932649589,
+      "learning_rate": 8.219514768335895e-06,
+      "loss": 0.2883,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6994991652754591,
+      "grad_norm": 0.45949619724721374,
+      "learning_rate": 8.217284899709907e-06,
+      "loss": 0.274,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6998330550918197,
+      "grad_norm": 0.5287932472507615,
+      "learning_rate": 8.215053938510602e-06,
+      "loss": 0.2867,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7001669449081803,
+      "grad_norm": 0.5189404491101839,
+      "learning_rate": 8.212821885495602e-06,
+      "loss": 0.2856,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7005008347245409,
+      "grad_norm": 0.5192937834492897,
+      "learning_rate": 8.2105887414229e-06,
+      "loss": 0.2904,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7008347245409015,
+      "grad_norm": 0.5019839710474957,
+      "learning_rate": 8.208354507050865e-06,
+      "loss": 0.274,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7011686143572621,
+      "grad_norm": 0.4936738359975307,
+      "learning_rate": 8.206119183138223e-06,
+      "loss": 0.2835,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7015025041736227,
+      "grad_norm": 0.5205405267162605,
+      "learning_rate": 8.203882770444086e-06,
+      "loss": 0.2956,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7018363939899833,
+      "grad_norm": 0.5218671134222453,
+      "learning_rate": 8.201645269727924e-06,
+      "loss": 0.2826,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7021702838063439,
+      "grad_norm": 0.4882422090724648,
+      "learning_rate": 8.199406681749585e-06,
+      "loss": 0.2828,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7025041736227045,
+      "grad_norm": 0.5525991077946277,
+      "learning_rate": 8.197167007269275e-06,
+      "loss": 0.2958,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7028380634390651,
+      "grad_norm": 0.5150016570638973,
+      "learning_rate": 8.194926247047584e-06,
+      "loss": 0.2832,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7031719532554257,
+      "grad_norm": 0.48114119864672267,
+      "learning_rate": 8.192684401845454e-06,
+      "loss": 0.2919,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7035058430717863,
+      "grad_norm": 0.4892744066558964,
+      "learning_rate": 8.190441472424213e-06,
+      "loss": 0.2827,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7038397328881469,
+      "grad_norm": 0.5955132324893855,
+      "learning_rate": 8.188197459545541e-06,
+      "loss": 0.2856,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7041736227045076,
+      "grad_norm": 0.5081849491512559,
+      "learning_rate": 8.185952363971495e-06,
+      "loss": 0.2926,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7045075125208681,
+      "grad_norm": 0.5139613528234572,
+      "learning_rate": 8.183706186464502e-06,
+      "loss": 0.2958,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7048414023372287,
+      "grad_norm": 0.5784600425498146,
+      "learning_rate": 8.181458927787347e-06,
+      "loss": 0.2906,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7051752921535893,
+      "grad_norm": 0.5278245659410805,
+      "learning_rate": 8.17921058870319e-06,
+      "loss": 0.2951,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7055091819699499,
+      "grad_norm": 0.48678904888434615,
+      "learning_rate": 8.176961169975553e-06,
+      "loss": 0.2924,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7058430717863106,
+      "grad_norm": 0.48393360879584035,
+      "learning_rate": 8.17471067236833e-06,
+      "loss": 0.2907,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7061769616026711,
+      "grad_norm": 0.49764180275330605,
+      "learning_rate": 8.172459096645778e-06,
+      "loss": 0.2786,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7065108514190317,
+      "grad_norm": 0.5431824042325069,
+      "learning_rate": 8.170206443572518e-06,
+      "loss": 0.2974,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7068447412353923,
+      "grad_norm": 0.5050620505231975,
+      "learning_rate": 8.167952713913541e-06,
+      "loss": 0.294,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7071786310517529,
+      "grad_norm": 0.5018609043365871,
+      "learning_rate": 8.165697908434204e-06,
+      "loss": 0.2895,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7075125208681136,
+      "grad_norm": 0.4444366156754223,
+      "learning_rate": 8.163442027900224e-06,
+      "loss": 0.2774,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7078464106844741,
+      "grad_norm": 0.5198842898264538,
+      "learning_rate": 8.161185073077686e-06,
+      "loss": 0.2806,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7081803005008347,
+      "grad_norm": 0.5347916910430861,
+      "learning_rate": 8.158927044733043e-06,
+      "loss": 0.2907,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7085141903171953,
+      "grad_norm": 0.4962113165683347,
+      "learning_rate": 8.156667943633107e-06,
+      "loss": 0.2789,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7088480801335559,
+      "grad_norm": 0.48048717164142474,
+      "learning_rate": 8.154407770545061e-06,
+      "loss": 0.2835,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7091819699499166,
+      "grad_norm": 0.5186774584368103,
+      "learning_rate": 8.152146526236444e-06,
+      "loss": 0.2921,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7095158597662772,
+      "grad_norm": 0.46941416274939474,
+      "learning_rate": 8.149884211475166e-06,
+      "loss": 0.2898,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7098497495826377,
+      "grad_norm": 0.4947939517173156,
+      "learning_rate": 8.147620827029493e-06,
+      "loss": 0.2958,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7101836393989983,
+      "grad_norm": 0.5256021276062515,
+      "learning_rate": 8.145356373668063e-06,
+      "loss": 0.2998,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7105175292153589,
+      "grad_norm": 0.5019317704483405,
+      "learning_rate": 8.143090852159871e-06,
+      "loss": 0.2905,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7108514190317196,
+      "grad_norm": 0.5126664203725181,
+      "learning_rate": 8.140824263274278e-06,
+      "loss": 0.3042,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7111853088480802,
+      "grad_norm": 0.5126695525842806,
+      "learning_rate": 8.138556607781003e-06,
+      "loss": 0.2941,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7115191986644407,
+      "grad_norm": 0.5136792594190253,
+      "learning_rate": 8.136287886450132e-06,
+      "loss": 0.3008,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7118530884808013,
+      "grad_norm": 0.4958057066490382,
+      "learning_rate": 8.13401810005211e-06,
+      "loss": 0.2808,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7121869782971619,
+      "grad_norm": 0.48635584955564876,
+      "learning_rate": 8.131747249357745e-06,
+      "loss": 0.2963,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7125208681135226,
+      "grad_norm": 0.45208697789589025,
+      "learning_rate": 8.129475335138207e-06,
+      "loss": 0.288,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7128547579298832,
+      "grad_norm": 0.48289713302823406,
+      "learning_rate": 8.127202358165025e-06,
+      "loss": 0.2901,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7131886477462437,
+      "grad_norm": 0.5089879388417422,
+      "learning_rate": 8.124928319210093e-06,
+      "loss": 0.2909,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7135225375626043,
+      "grad_norm": 0.4801898075755117,
+      "learning_rate": 8.12265321904566e-06,
+      "loss": 0.279,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7138564273789649,
+      "grad_norm": 0.48115579809768094,
+      "learning_rate": 8.120377058444336e-06,
+      "loss": 0.2867,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7141903171953256,
+      "grad_norm": 0.47522699930691775,
+      "learning_rate": 8.1180998381791e-06,
+      "loss": 0.2911,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7145242070116862,
+      "grad_norm": 0.4870803610968345,
+      "learning_rate": 8.11582155902328e-06,
+      "loss": 0.2866,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7148580968280468,
+      "grad_norm": 0.5237820107786503,
+      "learning_rate": 8.113542221750567e-06,
+      "loss": 0.2809,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7151919866444073,
+      "grad_norm": 0.4778114933179815,
+      "learning_rate": 8.111261827135016e-06,
+      "loss": 0.2747,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7155258764607679,
+      "grad_norm": 0.5086920427184559,
+      "learning_rate": 8.108980375951034e-06,
+      "loss": 0.2765,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7158597662771286,
+      "grad_norm": 0.4709538865771709,
+      "learning_rate": 8.10669786897339e-06,
+      "loss": 0.2802,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7161936560934892,
+      "grad_norm": 0.48985252948955976,
+      "learning_rate": 8.104414306977213e-06,
+      "loss": 0.296,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7165275459098498,
+      "grad_norm": 0.457494815481562,
+      "learning_rate": 8.102129690737988e-06,
+      "loss": 0.2736,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7168614357262103,
+      "grad_norm": 0.5239585319073947,
+      "learning_rate": 8.099844021031559e-06,
+      "loss": 0.2804,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7171953255425709,
+      "grad_norm": 0.5057981746873187,
+      "learning_rate": 8.09755729863413e-06,
+      "loss": 0.2965,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7175292153589315,
+      "grad_norm": 0.4952615832583428,
+      "learning_rate": 8.095269524322256e-06,
+      "loss": 0.2753,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7178631051752922,
+      "grad_norm": 0.5135971945744714,
+      "learning_rate": 8.092980698872855e-06,
+      "loss": 0.3019,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7181969949916528,
+      "grad_norm": 0.4861131267330874,
+      "learning_rate": 8.0906908230632e-06,
+      "loss": 0.2925,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7185308848080133,
+      "grad_norm": 0.5143709472372887,
+      "learning_rate": 8.088399897670922e-06,
+      "loss": 0.2836,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7188647746243739,
+      "grad_norm": 0.5150582867966279,
+      "learning_rate": 8.086107923474008e-06,
+      "loss": 0.3032,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7191986644407345,
+      "grad_norm": 0.48920188686686356,
+      "learning_rate": 8.0838149012508e-06,
+      "loss": 0.2829,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7195325542570952,
+      "grad_norm": 0.5103366907293816,
+      "learning_rate": 8.081520831779995e-06,
+      "loss": 0.2713,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7198664440734558,
+      "grad_norm": 0.4989415394241823,
+      "learning_rate": 8.079225715840646e-06,
+      "loss": 0.284,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7202003338898164,
+      "grad_norm": 0.5162604149350071,
+      "learning_rate": 8.076929554212166e-06,
+      "loss": 0.2926,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7205342237061769,
+      "grad_norm": 0.5082355814779143,
+      "learning_rate": 8.07463234767432e-06,
+      "loss": 0.2901,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7208681135225375,
+      "grad_norm": 0.5103702708919696,
+      "learning_rate": 8.072334097007224e-06,
+      "loss": 0.2877,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7212020033388982,
+      "grad_norm": 0.4498253383352107,
+      "learning_rate": 8.070034802991356e-06,
+      "loss": 0.2693,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7215358931552588,
+      "grad_norm": 0.48096738336069506,
+      "learning_rate": 8.067734466407539e-06,
+      "loss": 0.2861,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7218697829716194,
+      "grad_norm": 0.4844793827147407,
+      "learning_rate": 8.06543308803696e-06,
+      "loss": 0.2935,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7222036727879799,
+      "grad_norm": 0.5079219978692142,
+      "learning_rate": 8.063130668661154e-06,
+      "loss": 0.2996,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7225375626043405,
+      "grad_norm": 0.48542421304781286,
+      "learning_rate": 8.06082720906201e-06,
+      "loss": 0.2887,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7228714524207012,
+      "grad_norm": 0.4930196860532299,
+      "learning_rate": 8.058522710021773e-06,
+      "loss": 0.2883,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7232053422370618,
+      "grad_norm": 0.5153551299978163,
+      "learning_rate": 8.056217172323035e-06,
+      "loss": 0.2968,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7235392320534224,
+      "grad_norm": 0.5048743874563972,
+      "learning_rate": 8.053910596748748e-06,
+      "loss": 0.2816,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7238731218697829,
+      "grad_norm": 0.5425973267478111,
+      "learning_rate": 8.051602984082212e-06,
+      "loss": 0.2841,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7242070116861435,
+      "grad_norm": 0.48134984026505834,
+      "learning_rate": 8.04929433510708e-06,
+      "loss": 0.282,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7245409015025042,
+      "grad_norm": 0.5293987713014412,
+      "learning_rate": 8.04698465060736e-06,
+      "loss": 0.2815,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7248747913188648,
+      "grad_norm": 0.5242140026472704,
+      "learning_rate": 8.044673931367404e-06,
+      "loss": 0.2827,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7252086811352254,
+      "grad_norm": 0.5143452679473568,
+      "learning_rate": 8.042362178171926e-06,
+      "loss": 0.2803,
+      "step": 2172
+    },
+    {
+      "epoch": 0.725542570951586,
+      "grad_norm": 0.5030353380512003,
+      "learning_rate": 8.04004939180598e-06,
+      "loss": 0.2926,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7258764607679465,
+      "grad_norm": 0.49678278757411315,
+      "learning_rate": 8.037735573054979e-06,
+      "loss": 0.2888,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7262103505843072,
+      "grad_norm": 0.5102242834010907,
+      "learning_rate": 8.035420722704683e-06,
+      "loss": 0.2944,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7265442404006678,
+      "grad_norm": 0.47078543554383023,
+      "learning_rate": 8.033104841541204e-06,
+      "loss": 0.2899,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7268781302170284,
+      "grad_norm": 0.4683102852318039,
+      "learning_rate": 8.030787930351005e-06,
+      "loss": 0.2938,
+      "step": 2177
+    },
+    {
+      "epoch": 0.727212020033389,
+      "grad_norm": 0.4718983730210831,
+      "learning_rate": 8.028469989920893e-06,
+      "loss": 0.2822,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7275459098497495,
+      "grad_norm": 0.4999078944180634,
+      "learning_rate": 8.026151021038032e-06,
+      "loss": 0.2916,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7278797996661102,
+      "grad_norm": 0.5061703222019475,
+      "learning_rate": 8.02383102448993e-06,
+      "loss": 0.2945,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7282136894824708,
+      "grad_norm": 0.46341689310677964,
+      "learning_rate": 8.021510001064448e-06,
+      "loss": 0.2887,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7285475792988314,
+      "grad_norm": 0.47813852139548313,
+      "learning_rate": 8.01918795154979e-06,
+      "loss": 0.2891,
+      "step": 2182
+    },
+    {
+      "epoch": 0.728881469115192,
+      "grad_norm": 0.5237827919530572,
+      "learning_rate": 8.016864876734514e-06,
+      "loss": 0.2795,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7292153589315525,
+      "grad_norm": 0.5157099380345735,
+      "learning_rate": 8.014540777407525e-06,
+      "loss": 0.2926,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7295492487479132,
+      "grad_norm": 0.47224513100438364,
+      "learning_rate": 8.012215654358073e-06,
+      "loss": 0.299,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7298831385642738,
+      "grad_norm": 0.4776893456231562,
+      "learning_rate": 8.00988950837576e-06,
+      "loss": 0.2867,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7302170283806344,
+      "grad_norm": 0.5017081523912267,
+      "learning_rate": 8.007562340250533e-06,
+      "loss": 0.2982,
+      "step": 2187
+    },
+    {
+      "epoch": 0.730550918196995,
+      "grad_norm": 0.5272329430906135,
+      "learning_rate": 8.005234150772685e-06,
+      "loss": 0.2845,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7308848080133556,
+      "grad_norm": 0.4883650570273218,
+      "learning_rate": 8.002904940732857e-06,
+      "loss": 0.2934,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7312186978297162,
+      "grad_norm": 0.5124416131629178,
+      "learning_rate": 8.000574710922037e-06,
+      "loss": 0.2906,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7315525876460768,
+      "grad_norm": 0.4993730485519076,
+      "learning_rate": 7.99824346213156e-06,
+      "loss": 0.2927,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7318864774624374,
+      "grad_norm": 0.5069744567689942,
+      "learning_rate": 7.995911195153105e-06,
+      "loss": 0.2795,
+      "step": 2192
+    },
+    {
+      "epoch": 0.732220367278798,
+      "grad_norm": 0.5372099168162648,
+      "learning_rate": 7.993577910778698e-06,
+      "loss": 0.2972,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7325542570951586,
+      "grad_norm": 0.4921818006172939,
+      "learning_rate": 7.991243609800711e-06,
+      "loss": 0.2797,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7328881469115192,
+      "grad_norm": 0.5112752185661475,
+      "learning_rate": 7.98890829301186e-06,
+      "loss": 0.2926,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7332220367278798,
+      "grad_norm": 0.516918705684611,
+      "learning_rate": 7.986571961205206e-06,
+      "loss": 0.2937,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7335559265442404,
+      "grad_norm": 0.5214283751169844,
+      "learning_rate": 7.984234615174156e-06,
+      "loss": 0.2896,
+      "step": 2197
+    },
+    {
+      "epoch": 0.733889816360601,
+      "grad_norm": 0.5031220914489871,
+      "learning_rate": 7.981896255712458e-06,
+      "loss": 0.283,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7342237061769616,
+      "grad_norm": 0.5125279363026655,
+      "learning_rate": 7.979556883614212e-06,
+      "loss": 0.299,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7345575959933222,
+      "grad_norm": 0.5300856648090128,
+      "learning_rate": 7.977216499673854e-06,
+      "loss": 0.2942,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7348914858096828,
+      "grad_norm": 0.5248390268742231,
+      "learning_rate": 7.974875104686164e-06,
+      "loss": 0.2784,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7352253756260434,
+      "grad_norm": 0.5390844513277494,
+      "learning_rate": 7.97253269944627e-06,
+      "loss": 0.2921,
+      "step": 2202
+    },
+    {
+      "epoch": 0.735559265442404,
+      "grad_norm": 0.5392769706299533,
+      "learning_rate": 7.970189284749641e-06,
+      "loss": 0.3028,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7358931552587646,
+      "grad_norm": 0.5012855660434895,
+      "learning_rate": 7.967844861392087e-06,
+      "loss": 0.2871,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7362270450751253,
+      "grad_norm": 0.4777620453347636,
+      "learning_rate": 7.965499430169764e-06,
+      "loss": 0.2882,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7365609348914858,
+      "grad_norm": 0.5187082674630245,
+      "learning_rate": 7.963152991879167e-06,
+      "loss": 0.2827,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7368948247078464,
+      "grad_norm": 0.5226436146054446,
+      "learning_rate": 7.960805547317134e-06,
+      "loss": 0.2876,
+      "step": 2207
+    },
+    {
+      "epoch": 0.737228714524207,
+      "grad_norm": 0.5020725158916941,
+      "learning_rate": 7.958457097280845e-06,
+      "loss": 0.2755,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7375626043405676,
+      "grad_norm": 0.48038844173062095,
+      "learning_rate": 7.956107642567824e-06,
+      "loss": 0.2665,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7378964941569283,
+      "grad_norm": 0.5646248266652931,
+      "learning_rate": 7.95375718397593e-06,
+      "loss": 0.287,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7382303839732888,
+      "grad_norm": 0.539881748190849,
+      "learning_rate": 7.95140572230337e-06,
+      "loss": 0.3016,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7385642737896494,
+      "grad_norm": 0.514028033946739,
+      "learning_rate": 7.949053258348684e-06,
+      "loss": 0.2915,
+      "step": 2212
+    },
+    {
+      "epoch": 0.73889816360601,
+      "grad_norm": 0.5325742726652537,
+      "learning_rate": 7.946699792910761e-06,
+      "loss": 0.2905,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7392320534223706,
+      "grad_norm": 0.5322473359324987,
+      "learning_rate": 7.944345326788825e-06,
+      "loss": 0.2922,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7395659432387313,
+      "grad_norm": 0.5069765088429906,
+      "learning_rate": 7.941989860782438e-06,
+      "loss": 0.2811,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7398998330550918,
+      "grad_norm": 0.4691025964356952,
+      "learning_rate": 7.939633395691505e-06,
+      "loss": 0.2759,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7402337228714524,
+      "grad_norm": 0.5049921924101604,
+      "learning_rate": 7.93727593231627e-06,
+      "loss": 0.2971,
+      "step": 2217
+    },
+    {
+      "epoch": 0.740567612687813,
+      "grad_norm": 0.5409885911998696,
+      "learning_rate": 7.934917471457316e-06,
+      "loss": 0.2844,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7409015025041736,
+      "grad_norm": 0.49615879538179913,
+      "learning_rate": 7.932558013915561e-06,
+      "loss": 0.2903,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7412353923205343,
+      "grad_norm": 0.491287647902513,
+      "learning_rate": 7.93019756049227e-06,
+      "loss": 0.2839,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7415692821368948,
+      "grad_norm": 0.5115579267977318,
+      "learning_rate": 7.927836111989035e-06,
+      "loss": 0.2911,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7419031719532554,
+      "grad_norm": 0.5024397354975961,
+      "learning_rate": 7.925473669207797e-06,
+      "loss": 0.2807,
+      "step": 2222
+    },
+    {
+      "epoch": 0.742237061769616,
+      "grad_norm": 0.4878648989050499,
+      "learning_rate": 7.923110232950826e-06,
+      "loss": 0.2831,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7425709515859766,
+      "grad_norm": 0.46101569167033596,
+      "learning_rate": 7.920745804020733e-06,
+      "loss": 0.2811,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7429048414023373,
+      "grad_norm": 0.5064073155235714,
+      "learning_rate": 7.918380383220466e-06,
+      "loss": 0.2829,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7432387312186979,
+      "grad_norm": 0.5188715344646577,
+      "learning_rate": 7.916013971353309e-06,
+      "loss": 0.3009,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7435726210350584,
+      "grad_norm": 0.529896444730491,
+      "learning_rate": 7.913646569222885e-06,
+      "loss": 0.2984,
+      "step": 2227
+    },
+    {
+      "epoch": 0.743906510851419,
+      "grad_norm": 0.49736758427727973,
+      "learning_rate": 7.911278177633151e-06,
+      "loss": 0.2823,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7442404006677796,
+      "grad_norm": 0.4854435696379315,
+      "learning_rate": 7.9089087973884e-06,
+      "loss": 0.2906,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7445742904841403,
+      "grad_norm": 0.5040858187829735,
+      "learning_rate": 7.906538429293262e-06,
+      "loss": 0.287,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7449081803005009,
+      "grad_norm": 0.5071299274344825,
+      "learning_rate": 7.904167074152703e-06,
+      "loss": 0.2863,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7452420701168614,
+      "grad_norm": 0.535211889826302,
+      "learning_rate": 7.901794732772017e-06,
+      "loss": 0.2974,
+      "step": 2232
+    },
+    {
+      "epoch": 0.745575959933222,
+      "grad_norm": 0.5371017442818009,
+      "learning_rate": 7.899421405956847e-06,
+      "loss": 0.2907,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7459098497495826,
+      "grad_norm": 0.5002397706374389,
+      "learning_rate": 7.897047094513157e-06,
+      "loss": 0.2886,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7462437395659433,
+      "grad_norm": 0.4805014205889334,
+      "learning_rate": 7.894671799247251e-06,
+      "loss": 0.2762,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7465776293823039,
+      "grad_norm": 0.49481928752143206,
+      "learning_rate": 7.89229552096577e-06,
+      "loss": 0.2702,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7469115191986644,
+      "grad_norm": 0.5568171926794035,
+      "learning_rate": 7.889918260475685e-06,
+      "loss": 0.2673,
+      "step": 2237
+    },
+    {
+      "epoch": 0.747245409015025,
+      "grad_norm": 0.513571599364459,
+      "learning_rate": 7.8875400185843e-06,
+      "loss": 0.2892,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7475792988313856,
+      "grad_norm": 0.5242461696790308,
+      "learning_rate": 7.885160796099254e-06,
+      "loss": 0.2914,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7479131886477463,
+      "grad_norm": 0.5132155975985135,
+      "learning_rate": 7.88278059382852e-06,
+      "loss": 0.2819,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7482470784641069,
+      "grad_norm": 0.5014654258147493,
+      "learning_rate": 7.880399412580401e-06,
+      "loss": 0.2888,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7485809682804675,
+      "grad_norm": 0.5119089595404953,
+      "learning_rate": 7.878017253163533e-06,
+      "loss": 0.281,
+      "step": 2242
+    },
+    {
+      "epoch": 0.748914858096828,
+      "grad_norm": 0.5051566761991256,
+      "learning_rate": 7.875634116386888e-06,
+      "loss": 0.2769,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7492487479131886,
+      "grad_norm": 0.5228577605881422,
+      "learning_rate": 7.873250003059765e-06,
+      "loss": 0.2824,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7495826377295493,
+      "grad_norm": 0.5019091753002975,
+      "learning_rate": 7.870864913991796e-06,
+      "loss": 0.2964,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7499165275459099,
+      "grad_norm": 0.505464680255568,
+      "learning_rate": 7.868478849992944e-06,
+      "loss": 0.2856,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7502504173622705,
+      "grad_norm": 0.4912067969345575,
+      "learning_rate": 7.866091811873509e-06,
+      "loss": 0.2899,
+      "step": 2247
+    },
+    {
+      "epoch": 0.750584307178631,
+      "grad_norm": 0.527632201555814,
+      "learning_rate": 7.863703800444111e-06,
+      "loss": 0.2918,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7509181969949916,
+      "grad_norm": 0.5011141227433704,
+      "learning_rate": 7.86131481651571e-06,
+      "loss": 0.2855,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7512520868113522,
+      "grad_norm": 0.4845713490240836,
+      "learning_rate": 7.85892486089959e-06,
+      "loss": 0.2972,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7515859766277129,
+      "grad_norm": 0.47848834934465,
+      "learning_rate": 7.856533934407371e-06,
+      "loss": 0.2763,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7519198664440735,
+      "grad_norm": 0.485665749222131,
+      "learning_rate": 7.854142037850995e-06,
+      "loss": 0.2945,
+      "step": 2252
+    },
+    {
+      "epoch": 0.752253756260434,
+      "grad_norm": 0.5015490974211299,
+      "learning_rate": 7.851749172042741e-06,
+      "loss": 0.2868,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7525876460767946,
+      "grad_norm": 0.4709318719613135,
+      "learning_rate": 7.849355337795213e-06,
+      "loss": 0.2877,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7529215358931552,
+      "grad_norm": 0.47572707113042,
+      "learning_rate": 7.846960535921344e-06,
+      "loss": 0.2717,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7532554257095159,
+      "grad_norm": 0.48028859848935984,
+      "learning_rate": 7.844564767234398e-06,
+      "loss": 0.274,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7535893155258765,
+      "grad_norm": 0.5171941657180035,
+      "learning_rate": 7.842168032547963e-06,
+      "loss": 0.2848,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7539232053422371,
+      "grad_norm": 0.5011112893932064,
+      "learning_rate": 7.83977033267596e-06,
+      "loss": 0.2916,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7542570951585976,
+      "grad_norm": 0.4802244118857508,
+      "learning_rate": 7.837371668432635e-06,
+      "loss": 0.2894,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7545909849749582,
+      "grad_norm": 0.5151917612469675,
+      "learning_rate": 7.83497204063256e-06,
+      "loss": 0.2792,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7549248747913189,
+      "grad_norm": 0.5341015513307913,
+      "learning_rate": 7.832571450090639e-06,
+      "loss": 0.2734,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7552587646076795,
+      "grad_norm": 0.5161595016459629,
+      "learning_rate": 7.8301698976221e-06,
+      "loss": 0.2948,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7555926544240401,
+      "grad_norm": 0.48032557876155074,
+      "learning_rate": 7.827767384042495e-06,
+      "loss": 0.2788,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7559265442404006,
+      "grad_norm": 0.5054263438828717,
+      "learning_rate": 7.825363910167709e-06,
+      "loss": 0.2991,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7562604340567612,
+      "grad_norm": 0.5307951784961427,
+      "learning_rate": 7.822959476813946e-06,
+      "loss": 0.2921,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7565943238731219,
+      "grad_norm": 0.5211515839668528,
+      "learning_rate": 7.820554084797739e-06,
+      "loss": 0.2819,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7569282136894825,
+      "grad_norm": 0.4863515742288066,
+      "learning_rate": 7.81814773493595e-06,
+      "loss": 0.2807,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7572621035058431,
+      "grad_norm": 0.47229071006124373,
+      "learning_rate": 7.81574042804576e-06,
+      "loss": 0.2961,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7575959933222036,
+      "grad_norm": 0.4796243298481503,
+      "learning_rate": 7.81333216494468e-06,
+      "loss": 0.2734,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7579298831385642,
+      "grad_norm": 0.5516199609177617,
+      "learning_rate": 7.810922946450543e-06,
+      "loss": 0.2958,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7582637729549249,
+      "grad_norm": 0.5384698975964798,
+      "learning_rate": 7.808512773381506e-06,
+      "loss": 0.2994,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7585976627712855,
+      "grad_norm": 0.5069435286055777,
+      "learning_rate": 7.806101646556052e-06,
+      "loss": 0.2837,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7589315525876461,
+      "grad_norm": 0.5511530484462736,
+      "learning_rate": 7.803689566792989e-06,
+      "loss": 0.2972,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7592654424040067,
+      "grad_norm": 0.533707967176556,
+      "learning_rate": 7.801276534911442e-06,
+      "loss": 0.3051,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7595993322203672,
+      "grad_norm": 0.4662764416613889,
+      "learning_rate": 7.79886255173087e-06,
+      "loss": 0.2786,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7599332220367279,
+      "grad_norm": 0.5321502246829226,
+      "learning_rate": 7.796447618071047e-06,
+      "loss": 0.2838,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7602671118530885,
+      "grad_norm": 0.5402824041809543,
+      "learning_rate": 7.79403173475207e-06,
+      "loss": 0.2874,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7606010016694491,
+      "grad_norm": 0.5078085257991221,
+      "learning_rate": 7.791614902594363e-06,
+      "loss": 0.2861,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7609348914858097,
+      "grad_norm": 0.48852488074510797,
+      "learning_rate": 7.789197122418668e-06,
+      "loss": 0.2878,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7612687813021702,
+      "grad_norm": 0.5120389150254007,
+      "learning_rate": 7.786778395046053e-06,
+      "loss": 0.2792,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7616026711185309,
+      "grad_norm": 0.507711153574066,
+      "learning_rate": 7.784358721297902e-06,
+      "loss": 0.2857,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7619365609348915,
+      "grad_norm": 0.4896633104440327,
+      "learning_rate": 7.781938101995928e-06,
+      "loss": 0.2915,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7622704507512521,
+      "grad_norm": 0.5031612370486945,
+      "learning_rate": 7.779516537962154e-06,
+      "loss": 0.2834,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7626043405676127,
+      "grad_norm": 0.510461214871243,
+      "learning_rate": 7.777094030018936e-06,
+      "loss": 0.29,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7629382303839732,
+      "grad_norm": 0.4805508004395844,
+      "learning_rate": 7.774670578988946e-06,
+      "loss": 0.2826,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7632721202003339,
+      "grad_norm": 0.4952054828864801,
+      "learning_rate": 7.772246185695174e-06,
+      "loss": 0.2772,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7636060100166945,
+      "grad_norm": 0.4762498431160314,
+      "learning_rate": 7.769820850960928e-06,
+      "loss": 0.2775,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7639398998330551,
+      "grad_norm": 0.46081880675840103,
+      "learning_rate": 7.767394575609843e-06,
+      "loss": 0.3002,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7642737896494157,
+      "grad_norm": 0.47181166141765735,
+      "learning_rate": 7.76496736046587e-06,
+      "loss": 0.2812,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7646076794657763,
+      "grad_norm": 0.5204480790627193,
+      "learning_rate": 7.762539206353276e-06,
+      "loss": 0.2964,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7649415692821369,
+      "grad_norm": 0.4821837845190929,
+      "learning_rate": 7.76011011409665e-06,
+      "loss": 0.2816,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7652754590984975,
+      "grad_norm": 0.48095943312797,
+      "learning_rate": 7.757680084520902e-06,
+      "loss": 0.2871,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7656093489148581,
+      "grad_norm": 0.4795910788930229,
+      "learning_rate": 7.755249118451256e-06,
+      "loss": 0.2751,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7659432387312187,
+      "grad_norm": 0.49350399366778563,
+      "learning_rate": 7.752817216713254e-06,
+      "loss": 0.2869,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7662771285475793,
+      "grad_norm": 0.5027663673696087,
+      "learning_rate": 7.750384380132756e-06,
+      "loss": 0.289,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7666110183639399,
+      "grad_norm": 0.47378994158810006,
+      "learning_rate": 7.747950609535947e-06,
+      "loss": 0.2713,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7669449081803005,
+      "grad_norm": 0.470937919215691,
+      "learning_rate": 7.745515905749315e-06,
+      "loss": 0.2903,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7672787979966611,
+      "grad_norm": 0.49491154728094966,
+      "learning_rate": 7.74308026959968e-06,
+      "loss": 0.282,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7676126878130217,
+      "grad_norm": 0.47858033851060067,
+      "learning_rate": 7.740643701914166e-06,
+      "loss": 0.2725,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7679465776293823,
+      "grad_norm": 0.5397725053302314,
+      "learning_rate": 7.738206203520223e-06,
+      "loss": 0.2904,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7682804674457429,
+      "grad_norm": 0.4832087274309526,
+      "learning_rate": 7.735767775245608e-06,
+      "loss": 0.288,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7686143572621035,
+      "grad_norm": 0.4724703717457446,
+      "learning_rate": 7.733328417918405e-06,
+      "loss": 0.2707,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7689482470784641,
+      "grad_norm": 0.46376462113706035,
+      "learning_rate": 7.730888132367e-06,
+      "loss": 0.2822,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7692821368948247,
+      "grad_norm": 0.5051668836713441,
+      "learning_rate": 7.728446919420108e-06,
+      "loss": 0.2905,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7696160267111853,
+      "grad_norm": 0.4760676314282702,
+      "learning_rate": 7.726004779906748e-06,
+      "loss": 0.2829,
+      "step": 2305
+    },
+    {
+      "epoch": 0.769949916527546,
+      "grad_norm": 0.5022629228032721,
+      "learning_rate": 7.723561714656258e-06,
+      "loss": 0.2867,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7702838063439065,
+      "grad_norm": 0.479183453283485,
+      "learning_rate": 7.721117724498293e-06,
+      "loss": 0.2911,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7706176961602671,
+      "grad_norm": 0.47653921520739817,
+      "learning_rate": 7.718672810262818e-06,
+      "loss": 0.2817,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7709515859766277,
+      "grad_norm": 0.4604658261404663,
+      "learning_rate": 7.716226972780111e-06,
+      "loss": 0.2764,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7712854757929883,
+      "grad_norm": 0.4852251251342805,
+      "learning_rate": 7.71378021288077e-06,
+      "loss": 0.285,
+      "step": 2310
+    },
+    {
+      "epoch": 0.771619365609349,
+      "grad_norm": 0.5234983224715798,
+      "learning_rate": 7.7113325313957e-06,
+      "loss": 0.2971,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7719532554257095,
+      "grad_norm": 0.49048097552012443,
+      "learning_rate": 7.70888392915612e-06,
+      "loss": 0.2857,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7722871452420701,
+      "grad_norm": 0.4721276587550153,
+      "learning_rate": 7.706434406993566e-06,
+      "loss": 0.2841,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7726210350584307,
+      "grad_norm": 0.4884205563113079,
+      "learning_rate": 7.703983965739878e-06,
+      "loss": 0.2772,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7729549248747913,
+      "grad_norm": 0.4915323030792945,
+      "learning_rate": 7.701532606227216e-06,
+      "loss": 0.2883,
+      "step": 2315
+    },
+    {
+      "epoch": 0.773288814691152,
+      "grad_norm": 0.4935917458850107,
+      "learning_rate": 7.69908032928805e-06,
+      "loss": 0.2852,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7736227045075125,
+      "grad_norm": 0.46754516010247876,
+      "learning_rate": 7.696627135755159e-06,
+      "loss": 0.2874,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7739565943238731,
+      "grad_norm": 0.46940800066184196,
+      "learning_rate": 7.694173026461634e-06,
+      "loss": 0.283,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7742904841402337,
+      "grad_norm": 0.5157969841670057,
+      "learning_rate": 7.69171800224088e-06,
+      "loss": 0.2924,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7746243739565943,
+      "grad_norm": 0.48700857541759557,
+      "learning_rate": 7.68926206392661e-06,
+      "loss": 0.2905,
+      "step": 2320
+    },
+    {
+      "epoch": 0.774958263772955,
+      "grad_norm": 0.47252248694152965,
+      "learning_rate": 7.686805212352848e-06,
+      "loss": 0.2905,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7752921535893156,
+      "grad_norm": 0.4709488955561574,
+      "learning_rate": 7.684347448353926e-06,
+      "loss": 0.2731,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7756260434056761,
+      "grad_norm": 0.5249174706651818,
+      "learning_rate": 7.681888772764493e-06,
+      "loss": 0.2847,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7759599332220367,
+      "grad_norm": 0.5115124707839976,
+      "learning_rate": 7.679429186419496e-06,
+      "loss": 0.2926,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7762938230383973,
+      "grad_norm": 0.4795361601243223,
+      "learning_rate": 7.6769686901542e-06,
+      "loss": 0.2863,
+      "step": 2325
+    },
+    {
+      "epoch": 0.776627712854758,
+      "grad_norm": 0.516956908609677,
+      "learning_rate": 7.67450728480418e-06,
+      "loss": 0.2834,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7769616026711186,
+      "grad_norm": 0.49202138011803104,
+      "learning_rate": 7.672044971205315e-06,
+      "loss": 0.2739,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7772954924874791,
+      "grad_norm": 0.4887411616992647,
+      "learning_rate": 7.669581750193791e-06,
+      "loss": 0.2939,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7776293823038397,
+      "grad_norm": 0.47643931107347165,
+      "learning_rate": 7.667117622606108e-06,
+      "loss": 0.2841,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7779632721202003,
+      "grad_norm": 0.46804078900960283,
+      "learning_rate": 7.664652589279071e-06,
+      "loss": 0.2793,
+      "step": 2330
+    },
+    {
+      "epoch": 0.778297161936561,
+      "grad_norm": 0.46995193127808743,
+      "learning_rate": 7.66218665104979e-06,
+      "loss": 0.2778,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7786310517529216,
+      "grad_norm": 0.4927704500605226,
+      "learning_rate": 7.65971980875569e-06,
+      "loss": 0.2958,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7789649415692821,
+      "grad_norm": 0.4691366472585806,
+      "learning_rate": 7.657252063234493e-06,
+      "loss": 0.2798,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7792988313856427,
+      "grad_norm": 0.49859740608937436,
+      "learning_rate": 7.654783415324234e-06,
+      "loss": 0.2802,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7796327212020033,
+      "grad_norm": 0.47233970097910866,
+      "learning_rate": 7.652313865863254e-06,
+      "loss": 0.2811,
+      "step": 2335
+    },
+    {
+      "epoch": 0.779966611018364,
+      "grad_norm": 0.5118055490604365,
+      "learning_rate": 7.649843415690198e-06,
+      "loss": 0.2859,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7803005008347246,
+      "grad_norm": 0.5038607229706572,
+      "learning_rate": 7.647372065644019e-06,
+      "loss": 0.2858,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7806343906510852,
+      "grad_norm": 0.4892572205461613,
+      "learning_rate": 7.644899816563975e-06,
+      "loss": 0.2794,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7809682804674457,
+      "grad_norm": 0.4788308914453706,
+      "learning_rate": 7.642426669289627e-06,
+      "loss": 0.2859,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7813021702838063,
+      "grad_norm": 0.47694416396462425,
+      "learning_rate": 7.639952624660847e-06,
+      "loss": 0.303,
+      "step": 2340
+    },
+    {
+      "epoch": 0.781636060100167,
+      "grad_norm": 0.5208191627461385,
+      "learning_rate": 7.637477683517802e-06,
+      "loss": 0.2867,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7819699499165276,
+      "grad_norm": 0.5108306778638265,
+      "learning_rate": 7.635001846700973e-06,
+      "loss": 0.2862,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7823038397328882,
+      "grad_norm": 0.5192665929414639,
+      "learning_rate": 7.632525115051142e-06,
+      "loss": 0.2978,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7826377295492487,
+      "grad_norm": 0.4920829250184859,
+      "learning_rate": 7.630047489409393e-06,
+      "loss": 0.2823,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7829716193656093,
+      "grad_norm": 0.47758246755462647,
+      "learning_rate": 7.627568970617114e-06,
+      "loss": 0.2775,
+      "step": 2345
+    },
+    {
+      "epoch": 0.78330550918197,
+      "grad_norm": 0.4757167405005467,
+      "learning_rate": 7.625089559515999e-06,
+      "loss": 0.2787,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7836393989983306,
+      "grad_norm": 0.4751729440033226,
+      "learning_rate": 7.622609256948043e-06,
+      "loss": 0.2834,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7839732888146912,
+      "grad_norm": 0.4870161461504156,
+      "learning_rate": 7.620128063755541e-06,
+      "loss": 0.2918,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7843071786310517,
+      "grad_norm": 0.5161106859005168,
+      "learning_rate": 7.617645980781097e-06,
+      "loss": 0.2957,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7846410684474123,
+      "grad_norm": 0.5533737170858487,
+      "learning_rate": 7.6151630088676125e-06,
+      "loss": 0.309,
+      "step": 2350
+    },
+    {
+      "epoch": 0.784974958263773,
+      "grad_norm": 0.4777347191546943,
+      "learning_rate": 7.612679148858291e-06,
+      "loss": 0.2867,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7853088480801336,
+      "grad_norm": 0.5048262049881393,
+      "learning_rate": 7.610194401596638e-06,
+      "loss": 0.2848,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7856427378964942,
+      "grad_norm": 0.5033921289870031,
+      "learning_rate": 7.607708767926463e-06,
+      "loss": 0.2899,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7859766277128547,
+      "grad_norm": 0.5041585674940261,
+      "learning_rate": 7.6052222486918725e-06,
+      "loss": 0.2887,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7863105175292153,
+      "grad_norm": 0.46520090153806426,
+      "learning_rate": 7.602734844737278e-06,
+      "loss": 0.2801,
+      "step": 2355
+    },
+    {
+      "epoch": 0.786644407345576,
+      "grad_norm": 0.47875240571738625,
+      "learning_rate": 7.600246556907386e-06,
+      "loss": 0.2889,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7869782971619366,
+      "grad_norm": 0.5088479120432765,
+      "learning_rate": 7.5977573860472065e-06,
+      "loss": 0.28,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7873121869782972,
+      "grad_norm": 0.5265954675929836,
+      "learning_rate": 7.595267333002052e-06,
+      "loss": 0.2911,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7876460767946578,
+      "grad_norm": 0.48406060024358155,
+      "learning_rate": 7.592776398617528e-06,
+      "loss": 0.2937,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7879799666110183,
+      "grad_norm": 0.46940941674435066,
+      "learning_rate": 7.590284583739547e-06,
+      "loss": 0.2817,
+      "step": 2360
+    },
+    {
+      "epoch": 0.788313856427379,
+      "grad_norm": 0.5093372965165733,
+      "learning_rate": 7.587791889214312e-06,
+      "loss": 0.2909,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7886477462437396,
+      "grad_norm": 0.5223971729657635,
+      "learning_rate": 7.585298315888332e-06,
+      "loss": 0.3005,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7889816360601002,
+      "grad_norm": 0.49878843509358384,
+      "learning_rate": 7.582803864608411e-06,
+      "loss": 0.2917,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7893155258764608,
+      "grad_norm": 0.4775202228253381,
+      "learning_rate": 7.580308536221651e-06,
+      "loss": 0.2784,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7896494156928213,
+      "grad_norm": 0.46705950530380613,
+      "learning_rate": 7.5778123315754535e-06,
+      "loss": 0.2808,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7899833055091819,
+      "grad_norm": 0.518928112509247,
+      "learning_rate": 7.5753152515175165e-06,
+      "loss": 0.296,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7903171953255426,
+      "grad_norm": 0.4580381928397729,
+      "learning_rate": 7.572817296895835e-06,
+      "loss": 0.2769,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7906510851419032,
+      "grad_norm": 0.4827224127377342,
+      "learning_rate": 7.570318468558704e-06,
+      "loss": 0.2889,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7909849749582638,
+      "grad_norm": 0.4821409370969341,
+      "learning_rate": 7.567818767354708e-06,
+      "loss": 0.2815,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7913188647746243,
+      "grad_norm": 0.4950307906933506,
+      "learning_rate": 7.565318194132736e-06,
+      "loss": 0.2845,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7916527545909849,
+      "grad_norm": 0.4589934063809129,
+      "learning_rate": 7.56281674974197e-06,
+      "loss": 0.2829,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7919866444073456,
+      "grad_norm": 0.4998009134140173,
+      "learning_rate": 7.560314435031886e-06,
+      "loss": 0.2826,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7923205342237062,
+      "grad_norm": 0.47352659128796865,
+      "learning_rate": 7.557811250852256e-06,
+      "loss": 0.2848,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7926544240400668,
+      "grad_norm": 0.4792846832753459,
+      "learning_rate": 7.5553071980531526e-06,
+      "loss": 0.2746,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7929883138564274,
+      "grad_norm": 0.45844947507747746,
+      "learning_rate": 7.5528022774849344e-06,
+      "loss": 0.2814,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7933222036727879,
+      "grad_norm": 0.4931586410441206,
+      "learning_rate": 7.550296489998264e-06,
+      "loss": 0.2828,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7936560934891486,
+      "grad_norm": 0.47843116646312484,
+      "learning_rate": 7.547789836444088e-06,
+      "loss": 0.2679,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7939899833055092,
+      "grad_norm": 0.4900375037688527,
+      "learning_rate": 7.54528231767366e-06,
+      "loss": 0.2696,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7943238731218698,
+      "grad_norm": 0.5224428336323976,
+      "learning_rate": 7.542773934538515e-06,
+      "loss": 0.3028,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7946577629382304,
+      "grad_norm": 0.4889197861899141,
+      "learning_rate": 7.540264687890489e-06,
+      "loss": 0.2781,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7949916527545909,
+      "grad_norm": 0.511746742283747,
+      "learning_rate": 7.537754578581711e-06,
+      "loss": 0.2957,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7953255425709516,
+      "grad_norm": 0.4912543182543457,
+      "learning_rate": 7.535243607464597e-06,
+      "loss": 0.2993,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7956594323873122,
+      "grad_norm": 0.49394982010776783,
+      "learning_rate": 7.532731775391864e-06,
+      "loss": 0.2854,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7959933222036728,
+      "grad_norm": 0.49325075483994407,
+      "learning_rate": 7.530219083216512e-06,
+      "loss": 0.2785,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7963272120200334,
+      "grad_norm": 0.4891249734044441,
+      "learning_rate": 7.527705531791844e-06,
+      "loss": 0.2677,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7966611018363939,
+      "grad_norm": 0.4933458176418238,
+      "learning_rate": 7.525191121971447e-06,
+      "loss": 0.2882,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7969949916527546,
+      "grad_norm": 0.4988985703557823,
+      "learning_rate": 7.5226758546092e-06,
+      "loss": 0.2858,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7973288814691152,
+      "grad_norm": 0.4891806194810961,
+      "learning_rate": 7.520159730559278e-06,
+      "loss": 0.2841,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7976627712854758,
+      "grad_norm": 0.5233081175948001,
+      "learning_rate": 7.51764275067614e-06,
+      "loss": 0.2866,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7979966611018364,
+      "grad_norm": 0.48881353911075914,
+      "learning_rate": 7.51512491581454e-06,
+      "loss": 0.2965,
+      "step": 2390
+    },
+    {
+      "epoch": 0.798330550918197,
+      "grad_norm": 0.47516173029945874,
+      "learning_rate": 7.512606226829524e-06,
+      "loss": 0.2874,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7986644407345576,
+      "grad_norm": 0.5320496352904915,
+      "learning_rate": 7.510086684576425e-06,
+      "loss": 0.2814,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7989983305509182,
+      "grad_norm": 0.5101849352098333,
+      "learning_rate": 7.5075662899108636e-06,
+      "loss": 0.2746,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7993322203672788,
+      "grad_norm": 0.48367765820563374,
+      "learning_rate": 7.505045043688758e-06,
+      "loss": 0.2843,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7996661101836394,
+      "grad_norm": 0.4784742296101321,
+      "learning_rate": 7.502522946766306e-06,
+      "loss": 0.2871,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48440101786988327,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2898,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8003338898163606,
+      "grad_norm": 0.49189918439983754,
+      "learning_rate": 7.497476204246621e-06,
+      "loss": 0.2955,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8006677796327212,
+      "grad_norm": 0.47599708733397816,
+      "learning_rate": 7.494951560363235e-06,
+      "loss": 0.3067,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8010016694490818,
+      "grad_norm": 0.4847547400907571,
+      "learning_rate": 7.4924260692072e-06,
+      "loss": 0.2831,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8013355592654424,
+      "grad_norm": 0.4646081120785199,
+      "learning_rate": 7.489899731636159e-06,
+      "loss": 0.2779,
+      "step": 2400
+    },
+    {
+      "epoch": 0.801669449081803,
+      "grad_norm": 0.4938822126951617,
+      "learning_rate": 7.487372548508041e-06,
+      "loss": 0.2782,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8020033388981636,
+      "grad_norm": 0.5024820467688039,
+      "learning_rate": 7.484844520681066e-06,
+      "loss": 0.2791,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8023372287145242,
+      "grad_norm": 0.5313729034281754,
+      "learning_rate": 7.482315649013741e-06,
+      "loss": 0.2837,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8026711185308848,
+      "grad_norm": 0.4832141425917904,
+      "learning_rate": 7.479785934364856e-06,
+      "loss": 0.2859,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8030050083472454,
+      "grad_norm": 0.5118714146542723,
+      "learning_rate": 7.477255377593489e-06,
+      "loss": 0.2914,
+      "step": 2405
+    },
+    {
+      "epoch": 0.803338898163606,
+      "grad_norm": 0.4855499962190082,
+      "learning_rate": 7.474723979559003e-06,
+      "loss": 0.2991,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8036727879799667,
+      "grad_norm": 0.5123245248596919,
+      "learning_rate": 7.4721917411210484e-06,
+      "loss": 0.2877,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8040066777963272,
+      "grad_norm": 0.5037312749840179,
+      "learning_rate": 7.4696586631395626e-06,
+      "loss": 0.2783,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8043405676126878,
+      "grad_norm": 0.5394065723507806,
+      "learning_rate": 7.4671247464747635e-06,
+      "loss": 0.2892,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8046744574290484,
+      "grad_norm": 0.5329879964450243,
+      "learning_rate": 7.4645899919871545e-06,
+      "loss": 0.2833,
+      "step": 2410
+    },
+    {
+      "epoch": 0.805008347245409,
+      "grad_norm": 0.5095488451427401,
+      "learning_rate": 7.462054400537529e-06,
+      "loss": 0.2852,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8053422370617697,
+      "grad_norm": 0.49508212392105305,
+      "learning_rate": 7.4595179729869584e-06,
+      "loss": 0.281,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8056761268781302,
+      "grad_norm": 0.5203268419619876,
+      "learning_rate": 7.456980710196801e-06,
+      "loss": 0.2918,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8060100166944908,
+      "grad_norm": 0.5143923674268233,
+      "learning_rate": 7.454442613028696e-06,
+      "loss": 0.2853,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8063439065108514,
+      "grad_norm": 0.5016629747721937,
+      "learning_rate": 7.451903682344572e-06,
+      "loss": 0.2841,
+      "step": 2415
+    },
+    {
+      "epoch": 0.806677796327212,
+      "grad_norm": 0.5138484062733644,
+      "learning_rate": 7.449363919006635e-06,
+      "loss": 0.2819,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8070116861435727,
+      "grad_norm": 0.5499453438811575,
+      "learning_rate": 7.446823323877375e-06,
+      "loss": 0.2882,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8073455759599332,
+      "grad_norm": 0.5045221784290563,
+      "learning_rate": 7.444281897819563e-06,
+      "loss": 0.2842,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8076794657762938,
+      "grad_norm": 0.48675616518989995,
+      "learning_rate": 7.44173964169626e-06,
+      "loss": 0.2892,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8080133555926544,
+      "grad_norm": 0.5064839881021352,
+      "learning_rate": 7.439196556370796e-06,
+      "loss": 0.3007,
+      "step": 2420
+    },
+    {
+      "epoch": 0.808347245409015,
+      "grad_norm": 0.4721143036482007,
+      "learning_rate": 7.436652642706795e-06,
+      "loss": 0.2787,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8086811352253757,
+      "grad_norm": 0.5248943333892839,
+      "learning_rate": 7.434107901568155e-06,
+      "loss": 0.2888,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8090150250417363,
+      "grad_norm": 0.5407968051993631,
+      "learning_rate": 7.431562333819054e-06,
+      "loss": 0.3006,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8093489148580968,
+      "grad_norm": 0.48556448180688494,
+      "learning_rate": 7.429015940323961e-06,
+      "loss": 0.2869,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8096828046744574,
+      "grad_norm": 0.5077637560320363,
+      "learning_rate": 7.426468721947611e-06,
+      "loss": 0.2825,
+      "step": 2425
+    },
+    {
+      "epoch": 0.810016694490818,
+      "grad_norm": 0.46518678335621205,
+      "learning_rate": 7.423920679555029e-06,
+      "loss": 0.2662,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8103505843071787,
+      "grad_norm": 0.4758557358565174,
+      "learning_rate": 7.421371814011516e-06,
+      "loss": 0.2836,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8106844741235393,
+      "grad_norm": 0.5225388559750322,
+      "learning_rate": 7.418822126182655e-06,
+      "loss": 0.2769,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8110183639398998,
+      "grad_norm": 0.5260219867590993,
+      "learning_rate": 7.416271616934308e-06,
+      "loss": 0.2988,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8113522537562604,
+      "grad_norm": 0.4716686678555991,
+      "learning_rate": 7.41372028713261e-06,
+      "loss": 0.2806,
+      "step": 2430
+    },
+    {
+      "epoch": 0.811686143572621,
+      "grad_norm": 0.47697516764584275,
+      "learning_rate": 7.4111681376439824e-06,
+      "loss": 0.2695,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8120200333889817,
+      "grad_norm": 0.5035126160362945,
+      "learning_rate": 7.4086151693351235e-06,
+      "loss": 0.282,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8123539232053423,
+      "grad_norm": 0.476827966254471,
+      "learning_rate": 7.406061383073007e-06,
+      "loss": 0.2696,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8126878130217028,
+      "grad_norm": 0.5080956810847602,
+      "learning_rate": 7.403506779724883e-06,
+      "loss": 0.2946,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8130217028380634,
+      "grad_norm": 0.47293852937943937,
+      "learning_rate": 7.400951360158285e-06,
+      "loss": 0.2727,
+      "step": 2435
+    },
+    {
+      "epoch": 0.813355592654424,
+      "grad_norm": 0.4816221877791688,
+      "learning_rate": 7.398395125241019e-06,
+      "loss": 0.2881,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8136894824707847,
+      "grad_norm": 0.4726810337288391,
+      "learning_rate": 7.395838075841169e-06,
+      "loss": 0.2911,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8140233722871453,
+      "grad_norm": 0.47993556437804563,
+      "learning_rate": 7.393280212827095e-06,
+      "loss": 0.2888,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8143572621035059,
+      "grad_norm": 0.5238610203253281,
+      "learning_rate": 7.390721537067436e-06,
+      "loss": 0.2842,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8146911519198664,
+      "grad_norm": 0.4569325607121092,
+      "learning_rate": 7.388162049431102e-06,
+      "loss": 0.2844,
+      "step": 2440
+    },
+    {
+      "epoch": 0.815025041736227,
+      "grad_norm": 0.4640308581102319,
+      "learning_rate": 7.385601750787285e-06,
+      "loss": 0.2767,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8153589315525877,
+      "grad_norm": 0.5132125975311789,
+      "learning_rate": 7.383040642005448e-06,
+      "loss": 0.2884,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8156928213689483,
+      "grad_norm": 0.5117784807889557,
+      "learning_rate": 7.38047872395533e-06,
+      "loss": 0.2945,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8160267111853089,
+      "grad_norm": 0.46123847459698025,
+      "learning_rate": 7.377915997506945e-06,
+      "loss": 0.289,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8163606010016694,
+      "grad_norm": 0.49329107901371533,
+      "learning_rate": 7.375352463530583e-06,
+      "loss": 0.2856,
+      "step": 2445
+    },
+    {
+      "epoch": 0.81669449081803,
+      "grad_norm": 0.5268299174305398,
+      "learning_rate": 7.3727881228968045e-06,
+      "loss": 0.2928,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8170283806343906,
+      "grad_norm": 0.5264669500698718,
+      "learning_rate": 7.370222976476449e-06,
+      "loss": 0.2972,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8173622704507513,
+      "grad_norm": 0.4706565759386983,
+      "learning_rate": 7.367657025140623e-06,
+      "loss": 0.296,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8176961602671119,
+      "grad_norm": 0.4809765576786834,
+      "learning_rate": 7.365090269760715e-06,
+      "loss": 0.2872,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8180300500834724,
+      "grad_norm": 0.5339765798813707,
+      "learning_rate": 7.3625227112083796e-06,
+      "loss": 0.2969,
+      "step": 2450
+    },
+    {
+      "epoch": 0.818363939899833,
+      "grad_norm": 0.5161637118001101,
+      "learning_rate": 7.359954350355547e-06,
+      "loss": 0.291,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8186978297161936,
+      "grad_norm": 0.48135701165187184,
+      "learning_rate": 7.357385188074419e-06,
+      "loss": 0.2847,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8190317195325543,
+      "grad_norm": 0.4708563453417966,
+      "learning_rate": 7.354815225237468e-06,
+      "loss": 0.2845,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8193656093489149,
+      "grad_norm": 0.5063996748860169,
+      "learning_rate": 7.352244462717444e-06,
+      "loss": 0.2822,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8196994991652755,
+      "grad_norm": 0.560412598544782,
+      "learning_rate": 7.349672901387362e-06,
+      "loss": 0.2882,
+      "step": 2455
+    },
+    {
+      "epoch": 0.820033388981636,
+      "grad_norm": 0.5115360028391183,
+      "learning_rate": 7.34710054212051e-06,
+      "loss": 0.2867,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8203672787979966,
+      "grad_norm": 0.57049610624446,
+      "learning_rate": 7.344527385790449e-06,
+      "loss": 0.2921,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8207011686143573,
+      "grad_norm": 0.49563994129883543,
+      "learning_rate": 7.341953433271011e-06,
+      "loss": 0.2839,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8210350584307179,
+      "grad_norm": 0.49270964739317613,
+      "learning_rate": 7.339378685436296e-06,
+      "loss": 0.29,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8213689482470785,
+      "grad_norm": 0.5123037620333541,
+      "learning_rate": 7.336803143160674e-06,
+      "loss": 0.2893,
+      "step": 2460
+    },
+    {
+      "epoch": 0.821702838063439,
+      "grad_norm": 0.4994694613145348,
+      "learning_rate": 7.334226807318787e-06,
+      "loss": 0.2868,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8220367278797996,
+      "grad_norm": 0.498546324840962,
+      "learning_rate": 7.331649678785545e-06,
+      "loss": 0.2978,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8223706176961603,
+      "grad_norm": 0.5298987830095987,
+      "learning_rate": 7.329071758436128e-06,
+      "loss": 0.2753,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8227045075125209,
+      "grad_norm": 0.49676745482837376,
+      "learning_rate": 7.3264930471459835e-06,
+      "loss": 0.2887,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8230383973288815,
+      "grad_norm": 0.49773337751928975,
+      "learning_rate": 7.3239135457908295e-06,
+      "loss": 0.2926,
+      "step": 2465
+    },
+    {
+      "epoch": 0.823372287145242,
+      "grad_norm": 0.5110251241496923,
+      "learning_rate": 7.32133325524665e-06,
+      "loss": 0.2876,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8237061769616026,
+      "grad_norm": 0.5163471783881414,
+      "learning_rate": 7.318752176389701e-06,
+      "loss": 0.288,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8240400667779633,
+      "grad_norm": 0.5300058832522858,
+      "learning_rate": 7.3161703100964995e-06,
+      "loss": 0.2835,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8243739565943239,
+      "grad_norm": 0.4955729877687883,
+      "learning_rate": 7.313587657243837e-06,
+      "loss": 0.3002,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8247078464106845,
+      "grad_norm": 0.5439200752308094,
+      "learning_rate": 7.311004218708769e-06,
+      "loss": 0.2897,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8250417362270451,
+      "grad_norm": 0.5281932809621575,
+      "learning_rate": 7.308419995368616e-06,
+      "loss": 0.2881,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8253756260434056,
+      "grad_norm": 0.5742448596547415,
+      "learning_rate": 7.30583498810097e-06,
+      "loss": 0.291,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8257095158597663,
+      "grad_norm": 0.4848004695858507,
+      "learning_rate": 7.303249197783682e-06,
+      "loss": 0.2815,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8260434056761269,
+      "grad_norm": 0.5116227508879949,
+      "learning_rate": 7.300662625294874e-06,
+      "loss": 0.2896,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8263772954924875,
+      "grad_norm": 0.5150292840621444,
+      "learning_rate": 7.2980752715129354e-06,
+      "loss": 0.2847,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8267111853088481,
+      "grad_norm": 0.5827117614513208,
+      "learning_rate": 7.295487137316517e-06,
+      "loss": 0.3028,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8270450751252086,
+      "grad_norm": 0.5051458851853271,
+      "learning_rate": 7.292898223584535e-06,
+      "loss": 0.2941,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8273789649415693,
+      "grad_norm": 0.4886017437203344,
+      "learning_rate": 7.290308531196171e-06,
+      "loss": 0.2823,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8277128547579299,
+      "grad_norm": 0.48691817881921834,
+      "learning_rate": 7.287718061030872e-06,
+      "loss": 0.2857,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8280467445742905,
+      "grad_norm": 0.5595926556917534,
+      "learning_rate": 7.285126813968346e-06,
+      "loss": 0.2866,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8283806343906511,
+      "grad_norm": 0.5280581658166001,
+      "learning_rate": 7.282534790888571e-06,
+      "loss": 0.2897,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8287145242070116,
+      "grad_norm": 0.4901923888247075,
+      "learning_rate": 7.27994199267178e-06,
+      "loss": 0.2798,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8290484140233723,
+      "grad_norm": 0.47817451213723716,
+      "learning_rate": 7.27734842019848e-06,
+      "loss": 0.2887,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8293823038397329,
+      "grad_norm": 0.4872967193702052,
+      "learning_rate": 7.274754074349429e-06,
+      "loss": 0.2804,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8297161936560935,
+      "grad_norm": 0.5131630657291192,
+      "learning_rate": 7.2721589560056575e-06,
+      "loss": 0.2866,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8300500834724541,
+      "grad_norm": 0.49467092325038503,
+      "learning_rate": 7.269563066048451e-06,
+      "loss": 0.2839,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8303839732888146,
+      "grad_norm": 0.46793641241040623,
+      "learning_rate": 7.266966405359361e-06,
+      "loss": 0.2731,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8307178631051753,
+      "grad_norm": 0.49418094934563983,
+      "learning_rate": 7.264368974820202e-06,
+      "loss": 0.2748,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8310517529215359,
+      "grad_norm": 0.538073040489842,
+      "learning_rate": 7.2617707753130465e-06,
+      "loss": 0.2955,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8313856427378965,
+      "grad_norm": 0.5627201821243142,
+      "learning_rate": 7.259171807720231e-06,
+      "loss": 0.2998,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8317195325542571,
+      "grad_norm": 0.5091374153444668,
+      "learning_rate": 7.256572072924347e-06,
+      "loss": 0.2885,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8320534223706177,
+      "grad_norm": 0.5050790730749642,
+      "learning_rate": 7.253971571808258e-06,
+      "loss": 0.2859,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8323873121869783,
+      "grad_norm": 0.4776891889789473,
+      "learning_rate": 7.2513703052550746e-06,
+      "loss": 0.2825,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8327212020033389,
+      "grad_norm": 0.5147700580297773,
+      "learning_rate": 7.248768274148178e-06,
+      "loss": 0.2895,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8330550918196995,
+      "grad_norm": 0.5085344827347269,
+      "learning_rate": 7.246165479371201e-06,
+      "loss": 0.2802,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8333889816360601,
+      "grad_norm": 0.4799569449219186,
+      "learning_rate": 7.243561921808041e-06,
+      "loss": 0.2893,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8337228714524207,
+      "grad_norm": 0.49485885939897517,
+      "learning_rate": 7.240957602342853e-06,
+      "loss": 0.2849,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8340567612687813,
+      "grad_norm": 0.47644336556969413,
+      "learning_rate": 7.238352521860049e-06,
+      "loss": 0.2779,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8343906510851419,
+      "grad_norm": 0.5325041303939819,
+      "learning_rate": 7.235746681244302e-06,
+      "loss": 0.2865,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8347245409015025,
+      "grad_norm": 0.4812331161000099,
+      "learning_rate": 7.233140081380541e-06,
+      "loss": 0.2867,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8350584307178631,
+      "grad_norm": 0.49998337759596806,
+      "learning_rate": 7.230532723153956e-06,
+      "loss": 0.2918,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8353923205342237,
+      "grad_norm": 0.4780540555669566,
+      "learning_rate": 7.2279246074499895e-06,
+      "loss": 0.2851,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8357262103505843,
+      "grad_norm": 0.5482556951926307,
+      "learning_rate": 7.225315735154346e-06,
+      "loss": 0.2776,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8360601001669449,
+      "grad_norm": 0.47137753094241464,
+      "learning_rate": 7.222706107152986e-06,
+      "loss": 0.284,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8363939899833055,
+      "grad_norm": 0.5087111947462134,
+      "learning_rate": 7.220095724332122e-06,
+      "loss": 0.2913,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8367278797996661,
+      "grad_norm": 0.45688406968366535,
+      "learning_rate": 7.2174845875782305e-06,
+      "loss": 0.2777,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8370617696160267,
+      "grad_norm": 0.49493806629429277,
+      "learning_rate": 7.214872697778037e-06,
+      "loss": 0.2925,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8373956594323874,
+      "grad_norm": 0.4892791341764886,
+      "learning_rate": 7.212260055818528e-06,
+      "loss": 0.2785,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8377295492487479,
+      "grad_norm": 0.49748340254330115,
+      "learning_rate": 7.209646662586941e-06,
+      "loss": 0.2883,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8380634390651085,
+      "grad_norm": 0.47081742018316924,
+      "learning_rate": 7.207032518970774e-06,
+      "loss": 0.2931,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8383973288814691,
+      "grad_norm": 0.47307382814504984,
+      "learning_rate": 7.204417625857772e-06,
+      "loss": 0.2821,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8387312186978297,
+      "grad_norm": 0.4688308634499861,
+      "learning_rate": 7.2018019841359435e-06,
+      "loss": 0.287,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8390651085141904,
+      "grad_norm": 0.48523130383393387,
+      "learning_rate": 7.199185594693545e-06,
+      "loss": 0.2902,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8393989983305509,
+      "grad_norm": 0.4981201279186025,
+      "learning_rate": 7.1965684584190894e-06,
+      "loss": 0.2978,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8397328881469115,
+      "grad_norm": 0.472258297610023,
+      "learning_rate": 7.193950576201343e-06,
+      "loss": 0.2908,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8400667779632721,
+      "grad_norm": 0.4615259995547763,
+      "learning_rate": 7.191331948929323e-06,
+      "loss": 0.2847,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8404006677796327,
+      "grad_norm": 0.46462996090705067,
+      "learning_rate": 7.188712577492305e-06,
+      "loss": 0.2843,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8407345575959934,
+      "grad_norm": 0.4703294901800178,
+      "learning_rate": 7.18609246277981e-06,
+      "loss": 0.2868,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8410684474123539,
+      "grad_norm": 0.478605444107316,
+      "learning_rate": 7.18347160568162e-06,
+      "loss": 0.2762,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8414023372287145,
+      "grad_norm": 0.4387302986760891,
+      "learning_rate": 7.180850007087761e-06,
+      "loss": 0.2731,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8417362270450751,
+      "grad_norm": 0.46033321867871313,
+      "learning_rate": 7.178227667888513e-06,
+      "loss": 0.2852,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8420701168614357,
+      "grad_norm": 0.47725856709834125,
+      "learning_rate": 7.1756045889744145e-06,
+      "loss": 0.275,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8424040066777964,
+      "grad_norm": 0.4558975245671861,
+      "learning_rate": 7.1729807712362455e-06,
+      "loss": 0.2731,
+      "step": 2523
+    },
+    {
+      "epoch": 0.842737896494157,
+      "grad_norm": 0.47168690426049364,
+      "learning_rate": 7.170356215565042e-06,
+      "loss": 0.2739,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8430717863105175,
+      "grad_norm": 0.47330558016393376,
+      "learning_rate": 7.1677309228520865e-06,
+      "loss": 0.287,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8434056761268781,
+      "grad_norm": 0.5000067696596497,
+      "learning_rate": 7.16510489398892e-06,
+      "loss": 0.2844,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8437395659432387,
+      "grad_norm": 0.4517864993374981,
+      "learning_rate": 7.162478129867326e-06,
+      "loss": 0.2755,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8440734557595994,
+      "grad_norm": 0.4793565110644601,
+      "learning_rate": 7.1598506313793395e-06,
+      "loss": 0.3018,
+      "step": 2528
+    },
+    {
+      "epoch": 0.84440734557596,
+      "grad_norm": 0.48829400031750925,
+      "learning_rate": 7.157222399417245e-06,
+      "loss": 0.283,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8447412353923205,
+      "grad_norm": 0.4507006670532389,
+      "learning_rate": 7.154593434873578e-06,
+      "loss": 0.279,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8450751252086811,
+      "grad_norm": 0.5075502371559258,
+      "learning_rate": 7.151963738641121e-06,
+      "loss": 0.2735,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8454090150250417,
+      "grad_norm": 0.48894084663427717,
+      "learning_rate": 7.149333311612905e-06,
+      "loss": 0.287,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8457429048414024,
+      "grad_norm": 0.47529989048139876,
+      "learning_rate": 7.146702154682208e-06,
+      "loss": 0.273,
+      "step": 2533
+    },
+    {
+      "epoch": 0.846076794657763,
+      "grad_norm": 0.46111696746843484,
+      "learning_rate": 7.14407026874256e-06,
+      "loss": 0.2817,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8464106844741235,
+      "grad_norm": 0.5180039475947849,
+      "learning_rate": 7.141437654687733e-06,
+      "loss": 0.2804,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8467445742904841,
+      "grad_norm": 0.5351090903464721,
+      "learning_rate": 7.138804313411751e-06,
+      "loss": 0.2868,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8470784641068447,
+      "grad_norm": 0.47631434505763043,
+      "learning_rate": 7.136170245808881e-06,
+      "loss": 0.2917,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8474123539232054,
+      "grad_norm": 0.4889272187880478,
+      "learning_rate": 7.133535452773643e-06,
+      "loss": 0.2691,
+      "step": 2538
+    },
+    {
+      "epoch": 0.847746243739566,
+      "grad_norm": 0.48235945113854084,
+      "learning_rate": 7.130899935200794e-06,
+      "loss": 0.2639,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8480801335559266,
+      "grad_norm": 0.5419517058383483,
+      "learning_rate": 7.128263693985347e-06,
+      "loss": 0.2933,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8484140233722871,
+      "grad_norm": 0.49832609537317146,
+      "learning_rate": 7.1256267300225525e-06,
+      "loss": 0.2989,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8487479131886477,
+      "grad_norm": 0.5074970129431398,
+      "learning_rate": 7.12298904420791e-06,
+      "loss": 0.285,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8490818030050084,
+      "grad_norm": 0.47999232461208663,
+      "learning_rate": 7.120350637437166e-06,
+      "loss": 0.2767,
+      "step": 2543
+    },
+    {
+      "epoch": 0.849415692821369,
+      "grad_norm": 0.48888235330511803,
+      "learning_rate": 7.117711510606306e-06,
+      "loss": 0.2964,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8497495826377296,
+      "grad_norm": 0.48080870649969093,
+      "learning_rate": 7.11507166461157e-06,
+      "loss": 0.2828,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8500834724540901,
+      "grad_norm": 0.46284227701540887,
+      "learning_rate": 7.112431100349429e-06,
+      "loss": 0.2792,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8504173622704507,
+      "grad_norm": 0.4972129990902112,
+      "learning_rate": 7.10978981871661e-06,
+      "loss": 0.2779,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8507512520868113,
+      "grad_norm": 0.48881813503339727,
+      "learning_rate": 7.107147820610077e-06,
+      "loss": 0.2801,
+      "step": 2548
+    },
+    {
+      "epoch": 0.851085141903172,
+      "grad_norm": 0.4787436352919017,
+      "learning_rate": 7.1045051069270395e-06,
+      "loss": 0.2847,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8514190317195326,
+      "grad_norm": 0.5075712071705906,
+      "learning_rate": 7.101861678564947e-06,
+      "loss": 0.2966,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8517529215358931,
+      "grad_norm": 0.4697667139621827,
+      "learning_rate": 7.099217536421497e-06,
+      "loss": 0.2916,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8520868113522537,
+      "grad_norm": 0.46092149442512603,
+      "learning_rate": 7.096572681394625e-06,
+      "loss": 0.2704,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8524207011686143,
+      "grad_norm": 0.5107908709358153,
+      "learning_rate": 7.093927114382511e-06,
+      "loss": 0.2787,
+      "step": 2553
+    },
+    {
+      "epoch": 0.852754590984975,
+      "grad_norm": 0.5170528193225149,
+      "learning_rate": 7.091280836283574e-06,
+      "loss": 0.2974,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8530884808013356,
+      "grad_norm": 0.5040710312760669,
+      "learning_rate": 7.088633847996479e-06,
+      "loss": 0.2814,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8534223706176962,
+      "grad_norm": 0.4791918287815672,
+      "learning_rate": 7.085986150420128e-06,
+      "loss": 0.2823,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8537562604340567,
+      "grad_norm": 0.4811384119672847,
+      "learning_rate": 7.083337744453666e-06,
+      "loss": 0.2781,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8540901502504173,
+      "grad_norm": 0.49331746178767477,
+      "learning_rate": 7.080688630996477e-06,
+      "loss": 0.2721,
+      "step": 2558
+    },
+    {
+      "epoch": 0.854424040066778,
+      "grad_norm": 0.4828725305872411,
+      "learning_rate": 7.078038810948187e-06,
+      "loss": 0.2744,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8547579298831386,
+      "grad_norm": 0.4796563330548381,
+      "learning_rate": 7.07538828520866e-06,
+      "loss": 0.2808,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8550918196994992,
+      "grad_norm": 0.4627483544714393,
+      "learning_rate": 7.072737054678004e-06,
+      "loss": 0.2735,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8554257095158597,
+      "grad_norm": 0.4649883633518454,
+      "learning_rate": 7.070085120256559e-06,
+      "loss": 0.2754,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8557595993322203,
+      "grad_norm": 0.49202727174361893,
+      "learning_rate": 7.067432482844911e-06,
+      "loss": 0.2815,
+      "step": 2563
+    },
+    {
+      "epoch": 0.856093489148581,
+      "grad_norm": 0.4900456209382406,
+      "learning_rate": 7.064779143343881e-06,
+      "loss": 0.2935,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8564273789649416,
+      "grad_norm": 0.47742930030310426,
+      "learning_rate": 7.06212510265453e-06,
+      "loss": 0.2728,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8567612687813022,
+      "grad_norm": 0.49859264857916574,
+      "learning_rate": 7.059470361678156e-06,
+      "loss": 0.2828,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8570951585976627,
+      "grad_norm": 0.5137185639742206,
+      "learning_rate": 7.056814921316295e-06,
+      "loss": 0.3017,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8574290484140233,
+      "grad_norm": 0.48005008966395485,
+      "learning_rate": 7.054158782470722e-06,
+      "loss": 0.2749,
+      "step": 2568
+    },
+    {
+      "epoch": 0.857762938230384,
+      "grad_norm": 0.5030771244593911,
+      "learning_rate": 7.051501946043446e-06,
+      "loss": 0.296,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8580968280467446,
+      "grad_norm": 0.49961680480603426,
+      "learning_rate": 7.048844412936718e-06,
+      "loss": 0.2909,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8584307178631052,
+      "grad_norm": 0.47461543152167057,
+      "learning_rate": 7.046186184053019e-06,
+      "loss": 0.2867,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8587646076794658,
+      "grad_norm": 0.5011684384224094,
+      "learning_rate": 7.043527260295073e-06,
+      "loss": 0.2764,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8590984974958263,
+      "grad_norm": 0.4811084154454694,
+      "learning_rate": 7.040867642565834e-06,
+      "loss": 0.2867,
+      "step": 2573
+    },
+    {
+      "epoch": 0.859432387312187,
+      "grad_norm": 0.514000010529705,
+      "learning_rate": 7.038207331768496e-06,
+      "loss": 0.2823,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8597662771285476,
+      "grad_norm": 0.48489740770675627,
+      "learning_rate": 7.0355463288064865e-06,
+      "loss": 0.2779,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8601001669449082,
+      "grad_norm": 0.4709545447122008,
+      "learning_rate": 7.032884634583468e-06,
+      "loss": 0.276,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8604340567612688,
+      "grad_norm": 0.4935270999966106,
+      "learning_rate": 7.030222250003338e-06,
+      "loss": 0.2955,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8607679465776293,
+      "grad_norm": 0.4605938260199138,
+      "learning_rate": 7.027559175970229e-06,
+      "loss": 0.269,
+      "step": 2578
+    },
+    {
+      "epoch": 0.86110183639399,
+      "grad_norm": 0.4808711196496712,
+      "learning_rate": 7.024895413388508e-06,
+      "loss": 0.2752,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8614357262103506,
+      "grad_norm": 0.4909735953568922,
+      "learning_rate": 7.022230963162773e-06,
+      "loss": 0.2868,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8617696160267112,
+      "grad_norm": 0.48943391327136504,
+      "learning_rate": 7.019565826197859e-06,
+      "loss": 0.2841,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8621035058430718,
+      "grad_norm": 0.4591315296536973,
+      "learning_rate": 7.016900003398832e-06,
+      "loss": 0.2774,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8624373956594323,
+      "grad_norm": 0.4568097038397946,
+      "learning_rate": 7.014233495670994e-06,
+      "loss": 0.2823,
+      "step": 2583
+    },
+    {
+      "epoch": 0.862771285475793,
+      "grad_norm": 0.4701082494338221,
+      "learning_rate": 7.011566303919872e-06,
+      "loss": 0.2899,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8631051752921536,
+      "grad_norm": 0.4901781524513855,
+      "learning_rate": 7.008898429051235e-06,
+      "loss": 0.2877,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8634390651085142,
+      "grad_norm": 0.4672953857463387,
+      "learning_rate": 7.00622987197108e-06,
+      "loss": 0.2835,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8637729549248748,
+      "grad_norm": 0.4909068517352812,
+      "learning_rate": 7.003560633585631e-06,
+      "loss": 0.2926,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8641068447412354,
+      "grad_norm": 0.47084100590653744,
+      "learning_rate": 7.000890714801352e-06,
+      "loss": 0.2864,
+      "step": 2588
+    },
+    {
+      "epoch": 0.864440734557596,
+      "grad_norm": 0.5056695468284423,
+      "learning_rate": 6.99822011652493e-06,
+      "loss": 0.3013,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8647746243739566,
+      "grad_norm": 0.4870580900300368,
+      "learning_rate": 6.995548839663289e-06,
+      "loss": 0.2874,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8651085141903172,
+      "grad_norm": 0.48283308166752054,
+      "learning_rate": 6.992876885123578e-06,
+      "loss": 0.2866,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8654424040066778,
+      "grad_norm": 0.45436164014824676,
+      "learning_rate": 6.990204253813182e-06,
+      "loss": 0.2836,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8657762938230384,
+      "grad_norm": 0.44891707355265,
+      "learning_rate": 6.987530946639712e-06,
+      "loss": 0.2753,
+      "step": 2593
+    },
+    {
+      "epoch": 0.866110183639399,
+      "grad_norm": 0.47969313756097365,
+      "learning_rate": 6.984856964511007e-06,
+      "loss": 0.2947,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8664440734557596,
+      "grad_norm": 0.4884731571543168,
+      "learning_rate": 6.982182308335142e-06,
+      "loss": 0.2899,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8667779632721202,
+      "grad_norm": 0.46950810137093063,
+      "learning_rate": 6.979506979020412e-06,
+      "loss": 0.2903,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8671118530884808,
+      "grad_norm": 0.44728015846003233,
+      "learning_rate": 6.976830977475346e-06,
+      "loss": 0.2838,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8674457429048414,
+      "grad_norm": 0.4536455806909591,
+      "learning_rate": 6.974154304608703e-06,
+      "loss": 0.2735,
+      "step": 2598
+    },
+    {
+      "epoch": 0.867779632721202,
+      "grad_norm": 0.4625484694570935,
+      "learning_rate": 6.971476961329464e-06,
+      "loss": 0.2874,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8681135225375626,
+      "grad_norm": 0.472397451005472,
+      "learning_rate": 6.968798948546842e-06,
+      "loss": 0.2816,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8684474123539232,
+      "grad_norm": 0.5584719284022011,
+      "learning_rate": 6.9661202671702765e-06,
+      "loss": 0.2723,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8687813021702838,
+      "grad_norm": 0.48559704956365285,
+      "learning_rate": 6.9634409181094355e-06,
+      "loss": 0.2972,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8691151919866444,
+      "grad_norm": 0.45459424598324455,
+      "learning_rate": 6.9607609022742085e-06,
+      "loss": 0.2808,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8694490818030051,
+      "grad_norm": 0.47118769535782634,
+      "learning_rate": 6.958080220574719e-06,
+      "loss": 0.2726,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8697829716193656,
+      "grad_norm": 0.5135018458235371,
+      "learning_rate": 6.9553988739213075e-06,
+      "loss": 0.2933,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8701168614357262,
+      "grad_norm": 0.4561098481473763,
+      "learning_rate": 6.952716863224551e-06,
+      "loss": 0.2785,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8704507512520868,
+      "grad_norm": 0.47587059438554946,
+      "learning_rate": 6.950034189395243e-06,
+      "loss": 0.2816,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8707846410684474,
+      "grad_norm": 0.4661279393994581,
+      "learning_rate": 6.947350853344408e-06,
+      "loss": 0.2733,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8711185308848081,
+      "grad_norm": 0.4696464424860318,
+      "learning_rate": 6.9446668559832895e-06,
+      "loss": 0.2822,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8714524207011686,
+      "grad_norm": 0.47166084342855963,
+      "learning_rate": 6.941982198223362e-06,
+      "loss": 0.286,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8717863105175292,
+      "grad_norm": 0.4865325548302918,
+      "learning_rate": 6.939296880976322e-06,
+      "loss": 0.2878,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8721202003338898,
+      "grad_norm": 0.46926467823829393,
+      "learning_rate": 6.936610905154088e-06,
+      "loss": 0.2863,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8724540901502504,
+      "grad_norm": 0.44092354019546404,
+      "learning_rate": 6.933924271668802e-06,
+      "loss": 0.2774,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8727879799666111,
+      "grad_norm": 0.4724123068997372,
+      "learning_rate": 6.931236981432833e-06,
+      "loss": 0.2842,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8731218697829716,
+      "grad_norm": 0.48792596354112133,
+      "learning_rate": 6.928549035358772e-06,
+      "loss": 0.2867,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8734557595993322,
+      "grad_norm": 0.47338273446056206,
+      "learning_rate": 6.925860434359427e-06,
+      "loss": 0.2696,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8737896494156928,
+      "grad_norm": 0.5131948734031979,
+      "learning_rate": 6.923171179347838e-06,
+      "loss": 0.2942,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8741235392320534,
+      "grad_norm": 0.4936424594046395,
+      "learning_rate": 6.92048127123726e-06,
+      "loss": 0.2809,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8744574290484141,
+      "grad_norm": 0.47094636157882946,
+      "learning_rate": 6.917790710941171e-06,
+      "loss": 0.2835,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8747913188647746,
+      "grad_norm": 0.4682549956051104,
+      "learning_rate": 6.915099499373273e-06,
+      "loss": 0.2734,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8751252086811352,
+      "grad_norm": 0.5023502811144523,
+      "learning_rate": 6.912407637447487e-06,
+      "loss": 0.2842,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8754590984974958,
+      "grad_norm": 0.4700360040967753,
+      "learning_rate": 6.909715126077953e-06,
+      "loss": 0.2826,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8757929883138564,
+      "grad_norm": 0.46362909164986554,
+      "learning_rate": 6.907021966179039e-06,
+      "loss": 0.2726,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8761268781302171,
+      "grad_norm": 0.4805389290217823,
+      "learning_rate": 6.904328158665323e-06,
+      "loss": 0.2757,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8764607679465777,
+      "grad_norm": 0.4970288753509855,
+      "learning_rate": 6.901633704451614e-06,
+      "loss": 0.2841,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8767946577629382,
+      "grad_norm": 0.5004179481368005,
+      "learning_rate": 6.898938604452926e-06,
+      "loss": 0.2861,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8771285475792988,
+      "grad_norm": 0.5089159904102596,
+      "learning_rate": 6.89624285958451e-06,
+      "loss": 0.2892,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8774624373956594,
+      "grad_norm": 0.47111959008499743,
+      "learning_rate": 6.89354647076182e-06,
+      "loss": 0.2822,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8777963272120201,
+      "grad_norm": 0.46620613422878593,
+      "learning_rate": 6.890849438900542e-06,
+      "loss": 0.2882,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8781302170283807,
+      "grad_norm": 0.48050747245266034,
+      "learning_rate": 6.88815176491657e-06,
+      "loss": 0.2831,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8784641068447412,
+      "grad_norm": 0.4779962633962307,
+      "learning_rate": 6.88545344972602e-06,
+      "loss": 0.2848,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8787979966611018,
+      "grad_norm": 0.4769219451990421,
+      "learning_rate": 6.882754494245229e-06,
+      "loss": 0.2905,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8791318864774624,
+      "grad_norm": 0.4915291216499879,
+      "learning_rate": 6.880054899390744e-06,
+      "loss": 0.2902,
+      "step": 2633
+    },
+    {
+      "epoch": 0.879465776293823,
+      "grad_norm": 0.4954472174313394,
+      "learning_rate": 6.8773546660793374e-06,
+      "loss": 0.2911,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8797996661101837,
+      "grad_norm": 0.46298148611319373,
+      "learning_rate": 6.87465379522799e-06,
+      "loss": 0.2724,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8801335559265442,
+      "grad_norm": 0.4419334792493168,
+      "learning_rate": 6.87195228775391e-06,
+      "loss": 0.2763,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8804674457429048,
+      "grad_norm": 0.4886221392469825,
+      "learning_rate": 6.8692501445745094e-06,
+      "loss": 0.2734,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8808013355592654,
+      "grad_norm": 0.5359869367732673,
+      "learning_rate": 6.866547366607427e-06,
+      "loss": 0.3054,
+      "step": 2638
+    },
+    {
+      "epoch": 0.881135225375626,
+      "grad_norm": 0.491276242874801,
+      "learning_rate": 6.863843954770507e-06,
+      "loss": 0.2818,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8814691151919867,
+      "grad_norm": 0.46606499658008294,
+      "learning_rate": 6.861139909981818e-06,
+      "loss": 0.282,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8818030050083473,
+      "grad_norm": 0.4657454099023511,
+      "learning_rate": 6.858435233159638e-06,
+      "loss": 0.2944,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8821368948247078,
+      "grad_norm": 0.46998393511687225,
+      "learning_rate": 6.855729925222462e-06,
+      "loss": 0.2854,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8824707846410684,
+      "grad_norm": 0.5166724375708797,
+      "learning_rate": 6.853023987088997e-06,
+      "loss": 0.2852,
+      "step": 2643
+    },
+    {
+      "epoch": 0.882804674457429,
+      "grad_norm": 0.49780676223604287,
+      "learning_rate": 6.850317419678168e-06,
+      "loss": 0.2834,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8831385642737897,
+      "grad_norm": 0.4984120857584309,
+      "learning_rate": 6.847610223909109e-06,
+      "loss": 0.2916,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8834724540901503,
+      "grad_norm": 0.4796832389818675,
+      "learning_rate": 6.844902400701172e-06,
+      "loss": 0.2947,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8838063439065108,
+      "grad_norm": 0.48731437785971965,
+      "learning_rate": 6.842193950973919e-06,
+      "loss": 0.2867,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8841402337228714,
+      "grad_norm": 0.4594274203007557,
+      "learning_rate": 6.839484875647124e-06,
+      "loss": 0.28,
+      "step": 2648
+    },
+    {
+      "epoch": 0.884474123539232,
+      "grad_norm": 0.5441715164390831,
+      "learning_rate": 6.836775175640775e-06,
+      "loss": 0.2893,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8848080133555927,
+      "grad_norm": 0.49044152796355023,
+      "learning_rate": 6.834064851875075e-06,
+      "loss": 0.2921,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8851419031719533,
+      "grad_norm": 0.4766896550866435,
+      "learning_rate": 6.831353905270433e-06,
+      "loss": 0.2891,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8854757929883138,
+      "grad_norm": 0.4918257054340345,
+      "learning_rate": 6.828642336747474e-06,
+      "loss": 0.2963,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8858096828046744,
+      "grad_norm": 0.4914995495475317,
+      "learning_rate": 6.8259301472270314e-06,
+      "loss": 0.279,
+      "step": 2653
+    },
+    {
+      "epoch": 0.886143572621035,
+      "grad_norm": 0.4938999271205023,
+      "learning_rate": 6.823217337630153e-06,
+      "loss": 0.2899,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8864774624373957,
+      "grad_norm": 0.4625580706888065,
+      "learning_rate": 6.820503908878093e-06,
+      "loss": 0.2827,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8868113522537563,
+      "grad_norm": 0.4727924925615349,
+      "learning_rate": 6.817789861892318e-06,
+      "loss": 0.2846,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8871452420701169,
+      "grad_norm": 0.49111007993604033,
+      "learning_rate": 6.815075197594505e-06,
+      "loss": 0.2807,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8874791318864774,
+      "grad_norm": 0.5023042258323622,
+      "learning_rate": 6.812359916906539e-06,
+      "loss": 0.2848,
+      "step": 2658
+    },
+    {
+      "epoch": 0.887813021702838,
+      "grad_norm": 0.5073281550710147,
+      "learning_rate": 6.809644020750517e-06,
+      "loss": 0.2881,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8881469115191987,
+      "grad_norm": 0.4751906535051513,
+      "learning_rate": 6.806927510048738e-06,
+      "loss": 0.2928,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8884808013355593,
+      "grad_norm": 0.49012023414458106,
+      "learning_rate": 6.804210385723723e-06,
+      "loss": 0.2875,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8888146911519199,
+      "grad_norm": 0.5017067831602734,
+      "learning_rate": 6.801492648698188e-06,
+      "loss": 0.28,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8891485809682804,
+      "grad_norm": 0.5340614745684702,
+      "learning_rate": 6.798774299895066e-06,
+      "loss": 0.2891,
+      "step": 2663
+    },
+    {
+      "epoch": 0.889482470784641,
+      "grad_norm": 0.487082875312774,
+      "learning_rate": 6.79605534023749e-06,
+      "loss": 0.2778,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8898163606010017,
+      "grad_norm": 0.5136469487040188,
+      "learning_rate": 6.793335770648806e-06,
+      "loss": 0.28,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8901502504173623,
+      "grad_norm": 0.49640535725688534,
+      "learning_rate": 6.790615592052568e-06,
+      "loss": 0.2875,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8904841402337229,
+      "grad_norm": 0.47961689822559017,
+      "learning_rate": 6.787894805372532e-06,
+      "loss": 0.2779,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8908180300500834,
+      "grad_norm": 0.502222640977425,
+      "learning_rate": 6.785173411532666e-06,
+      "loss": 0.2871,
+      "step": 2668
+    },
+    {
+      "epoch": 0.891151919866444,
+      "grad_norm": 0.5167699800804186,
+      "learning_rate": 6.782451411457137e-06,
+      "loss": 0.2799,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8914858096828047,
+      "grad_norm": 0.4795207157629916,
+      "learning_rate": 6.779728806070327e-06,
+      "loss": 0.2917,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8918196994991653,
+      "grad_norm": 0.4943890861750615,
+      "learning_rate": 6.777005596296815e-06,
+      "loss": 0.2733,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8921535893155259,
+      "grad_norm": 0.4948029379495857,
+      "learning_rate": 6.774281783061391e-06,
+      "loss": 0.2923,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8924874791318865,
+      "grad_norm": 0.47003536144584385,
+      "learning_rate": 6.7715573672890455e-06,
+      "loss": 0.2852,
+      "step": 2673
+    },
+    {
+      "epoch": 0.892821368948247,
+      "grad_norm": 0.4926436547668656,
+      "learning_rate": 6.768832349904979e-06,
+      "loss": 0.2958,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8931552587646077,
+      "grad_norm": 0.47533692887286233,
+      "learning_rate": 6.766106731834592e-06,
+      "loss": 0.2822,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8934891485809683,
+      "grad_norm": 0.490790540961599,
+      "learning_rate": 6.76338051400349e-06,
+      "loss": 0.2937,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8938230383973289,
+      "grad_norm": 0.5215371561829931,
+      "learning_rate": 6.760653697337483e-06,
+      "loss": 0.2951,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8941569282136895,
+      "grad_norm": 0.4706995289152973,
+      "learning_rate": 6.757926282762583e-06,
+      "loss": 0.2811,
+      "step": 2678
+    },
+    {
+      "epoch": 0.89449081803005,
+      "grad_norm": 0.4726870515885244,
+      "learning_rate": 6.755198271205008e-06,
+      "loss": 0.2977,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8948247078464107,
+      "grad_norm": 0.47557600677039136,
+      "learning_rate": 6.752469663591174e-06,
+      "loss": 0.2811,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8951585976627713,
+      "grad_norm": 0.46277968940936104,
+      "learning_rate": 6.7497404608477035e-06,
+      "loss": 0.279,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8954924874791319,
+      "grad_norm": 0.468007706095983,
+      "learning_rate": 6.747010663901419e-06,
+      "loss": 0.2743,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8958263772954925,
+      "grad_norm": 0.48007983288975137,
+      "learning_rate": 6.744280273679345e-06,
+      "loss": 0.281,
+      "step": 2683
+    },
+    {
+      "epoch": 0.896160267111853,
+      "grad_norm": 0.48408732640809116,
+      "learning_rate": 6.74154929110871e-06,
+      "loss": 0.2956,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8964941569282137,
+      "grad_norm": 0.5021970773732836,
+      "learning_rate": 6.738817717116939e-06,
+      "loss": 0.2882,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8968280467445743,
+      "grad_norm": 0.4784985891824096,
+      "learning_rate": 6.736085552631663e-06,
+      "loss": 0.286,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8971619365609349,
+      "grad_norm": 0.46764803291157087,
+      "learning_rate": 6.733352798580708e-06,
+      "loss": 0.2851,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8974958263772955,
+      "grad_norm": 0.4367639928894207,
+      "learning_rate": 6.730619455892106e-06,
+      "loss": 0.2698,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8978297161936561,
+      "grad_norm": 0.4705117060959479,
+      "learning_rate": 6.727885525494083e-06,
+      "loss": 0.2904,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8981636060100167,
+      "grad_norm": 0.4770427768973616,
+      "learning_rate": 6.725151008315071e-06,
+      "loss": 0.2758,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8984974958263773,
+      "grad_norm": 0.4738981037969852,
+      "learning_rate": 6.722415905283696e-06,
+      "loss": 0.2841,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8988313856427379,
+      "grad_norm": 0.4828561505354717,
+      "learning_rate": 6.719680217328784e-06,
+      "loss": 0.2929,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8991652754590985,
+      "grad_norm": 0.47316669453084004,
+      "learning_rate": 6.716943945379364e-06,
+      "loss": 0.2992,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8994991652754591,
+      "grad_norm": 0.46467298423418174,
+      "learning_rate": 6.714207090364654e-06,
+      "loss": 0.2797,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8998330550918197,
+      "grad_norm": 0.46764408156153586,
+      "learning_rate": 6.7114696532140825e-06,
+      "loss": 0.2891,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9001669449081803,
+      "grad_norm": 0.4690293170670278,
+      "learning_rate": 6.7087316348572626e-06,
+      "loss": 0.2865,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9005008347245409,
+      "grad_norm": 0.4569321626678759,
+      "learning_rate": 6.705993036224017e-06,
+      "loss": 0.2732,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9008347245409015,
+      "grad_norm": 0.4854655794806803,
+      "learning_rate": 6.703253858244355e-06,
+      "loss": 0.2871,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9011686143572621,
+      "grad_norm": 0.47614297049904813,
+      "learning_rate": 6.70051410184849e-06,
+      "loss": 0.2963,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9015025041736227,
+      "grad_norm": 0.44761475362698383,
+      "learning_rate": 6.697773767966829e-06,
+      "loss": 0.2861,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9018363939899833,
+      "grad_norm": 0.44987982591314096,
+      "learning_rate": 6.695032857529973e-06,
+      "loss": 0.2769,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9021702838063439,
+      "grad_norm": 0.469646131694694,
+      "learning_rate": 6.692291371468724e-06,
+      "loss": 0.2802,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9025041736227045,
+      "grad_norm": 0.4710876846203227,
+      "learning_rate": 6.689549310714073e-06,
+      "loss": 0.2825,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9028380634390651,
+      "grad_norm": 0.48302982496064645,
+      "learning_rate": 6.686806676197214e-06,
+      "loss": 0.2794,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9031719532554258,
+      "grad_norm": 0.4528382900103332,
+      "learning_rate": 6.684063468849528e-06,
+      "loss": 0.2812,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9035058430717863,
+      "grad_norm": 0.4539495721867444,
+      "learning_rate": 6.681319689602596e-06,
+      "loss": 0.2867,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9038397328881469,
+      "grad_norm": 0.4670177978150963,
+      "learning_rate": 6.678575339388189e-06,
+      "loss": 0.2922,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9041736227045075,
+      "grad_norm": 0.47467566740896067,
+      "learning_rate": 6.6758304191382764e-06,
+      "loss": 0.2848,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9045075125208681,
+      "grad_norm": 0.46782059851986046,
+      "learning_rate": 6.673084929785019e-06,
+      "loss": 0.2844,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9048414023372288,
+      "grad_norm": 0.4809485251163933,
+      "learning_rate": 6.670338872260768e-06,
+      "loss": 0.2952,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9051752921535893,
+      "grad_norm": 0.4598207380236529,
+      "learning_rate": 6.667592247498074e-06,
+      "loss": 0.2779,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9055091819699499,
+      "grad_norm": 0.48594598195785754,
+      "learning_rate": 6.664845056429673e-06,
+      "loss": 0.2835,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9058430717863105,
+      "grad_norm": 0.46232974229797147,
+      "learning_rate": 6.662097299988502e-06,
+      "loss": 0.2869,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9061769616026711,
+      "grad_norm": 0.4521368411681244,
+      "learning_rate": 6.659348979107679e-06,
+      "loss": 0.283,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9065108514190318,
+      "grad_norm": 0.45184349148143177,
+      "learning_rate": 6.656600094720522e-06,
+      "loss": 0.2758,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9068447412353923,
+      "grad_norm": 0.4616252256275939,
+      "learning_rate": 6.653850647760538e-06,
+      "loss": 0.2778,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9071786310517529,
+      "grad_norm": 0.44874180665874897,
+      "learning_rate": 6.651100639161425e-06,
+      "loss": 0.2821,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9075125208681135,
+      "grad_norm": 0.4735867469418201,
+      "learning_rate": 6.6483500698570734e-06,
+      "loss": 0.2852,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9078464106844741,
+      "grad_norm": 0.4639973326533366,
+      "learning_rate": 6.645598940781559e-06,
+      "loss": 0.284,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9081803005008348,
+      "grad_norm": 0.4876243423661334,
+      "learning_rate": 6.642847252869153e-06,
+      "loss": 0.2724,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9085141903171954,
+      "grad_norm": 0.47095282678480793,
+      "learning_rate": 6.6400950070543155e-06,
+      "loss": 0.2874,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9088480801335559,
+      "grad_norm": 0.5146540823102966,
+      "learning_rate": 6.637342204271693e-06,
+      "loss": 0.2875,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9091819699499165,
+      "grad_norm": 0.5050330236323558,
+      "learning_rate": 6.634588845456122e-06,
+      "loss": 0.283,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9095158597662771,
+      "grad_norm": 0.45145202985929805,
+      "learning_rate": 6.631834931542631e-06,
+      "loss": 0.2859,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9098497495826378,
+      "grad_norm": 0.49127676215174354,
+      "learning_rate": 6.629080463466435e-06,
+      "loss": 0.2983,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9101836393989984,
+      "grad_norm": 0.4491633650106657,
+      "learning_rate": 6.626325442162936e-06,
+      "loss": 0.2741,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9105175292153589,
+      "grad_norm": 0.4475485115335679,
+      "learning_rate": 6.623569868567727e-06,
+      "loss": 0.2879,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9108514190317195,
+      "grad_norm": 0.4827658235814058,
+      "learning_rate": 6.620813743616584e-06,
+      "loss": 0.2878,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9111853088480801,
+      "grad_norm": 0.4995003127842459,
+      "learning_rate": 6.6180570682454734e-06,
+      "loss": 0.2852,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9115191986644408,
+      "grad_norm": 0.48510442140830107,
+      "learning_rate": 6.61529984339055e-06,
+      "loss": 0.2727,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9118530884808014,
+      "grad_norm": 0.45578492267234677,
+      "learning_rate": 6.61254206998815e-06,
+      "loss": 0.2798,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9121869782971619,
+      "grad_norm": 0.4680844186801238,
+      "learning_rate": 6.609783748974802e-06,
+      "loss": 0.2843,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9125208681135225,
+      "grad_norm": 0.4842803602397625,
+      "learning_rate": 6.607024881287213e-06,
+      "loss": 0.2962,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9128547579298831,
+      "grad_norm": 0.5207789863761901,
+      "learning_rate": 6.604265467862286e-06,
+      "loss": 0.2978,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9131886477462438,
+      "grad_norm": 0.47067548976401286,
+      "learning_rate": 6.6015055096371e-06,
+      "loss": 0.286,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9135225375626044,
+      "grad_norm": 0.4473385848147624,
+      "learning_rate": 6.5987450075489234e-06,
+      "loss": 0.2784,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9138564273789649,
+      "grad_norm": 0.4898520170046775,
+      "learning_rate": 6.595983962535209e-06,
+      "loss": 0.2737,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9141903171953255,
+      "grad_norm": 0.49117561859566045,
+      "learning_rate": 6.59322237553359e-06,
+      "loss": 0.2818,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9145242070116861,
+      "grad_norm": 0.46863831196368594,
+      "learning_rate": 6.5904602474818924e-06,
+      "loss": 0.2733,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9148580968280468,
+      "grad_norm": 0.4850144642644832,
+      "learning_rate": 6.587697579318118e-06,
+      "loss": 0.2819,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9151919866444074,
+      "grad_norm": 0.46513328843010404,
+      "learning_rate": 6.584934371980452e-06,
+      "loss": 0.2733,
+      "step": 2741
+    },
+    {
+      "epoch": 0.915525876460768,
+      "grad_norm": 0.4756530598048289,
+      "learning_rate": 6.582170626407271e-06,
+      "loss": 0.2853,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9158597662771285,
+      "grad_norm": 0.4463718902933659,
+      "learning_rate": 6.579406343537125e-06,
+      "loss": 0.2764,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9161936560934891,
+      "grad_norm": 0.509236751434957,
+      "learning_rate": 6.57664152430875e-06,
+      "loss": 0.2903,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9165275459098498,
+      "grad_norm": 0.46518032400336967,
+      "learning_rate": 6.573876169661065e-06,
+      "loss": 0.2743,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9168614357262104,
+      "grad_norm": 0.4572607393549725,
+      "learning_rate": 6.571110280533171e-06,
+      "loss": 0.2888,
+      "step": 2746
+    },
+    {
+      "epoch": 0.917195325542571,
+      "grad_norm": 0.5036855063543908,
+      "learning_rate": 6.56834385786435e-06,
+      "loss": 0.2858,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9175292153589315,
+      "grad_norm": 0.4718677812074385,
+      "learning_rate": 6.565576902594063e-06,
+      "loss": 0.2811,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9178631051752921,
+      "grad_norm": 0.4623942750329796,
+      "learning_rate": 6.562809415661954e-06,
+      "loss": 0.2901,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9181969949916527,
+      "grad_norm": 0.47399213591192507,
+      "learning_rate": 6.560041398007847e-06,
+      "loss": 0.2888,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9185308848080134,
+      "grad_norm": 0.5330149969361173,
+      "learning_rate": 6.5572728505717475e-06,
+      "loss": 0.2795,
+      "step": 2751
+    },
+    {
+      "epoch": 0.918864774624374,
+      "grad_norm": 0.5043446562355761,
+      "learning_rate": 6.554503774293841e-06,
+      "loss": 0.2827,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9191986644407345,
+      "grad_norm": 0.47812566524801814,
+      "learning_rate": 6.55173417011449e-06,
+      "loss": 0.2822,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9195325542570951,
+      "grad_norm": 0.45663696755589595,
+      "learning_rate": 6.548964038974236e-06,
+      "loss": 0.2828,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9198664440734557,
+      "grad_norm": 0.44892421149992107,
+      "learning_rate": 6.546193381813806e-06,
+      "loss": 0.2796,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9202003338898164,
+      "grad_norm": 0.49168088512887964,
+      "learning_rate": 6.543422199574096e-06,
+      "loss": 0.2905,
+      "step": 2756
+    },
+    {
+      "epoch": 0.920534223706177,
+      "grad_norm": 0.5146290210596555,
+      "learning_rate": 6.540650493196187e-06,
+      "loss": 0.295,
+      "step": 2757
+    },
+    {
+      "epoch": 0.9208681135225376,
+      "grad_norm": 0.4778074163876107,
+      "learning_rate": 6.537878263621335e-06,
+      "loss": 0.2746,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9212020033388981,
+      "grad_norm": 0.44852367046810726,
+      "learning_rate": 6.535105511790979e-06,
+      "loss": 0.2834,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9215358931552587,
+      "grad_norm": 0.43815392887249655,
+      "learning_rate": 6.532332238646726e-06,
+      "loss": 0.2791,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9218697829716194,
+      "grad_norm": 0.4604734991774878,
+      "learning_rate": 6.5295584451303694e-06,
+      "loss": 0.2739,
+      "step": 2761
+    },
+    {
+      "epoch": 0.92220367278798,
+      "grad_norm": 0.5011903703972732,
+      "learning_rate": 6.526784132183872e-06,
+      "loss": 0.2789,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9225375626043406,
+      "grad_norm": 0.48769829861064534,
+      "learning_rate": 6.524009300749377e-06,
+      "loss": 0.2862,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9228714524207011,
+      "grad_norm": 0.5029750704921034,
+      "learning_rate": 6.521233951769204e-06,
+      "loss": 0.2782,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9232053422370617,
+      "grad_norm": 0.5253251142529386,
+      "learning_rate": 6.518458086185846e-06,
+      "loss": 0.2835,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9235392320534224,
+      "grad_norm": 0.45846729953047993,
+      "learning_rate": 6.515681704941974e-06,
+      "loss": 0.2768,
+      "step": 2766
+    },
+    {
+      "epoch": 0.923873121869783,
+      "grad_norm": 0.46967465750392495,
+      "learning_rate": 6.512904808980429e-06,
+      "loss": 0.295,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9242070116861436,
+      "grad_norm": 0.47355941924857314,
+      "learning_rate": 6.510127399244235e-06,
+      "loss": 0.2911,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9245409015025041,
+      "grad_norm": 0.46747463057812094,
+      "learning_rate": 6.5073494766765835e-06,
+      "loss": 0.2905,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9248747913188647,
+      "grad_norm": 0.47294389200252956,
+      "learning_rate": 6.5045710422208445e-06,
+      "loss": 0.2766,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9252086811352254,
+      "grad_norm": 0.45226078665588415,
+      "learning_rate": 6.5017920968205575e-06,
+      "loss": 0.2618,
+      "step": 2771
+    },
+    {
+      "epoch": 0.925542570951586,
+      "grad_norm": 0.4714298067979066,
+      "learning_rate": 6.499012641419438e-06,
+      "loss": 0.2976,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9258764607679466,
+      "grad_norm": 0.44892422071480925,
+      "learning_rate": 6.496232676961379e-06,
+      "loss": 0.2708,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9262103505843072,
+      "grad_norm": 0.47664742783008623,
+      "learning_rate": 6.493452204390438e-06,
+      "loss": 0.2916,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9265442404006677,
+      "grad_norm": 0.47110198016708166,
+      "learning_rate": 6.4906712246508484e-06,
+      "loss": 0.2836,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9268781302170284,
+      "grad_norm": 0.47090347806101984,
+      "learning_rate": 6.487889738687019e-06,
+      "loss": 0.2853,
+      "step": 2776
+    },
+    {
+      "epoch": 0.927212020033389,
+      "grad_norm": 0.4875463686270971,
+      "learning_rate": 6.485107747443529e-06,
+      "loss": 0.2796,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9275459098497496,
+      "grad_norm": 0.4651806132206352,
+      "learning_rate": 6.482325251865125e-06,
+      "loss": 0.2794,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9278797996661102,
+      "grad_norm": 0.47449335188923775,
+      "learning_rate": 6.479542252896729e-06,
+      "loss": 0.2861,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9282136894824707,
+      "grad_norm": 0.4913821132425136,
+      "learning_rate": 6.476758751483435e-06,
+      "loss": 0.2821,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9285475792988314,
+      "grad_norm": 0.462747224933638,
+      "learning_rate": 6.473974748570504e-06,
+      "loss": 0.2813,
+      "step": 2781
+    },
+    {
+      "epoch": 0.928881469115192,
+      "grad_norm": 0.4379389586607412,
+      "learning_rate": 6.47119024510337e-06,
+      "loss": 0.285,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9292153589315526,
+      "grad_norm": 0.4535404568747794,
+      "learning_rate": 6.4684052420276335e-06,
+      "loss": 0.272,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9295492487479132,
+      "grad_norm": 0.46225203332369796,
+      "learning_rate": 6.465619740289071e-06,
+      "loss": 0.2799,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9298831385642737,
+      "grad_norm": 0.46094752065359396,
+      "learning_rate": 6.46283374083362e-06,
+      "loss": 0.2813,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9302170283806344,
+      "grad_norm": 0.5007157681310827,
+      "learning_rate": 6.460047244607397e-06,
+      "loss": 0.2728,
+      "step": 2786
+    },
+    {
+      "epoch": 0.930550918196995,
+      "grad_norm": 0.4777259650709117,
+      "learning_rate": 6.457260252556677e-06,
+      "loss": 0.2882,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9308848080133556,
+      "grad_norm": 0.4572050885040431,
+      "learning_rate": 6.454472765627911e-06,
+      "loss": 0.2808,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9312186978297162,
+      "grad_norm": 0.4675759219747415,
+      "learning_rate": 6.451684784767714e-06,
+      "loss": 0.2792,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9315525876460768,
+      "grad_norm": 0.4858206371801601,
+      "learning_rate": 6.4488963109228675e-06,
+      "loss": 0.2877,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9318864774624374,
+      "grad_norm": 0.4837924528240103,
+      "learning_rate": 6.446107345040328e-06,
+      "loss": 0.2794,
+      "step": 2791
+    },
+    {
+      "epoch": 0.932220367278798,
+      "grad_norm": 0.47960909383782296,
+      "learning_rate": 6.443317888067207e-06,
+      "loss": 0.293,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9325542570951586,
+      "grad_norm": 0.4683759282314836,
+      "learning_rate": 6.4405279409507984e-06,
+      "loss": 0.2815,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9328881469115192,
+      "grad_norm": 0.4404221379297429,
+      "learning_rate": 6.437737504638546e-06,
+      "loss": 0.272,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9332220367278798,
+      "grad_norm": 0.47098894560196924,
+      "learning_rate": 6.434946580078072e-06,
+      "loss": 0.2888,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9335559265442404,
+      "grad_norm": 0.4706595271935374,
+      "learning_rate": 6.432155168217158e-06,
+      "loss": 0.2868,
+      "step": 2796
+    },
+    {
+      "epoch": 0.933889816360601,
+      "grad_norm": 0.46578891492534436,
+      "learning_rate": 6.429363270003755e-06,
+      "loss": 0.2769,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9342237061769616,
+      "grad_norm": 0.47430288388020525,
+      "learning_rate": 6.426570886385974e-06,
+      "loss": 0.2899,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9345575959933222,
+      "grad_norm": 0.4688417620695341,
+      "learning_rate": 6.423778018312097e-06,
+      "loss": 0.2797,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9348914858096828,
+      "grad_norm": 0.4768670057680106,
+      "learning_rate": 6.420984666730565e-06,
+      "loss": 0.2732,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9352253756260434,
+      "grad_norm": 0.45068478020800085,
+      "learning_rate": 6.4181908325899885e-06,
+      "loss": 0.2937,
+      "step": 2801
+    },
+    {
+      "epoch": 0.935559265442404,
+      "grad_norm": 0.4886764864056523,
+      "learning_rate": 6.415396516839139e-06,
+      "loss": 0.2782,
+      "step": 2802
+    },
+    {
+      "epoch": 0.9358931552587646,
+      "grad_norm": 0.45896242713780383,
+      "learning_rate": 6.412601720426948e-06,
+      "loss": 0.2807,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9362270450751252,
+      "grad_norm": 0.4889214484691528,
+      "learning_rate": 6.409806444302519e-06,
+      "loss": 0.2876,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9365609348914858,
+      "grad_norm": 0.47712546382133414,
+      "learning_rate": 6.407010689415109e-06,
+      "loss": 0.2862,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9368948247078465,
+      "grad_norm": 0.49113773947154626,
+      "learning_rate": 6.404214456714143e-06,
+      "loss": 0.2836,
+      "step": 2806
+    },
+    {
+      "epoch": 0.937228714524207,
+      "grad_norm": 0.48125116788469374,
+      "learning_rate": 6.40141774714921e-06,
+      "loss": 0.2974,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9375626043405676,
+      "grad_norm": 0.44310509791277947,
+      "learning_rate": 6.3986205616700534e-06,
+      "loss": 0.2715,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9378964941569282,
+      "grad_norm": 0.4791256820303822,
+      "learning_rate": 6.395822901226587e-06,
+      "loss": 0.2831,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9382303839732888,
+      "grad_norm": 0.4748729080786393,
+      "learning_rate": 6.39302476676888e-06,
+      "loss": 0.2823,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9385642737896495,
+      "grad_norm": 0.46438686956055103,
+      "learning_rate": 6.390226159247165e-06,
+      "loss": 0.2778,
+      "step": 2811
+    },
+    {
+      "epoch": 0.93889816360601,
+      "grad_norm": 0.46030159105051766,
+      "learning_rate": 6.3874270796118345e-06,
+      "loss": 0.2815,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9392320534223706,
+      "grad_norm": 0.4626659052926879,
+      "learning_rate": 6.384627528813439e-06,
+      "loss": 0.2781,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9395659432387312,
+      "grad_norm": 0.4912881486054533,
+      "learning_rate": 6.381827507802696e-06,
+      "loss": 0.2869,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9398998330550918,
+      "grad_norm": 0.48495836498293743,
+      "learning_rate": 6.379027017530476e-06,
+      "loss": 0.2782,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9402337228714525,
+      "grad_norm": 0.4694858989462164,
+      "learning_rate": 6.376226058947811e-06,
+      "loss": 0.2835,
+      "step": 2816
+    },
+    {
+      "epoch": 0.940567612687813,
+      "grad_norm": 0.5183997969361961,
+      "learning_rate": 6.373424633005889e-06,
+      "loss": 0.3021,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9409015025041736,
+      "grad_norm": 0.4693642012877035,
+      "learning_rate": 6.370622740656066e-06,
+      "loss": 0.2943,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9412353923205342,
+      "grad_norm": 0.49561529770319357,
+      "learning_rate": 6.367820382849847e-06,
+      "loss": 0.2862,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9415692821368948,
+      "grad_norm": 0.5054481657471496,
+      "learning_rate": 6.365017560538896e-06,
+      "loss": 0.2927,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9419031719532555,
+      "grad_norm": 0.44411918505205317,
+      "learning_rate": 6.362214274675042e-06,
+      "loss": 0.2797,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9422370617696161,
+      "grad_norm": 0.46359295940469497,
+      "learning_rate": 6.359410526210259e-06,
+      "loss": 0.2983,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9425709515859766,
+      "grad_norm": 0.49233245636216105,
+      "learning_rate": 6.356606316096694e-06,
+      "loss": 0.2807,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9429048414023372,
+      "grad_norm": 0.49064494673368,
+      "learning_rate": 6.353801645286636e-06,
+      "loss": 0.2832,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9432387312186978,
+      "grad_norm": 0.47332422616656855,
+      "learning_rate": 6.35099651473254e-06,
+      "loss": 0.2907,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9435726210350585,
+      "grad_norm": 0.46508114096398245,
+      "learning_rate": 6.348190925387012e-06,
+      "loss": 0.2762,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9439065108514191,
+      "grad_norm": 0.4696591395286524,
+      "learning_rate": 6.345384878202817e-06,
+      "loss": 0.2952,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9442404006677796,
+      "grad_norm": 0.46747956176438554,
+      "learning_rate": 6.342578374132874e-06,
+      "loss": 0.274,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9445742904841402,
+      "grad_norm": 0.46513296218662153,
+      "learning_rate": 6.339771414130256e-06,
+      "loss": 0.2782,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9449081803005008,
+      "grad_norm": 0.46775898789600995,
+      "learning_rate": 6.336963999148193e-06,
+      "loss": 0.284,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9452420701168615,
+      "grad_norm": 0.46825564538196335,
+      "learning_rate": 6.334156130140068e-06,
+      "loss": 0.2762,
+      "step": 2831
+    },
+    {
+      "epoch": 0.9455759599332221,
+      "grad_norm": 0.48344939614057,
+      "learning_rate": 6.331347808059419e-06,
+      "loss": 0.2865,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9459098497495826,
+      "grad_norm": 0.4511690084703443,
+      "learning_rate": 6.328539033859939e-06,
+      "loss": 0.2587,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9462437395659432,
+      "grad_norm": 0.47045764933185225,
+      "learning_rate": 6.325729808495471e-06,
+      "loss": 0.2864,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9465776293823038,
+      "grad_norm": 0.4625001080291528,
+      "learning_rate": 6.3229201329200165e-06,
+      "loss": 0.2866,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9469115191986645,
+      "grad_norm": 0.4707234681044852,
+      "learning_rate": 6.3201100080877235e-06,
+      "loss": 0.2728,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9472454090150251,
+      "grad_norm": 0.49736066757161806,
+      "learning_rate": 6.3172994349528984e-06,
+      "loss": 0.2836,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9475792988313857,
+      "grad_norm": 0.4716403236971485,
+      "learning_rate": 6.314488414469995e-06,
+      "loss": 0.2895,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9479131886477462,
+      "grad_norm": 0.43914252684071553,
+      "learning_rate": 6.311676947593621e-06,
+      "loss": 0.2688,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9482470784641068,
+      "grad_norm": 0.4656750207952864,
+      "learning_rate": 6.308865035278539e-06,
+      "loss": 0.2799,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9485809682804675,
+      "grad_norm": 0.4555598846755603,
+      "learning_rate": 6.30605267847966e-06,
+      "loss": 0.2717,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9489148580968281,
+      "grad_norm": 0.48402090696000877,
+      "learning_rate": 6.303239878152041e-06,
+      "loss": 0.2817,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9492487479131887,
+      "grad_norm": 0.4691740643723942,
+      "learning_rate": 6.300426635250899e-06,
+      "loss": 0.2808,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9495826377295492,
+      "grad_norm": 0.48590736418734515,
+      "learning_rate": 6.297612950731596e-06,
+      "loss": 0.2786,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9499165275459098,
+      "grad_norm": 0.4808552985370258,
+      "learning_rate": 6.294798825549643e-06,
+      "loss": 0.281,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9502504173622704,
+      "grad_norm": 0.4632798369611852,
+      "learning_rate": 6.291984260660705e-06,
+      "loss": 0.2786,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9505843071786311,
+      "grad_norm": 0.4673289057441812,
+      "learning_rate": 6.28916925702059e-06,
+      "loss": 0.2877,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9509181969949917,
+      "grad_norm": 0.512982601817563,
+      "learning_rate": 6.286353815585262e-06,
+      "loss": 0.2874,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9512520868113522,
+      "grad_norm": 0.4741428510294219,
+      "learning_rate": 6.283537937310829e-06,
+      "loss": 0.2749,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9515859766277128,
+      "grad_norm": 0.4658972602393785,
+      "learning_rate": 6.280721623153549e-06,
+      "loss": 0.2899,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9519198664440734,
+      "grad_norm": 0.44282011902758206,
+      "learning_rate": 6.277904874069827e-06,
+      "loss": 0.2632,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9522537562604341,
+      "grad_norm": 0.45368478196188605,
+      "learning_rate": 6.275087691016217e-06,
+      "loss": 0.2726,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9525876460767947,
+      "grad_norm": 0.5067105474818324,
+      "learning_rate": 6.272270074949421e-06,
+      "loss": 0.2782,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9529215358931553,
+      "grad_norm": 0.5195510071042065,
+      "learning_rate": 6.269452026826285e-06,
+      "loss": 0.2927,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9532554257095158,
+      "grad_norm": 0.4757984069558446,
+      "learning_rate": 6.266633547603804e-06,
+      "loss": 0.2879,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9535893155258764,
+      "grad_norm": 0.47541432314317833,
+      "learning_rate": 6.263814638239118e-06,
+      "loss": 0.2848,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9539232053422371,
+      "grad_norm": 0.4835948819603942,
+      "learning_rate": 6.2609952996895185e-06,
+      "loss": 0.2892,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9542570951585977,
+      "grad_norm": 0.45882508794677584,
+      "learning_rate": 6.258175532912432e-06,
+      "loss": 0.276,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9545909849749583,
+      "grad_norm": 0.4702641998793967,
+      "learning_rate": 6.255355338865442e-06,
+      "loss": 0.2849,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9549248747913188,
+      "grad_norm": 0.4565534196055433,
+      "learning_rate": 6.252534718506268e-06,
+      "loss": 0.2775,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9552587646076794,
+      "grad_norm": 0.5060784040108589,
+      "learning_rate": 6.24971367279278e-06,
+      "loss": 0.281,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9555926544240401,
+      "grad_norm": 0.464065079202329,
+      "learning_rate": 6.24689220268299e-06,
+      "loss": 0.2837,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9559265442404007,
+      "grad_norm": 0.48176329891549513,
+      "learning_rate": 6.244070309135055e-06,
+      "loss": 0.278,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9562604340567613,
+      "grad_norm": 0.4812915467292244,
+      "learning_rate": 6.2412479931072754e-06,
+      "loss": 0.3042,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9565943238731218,
+      "grad_norm": 0.4611507821934498,
+      "learning_rate": 6.238425255558095e-06,
+      "loss": 0.2774,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9569282136894824,
+      "grad_norm": 0.48889832508698305,
+      "learning_rate": 6.235602097446102e-06,
+      "loss": 0.2757,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9572621035058431,
+      "grad_norm": 0.4579815377199059,
+      "learning_rate": 6.232778519730024e-06,
+      "loss": 0.302,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9575959933222037,
+      "grad_norm": 0.4386076360654089,
+      "learning_rate": 6.2299545233687346e-06,
+      "loss": 0.2698,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9579298831385643,
+      "grad_norm": 0.49036080040081254,
+      "learning_rate": 6.227130109321249e-06,
+      "loss": 0.2892,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9582637729549248,
+      "grad_norm": 0.4660044853324248,
+      "learning_rate": 6.224305278546723e-06,
+      "loss": 0.2764,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9585976627712854,
+      "grad_norm": 0.46428605811670526,
+      "learning_rate": 6.221480032004455e-06,
+      "loss": 0.2843,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9589315525876461,
+      "grad_norm": 0.46826279917123664,
+      "learning_rate": 6.218654370653881e-06,
+      "loss": 0.2875,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9592654424040067,
+      "grad_norm": 0.4746687406164823,
+      "learning_rate": 6.215828295454585e-06,
+      "loss": 0.2725,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9595993322203673,
+      "grad_norm": 0.4716017468951336,
+      "learning_rate": 6.213001807366287e-06,
+      "loss": 0.2878,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9599332220367279,
+      "grad_norm": 0.4927468753829959,
+      "learning_rate": 6.210174907348847e-06,
+      "loss": 0.2868,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9602671118530884,
+      "grad_norm": 0.4831828292002612,
+      "learning_rate": 6.207347596362265e-06,
+      "loss": 0.274,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9606010016694491,
+      "grad_norm": 0.4812104795520979,
+      "learning_rate": 6.204519875366681e-06,
+      "loss": 0.2862,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9609348914858097,
+      "grad_norm": 0.47964465648388716,
+      "learning_rate": 6.201691745322377e-06,
+      "loss": 0.2844,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9612687813021703,
+      "grad_norm": 0.49754662621571827,
+      "learning_rate": 6.198863207189768e-06,
+      "loss": 0.2816,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9616026711185309,
+      "grad_norm": 0.4722027412639167,
+      "learning_rate": 6.196034261929413e-06,
+      "loss": 0.2678,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9619365609348914,
+      "grad_norm": 0.4770166926976209,
+      "learning_rate": 6.193204910502006e-06,
+      "loss": 0.2954,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9622704507512521,
+      "grad_norm": 0.5082026481993905,
+      "learning_rate": 6.190375153868381e-06,
+      "loss": 0.2984,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9626043405676127,
+      "grad_norm": 0.49613397343852855,
+      "learning_rate": 6.187544992989508e-06,
+      "loss": 0.2882,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9629382303839733,
+      "grad_norm": 0.47681480226212747,
+      "learning_rate": 6.184714428826496e-06,
+      "loss": 0.2992,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9632721202003339,
+      "grad_norm": 0.5189518695125312,
+      "learning_rate": 6.181883462340589e-06,
+      "loss": 0.2818,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9636060100166944,
+      "grad_norm": 0.5273530005610674,
+      "learning_rate": 6.179052094493169e-06,
+      "loss": 0.2862,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9639398998330551,
+      "grad_norm": 0.4912750866956399,
+      "learning_rate": 6.176220326245754e-06,
+      "loss": 0.2806,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9642737896494157,
+      "grad_norm": 0.5092640448930034,
+      "learning_rate": 6.1733881585599985e-06,
+      "loss": 0.2929,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9646076794657763,
+      "grad_norm": 0.4910540052325372,
+      "learning_rate": 6.170555592397691e-06,
+      "loss": 0.2861,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9649415692821369,
+      "grad_norm": 0.47000503877871824,
+      "learning_rate": 6.1677226287207545e-06,
+      "loss": 0.2852,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9652754590984975,
+      "grad_norm": 0.4995094719428579,
+      "learning_rate": 6.164889268491253e-06,
+      "loss": 0.2859,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9656093489148581,
+      "grad_norm": 0.49674232136067853,
+      "learning_rate": 6.162055512671378e-06,
+      "loss": 0.2838,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9659432387312187,
+      "grad_norm": 0.5115382037875036,
+      "learning_rate": 6.15922136222346e-06,
+      "loss": 0.2957,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9662771285475793,
+      "grad_norm": 0.4369761738638269,
+      "learning_rate": 6.1563868181099596e-06,
+      "loss": 0.2716,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9666110183639399,
+      "grad_norm": 0.4692773714124014,
+      "learning_rate": 6.153551881293475e-06,
+      "loss": 0.2836,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9669449081803005,
+      "grad_norm": 0.462468549738759,
+      "learning_rate": 6.1507165527367376e-06,
+      "loss": 0.2839,
+      "step": 2896
+    },
+    {
+      "epoch": 0.967278797996661,
+      "grad_norm": 0.4965615663619004,
+      "learning_rate": 6.147880833402606e-06,
+      "loss": 0.2837,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9676126878130217,
+      "grad_norm": 0.48785218557133575,
+      "learning_rate": 6.145044724254079e-06,
+      "loss": 0.2884,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9679465776293823,
+      "grad_norm": 0.48572193295908295,
+      "learning_rate": 6.142208226254285e-06,
+      "loss": 0.2927,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9682804674457429,
+      "grad_norm": 0.5020508975804961,
+      "learning_rate": 6.139371340366482e-06,
+      "loss": 0.2746,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9686143572621035,
+      "grad_norm": 0.4734208511898631,
+      "learning_rate": 6.136534067554063e-06,
+      "loss": 0.291,
+      "step": 2901
+    },
+    {
+      "epoch": 0.968948247078464,
+      "grad_norm": 0.4839300150367573,
+      "learning_rate": 6.133696408780551e-06,
+      "loss": 0.2866,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9692821368948247,
+      "grad_norm": 0.4968348337445214,
+      "learning_rate": 6.130858365009601e-06,
+      "loss": 0.2743,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9696160267111853,
+      "grad_norm": 0.4608013188054011,
+      "learning_rate": 6.128019937204997e-06,
+      "loss": 0.2786,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9699499165275459,
+      "grad_norm": 0.5068651728834515,
+      "learning_rate": 6.1251811263306535e-06,
+      "loss": 0.2898,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9702838063439065,
+      "grad_norm": 0.47386353957773975,
+      "learning_rate": 6.1223419333506185e-06,
+      "loss": 0.2815,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9706176961602672,
+      "grad_norm": 0.4895251434135455,
+      "learning_rate": 6.119502359229065e-06,
+      "loss": 0.2928,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9709515859766277,
+      "grad_norm": 0.4412885688824866,
+      "learning_rate": 6.116662404930298e-06,
+      "loss": 0.2649,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9712854757929883,
+      "grad_norm": 0.46003716367775604,
+      "learning_rate": 6.1138220714187535e-06,
+      "loss": 0.2706,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9716193656093489,
+      "grad_norm": 0.5097301651934046,
+      "learning_rate": 6.110981359658992e-06,
+      "loss": 0.2827,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9719532554257095,
+      "grad_norm": 0.4656201876122148,
+      "learning_rate": 6.108140270615702e-06,
+      "loss": 0.2791,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9722871452420702,
+      "grad_norm": 0.48006257077726194,
+      "learning_rate": 6.105298805253709e-06,
+      "loss": 0.2804,
+      "step": 2912
+    },
+    {
+      "epoch": 0.9726210350584307,
+      "grad_norm": 0.4436923999030712,
+      "learning_rate": 6.102456964537955e-06,
+      "loss": 0.2758,
+      "step": 2913
+    },
+    {
+      "epoch": 0.9729549248747913,
+      "grad_norm": 0.4709304903190613,
+      "learning_rate": 6.0996147494335174e-06,
+      "loss": 0.2687,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9732888146911519,
+      "grad_norm": 0.4625950012744203,
+      "learning_rate": 6.0967721609055956e-06,
+      "loss": 0.2863,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9736227045075125,
+      "grad_norm": 0.4643972730709825,
+      "learning_rate": 6.093929199919517e-06,
+      "loss": 0.2834,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9739565943238732,
+      "grad_norm": 0.47592502580356555,
+      "learning_rate": 6.0910858674407406e-06,
+      "loss": 0.2827,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9742904841402337,
+      "grad_norm": 0.45513526024950157,
+      "learning_rate": 6.0882421644348435e-06,
+      "loss": 0.2842,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9746243739565943,
+      "grad_norm": 0.4621760520577419,
+      "learning_rate": 6.085398091867534e-06,
+      "loss": 0.2809,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9749582637729549,
+      "grad_norm": 0.448751658944793,
+      "learning_rate": 6.082553650704644e-06,
+      "loss": 0.2709,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9752921535893155,
+      "grad_norm": 0.4631672472634566,
+      "learning_rate": 6.079708841912133e-06,
+      "loss": 0.2751,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9756260434056762,
+      "grad_norm": 0.464410275962735,
+      "learning_rate": 6.076863666456081e-06,
+      "loss": 0.2841,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9759599332220368,
+      "grad_norm": 0.4628846945306602,
+      "learning_rate": 6.074018125302697e-06,
+      "loss": 0.2745,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9762938230383973,
+      "grad_norm": 0.4692021878786526,
+      "learning_rate": 6.071172219418309e-06,
+      "loss": 0.2758,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9766277128547579,
+      "grad_norm": 0.4662806030096645,
+      "learning_rate": 6.0683259497693736e-06,
+      "loss": 0.2984,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9769616026711185,
+      "grad_norm": 0.4741366595030662,
+      "learning_rate": 6.0654793173224706e-06,
+      "loss": 0.2936,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9772954924874792,
+      "grad_norm": 0.46663631439221465,
+      "learning_rate": 6.062632323044299e-06,
+      "loss": 0.2719,
+      "step": 2927
+    },
+    {
+      "epoch": 0.9776293823038398,
+      "grad_norm": 0.45363307574375855,
+      "learning_rate": 6.059784967901684e-06,
+      "loss": 0.283,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9779632721202003,
+      "grad_norm": 0.45872471113847246,
+      "learning_rate": 6.056937252861574e-06,
+      "loss": 0.2707,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9782971619365609,
+      "grad_norm": 0.4689491292322838,
+      "learning_rate": 6.054089178891039e-06,
+      "loss": 0.2885,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9786310517529215,
+      "grad_norm": 0.44487997579838373,
+      "learning_rate": 6.0512407469572675e-06,
+      "loss": 0.2726,
+      "step": 2931
+    },
+    {
+      "epoch": 0.9789649415692822,
+      "grad_norm": 0.4695763591049369,
+      "learning_rate": 6.048391958027574e-06,
+      "loss": 0.2723,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9792988313856428,
+      "grad_norm": 0.48324020724693145,
+      "learning_rate": 6.045542813069394e-06,
+      "loss": 0.2834,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9796327212020033,
+      "grad_norm": 0.46754214119861354,
+      "learning_rate": 6.0426933130502786e-06,
+      "loss": 0.2802,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9799666110183639,
+      "grad_norm": 0.44780183552190833,
+      "learning_rate": 6.039843458937905e-06,
+      "loss": 0.2691,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9803005008347245,
+      "grad_norm": 0.46819886609406547,
+      "learning_rate": 6.036993251700069e-06,
+      "loss": 0.2903,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9806343906510852,
+      "grad_norm": 0.479669404085268,
+      "learning_rate": 6.034142692304688e-06,
+      "loss": 0.2977,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9809682804674458,
+      "grad_norm": 0.46573387158461,
+      "learning_rate": 6.031291781719793e-06,
+      "loss": 0.2887,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9813021702838064,
+      "grad_norm": 0.4625017926604298,
+      "learning_rate": 6.028440520913545e-06,
+      "loss": 0.2712,
+      "step": 2939
+    },
+    {
+      "epoch": 0.9816360601001669,
+      "grad_norm": 0.44876297730872583,
+      "learning_rate": 6.02558891085421e-06,
+      "loss": 0.2724,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9819699499165275,
+      "grad_norm": 0.4420318562195873,
+      "learning_rate": 6.022736952510184e-06,
+      "loss": 0.2702,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9823038397328882,
+      "grad_norm": 0.4433314313047931,
+      "learning_rate": 6.019884646849978e-06,
+      "loss": 0.2764,
+      "step": 2942
+    },
+    {
+      "epoch": 0.9826377295492488,
+      "grad_norm": 0.47257590355226564,
+      "learning_rate": 6.017031994842218e-06,
+      "loss": 0.2812,
+      "step": 2943
+    },
+    {
+      "epoch": 0.9829716193656094,
+      "grad_norm": 0.4879227423529872,
+      "learning_rate": 6.014178997455653e-06,
+      "loss": 0.2909,
+      "step": 2944
+    },
+    {
+      "epoch": 0.9833055091819699,
+      "grad_norm": 0.460322086131244,
+      "learning_rate": 6.0113256556591415e-06,
+      "loss": 0.2536,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9836393989983305,
+      "grad_norm": 0.460048673151597,
+      "learning_rate": 6.008471970421667e-06,
+      "loss": 0.2735,
+      "step": 2946
+    },
+    {
+      "epoch": 0.9839732888146911,
+      "grad_norm": 0.4773223479943814,
+      "learning_rate": 6.0056179427123266e-06,
+      "loss": 0.2886,
+      "step": 2947
+    },
+    {
+      "epoch": 0.9843071786310518,
+      "grad_norm": 0.47868829888184333,
+      "learning_rate": 6.002763573500332e-06,
+      "loss": 0.2937,
+      "step": 2948
+    },
+    {
+      "epoch": 0.9846410684474124,
+      "grad_norm": 0.45564253718181325,
+      "learning_rate": 5.9999088637550126e-06,
+      "loss": 0.2797,
+      "step": 2949
+    },
+    {
+      "epoch": 0.9849749582637729,
+      "grad_norm": 0.4570071080291828,
+      "learning_rate": 5.997053814445812e-06,
+      "loss": 0.2781,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9853088480801335,
+      "grad_norm": 0.45639796006835764,
+      "learning_rate": 5.994198426542293e-06,
+      "loss": 0.2681,
+      "step": 2951
+    },
+    {
+      "epoch": 0.9856427378964941,
+      "grad_norm": 0.46644509725777683,
+      "learning_rate": 5.991342701014127e-06,
+      "loss": 0.2811,
+      "step": 2952
+    },
+    {
+      "epoch": 0.9859766277128548,
+      "grad_norm": 0.4925772054121782,
+      "learning_rate": 5.988486638831105e-06,
+      "loss": 0.2932,
+      "step": 2953
+    },
+    {
+      "epoch": 0.9863105175292154,
+      "grad_norm": 0.5031731761587677,
+      "learning_rate": 5.985630240963132e-06,
+      "loss": 0.2941,
+      "step": 2954
+    },
+    {
+      "epoch": 0.986644407345576,
+      "grad_norm": 0.451031521524845,
+      "learning_rate": 5.982773508380226e-06,
+      "loss": 0.2966,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9869782971619365,
+      "grad_norm": 0.465779673831049,
+      "learning_rate": 5.979916442052517e-06,
+      "loss": 0.2852,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9873121869782971,
+      "grad_norm": 0.4667483762582195,
+      "learning_rate": 5.977059042950252e-06,
+      "loss": 0.2896,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9876460767946578,
+      "grad_norm": 0.47108862726763673,
+      "learning_rate": 5.974201312043786e-06,
+      "loss": 0.2783,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9879799666110184,
+      "grad_norm": 0.4753322688072151,
+      "learning_rate": 5.971343250303592e-06,
+      "loss": 0.2846,
+      "step": 2959
+    },
+    {
+      "epoch": 0.988313856427379,
+      "grad_norm": 0.4764532731681164,
+      "learning_rate": 5.968484858700253e-06,
+      "loss": 0.2744,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9886477462437395,
+      "grad_norm": 0.44938863446477173,
+      "learning_rate": 5.965626138204463e-06,
+      "loss": 0.2754,
+      "step": 2961
+    },
+    {
+      "epoch": 0.9889816360601001,
+      "grad_norm": 0.49615488556490484,
+      "learning_rate": 5.9627670897870285e-06,
+      "loss": 0.2913,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9893155258764608,
+      "grad_norm": 0.5304209726745707,
+      "learning_rate": 5.959907714418868e-06,
+      "loss": 0.2821,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9896494156928214,
+      "grad_norm": 0.46871327400861573,
+      "learning_rate": 5.957048013071013e-06,
+      "loss": 0.2827,
+      "step": 2964
+    },
+    {
+      "epoch": 0.989983305509182,
+      "grad_norm": 0.4503915032269601,
+      "learning_rate": 5.954187986714599e-06,
+      "loss": 0.2716,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9903171953255425,
+      "grad_norm": 0.49059855910085504,
+      "learning_rate": 5.951327636320878e-06,
+      "loss": 0.2802,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9906510851419031,
+      "grad_norm": 0.5026481144718845,
+      "learning_rate": 5.948466962861212e-06,
+      "loss": 0.2859,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9909849749582638,
+      "grad_norm": 0.5020962999027283,
+      "learning_rate": 5.9456059673070685e-06,
+      "loss": 0.2824,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9913188647746244,
+      "grad_norm": 0.44020155466181976,
+      "learning_rate": 5.942744650630028e-06,
+      "loss": 0.2786,
+      "step": 2969
+    },
+    {
+      "epoch": 0.991652754590985,
+      "grad_norm": 0.45484469632911967,
+      "learning_rate": 5.9398830138017774e-06,
+      "loss": 0.2772,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9919866444073456,
+      "grad_norm": 0.4716582737715824,
+      "learning_rate": 5.937021057794116e-06,
+      "loss": 0.2927,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9923205342237061,
+      "grad_norm": 0.49428990437534237,
+      "learning_rate": 5.934158783578947e-06,
+      "loss": 0.2808,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9926544240400668,
+      "grad_norm": 0.48622873284027474,
+      "learning_rate": 5.9312961921282866e-06,
+      "loss": 0.2825,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9929883138564274,
+      "grad_norm": 0.4510872165093165,
+      "learning_rate": 5.928433284414253e-06,
+      "loss": 0.2811,
+      "step": 2974
+    },
+    {
+      "epoch": 0.993322203672788,
+      "grad_norm": 0.44850071606709674,
+      "learning_rate": 5.925570061409077e-06,
+      "loss": 0.2785,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9936560934891486,
+      "grad_norm": 0.46795219276756533,
+      "learning_rate": 5.922706524085094e-06,
+      "loss": 0.2761,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9939899833055091,
+      "grad_norm": 0.49247216031000435,
+      "learning_rate": 5.919842673414746e-06,
+      "loss": 0.2859,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9943238731218698,
+      "grad_norm": 0.46350426430124275,
+      "learning_rate": 5.9169785103705824e-06,
+      "loss": 0.2706,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9946577629382304,
+      "grad_norm": 0.4842123145431046,
+      "learning_rate": 5.914114035925256e-06,
+      "loss": 0.2795,
+      "step": 2979
+    },
+    {
+      "epoch": 0.994991652754591,
+      "grad_norm": 0.47555392503899213,
+      "learning_rate": 5.911249251051532e-06,
+      "loss": 0.28,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9953255425709516,
+      "grad_norm": 0.4760857881318245,
+      "learning_rate": 5.908384156722271e-06,
+      "loss": 0.2765,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9956594323873121,
+      "grad_norm": 0.5448692244472205,
+      "learning_rate": 5.90551875391045e-06,
+      "loss": 0.2906,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9959933222036728,
+      "grad_norm": 0.46007902884128965,
+      "learning_rate": 5.9026530435891405e-06,
+      "loss": 0.2843,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9963272120200334,
+      "grad_norm": 0.4485867599206855,
+      "learning_rate": 5.899787026731524e-06,
+      "loss": 0.2816,
+      "step": 2984
+    },
+    {
+      "epoch": 0.996661101836394,
+      "grad_norm": 0.4732867020682718,
+      "learning_rate": 5.8969207043108865e-06,
+      "loss": 0.2767,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9969949916527546,
+      "grad_norm": 0.49821885396778287,
+      "learning_rate": 5.894054077300614e-06,
+      "loss": 0.2866,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9973288814691152,
+      "grad_norm": 0.48760120461009476,
+      "learning_rate": 5.8911871466742e-06,
+      "loss": 0.2673,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9976627712854758,
+      "grad_norm": 0.4517019464166034,
+      "learning_rate": 5.888319913405238e-06,
+      "loss": 0.2804,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9979966611018364,
+      "grad_norm": 0.4695759037913214,
+      "learning_rate": 5.885452378467429e-06,
+      "loss": 0.2824,
+      "step": 2989
+    },
+    {
+      "epoch": 0.998330550918197,
+      "grad_norm": 0.4705021854805188,
+      "learning_rate": 5.8825845428345674e-06,
+      "loss": 0.2815,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9986644407345576,
+      "grad_norm": 0.4558418909176287,
+      "learning_rate": 5.879716407480559e-06,
+      "loss": 0.2885,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9989983305509182,
+      "grad_norm": 0.44509380887526606,
+      "learning_rate": 5.876847973379409e-06,
+      "loss": 0.2786,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9993322203672788,
+      "grad_norm": 0.4564647883594583,
+      "learning_rate": 5.873979241505219e-06,
+      "loss": 0.2869,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9996661101836394,
+      "grad_norm": 0.4561636909712985,
+      "learning_rate": 5.871110212832199e-06,
+      "loss": 0.2791,
+      "step": 2994
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46781827197476294,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.2795,
+      "step": 2995
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.28170257806777954,
+      "eval_runtime": 443.812,
+      "eval_samples_per_second": 45.465,
+      "eval_steps_per_second": 0.356,
+      "step": 2995
+    },
+    {
+      "epoch": 1.0003338898163605,
+      "grad_norm": 0.47930623150126495,
+      "learning_rate": 5.86537126898699e-06,
+      "loss": 0.259,
+      "step": 2996
+    },
+    {
+      "epoch": 1.0006677796327212,
+      "grad_norm": 0.46606797452682325,
+      "learning_rate": 5.862501355763719e-06,
+      "loss": 0.2618,
+      "step": 2997
+    },
+    {
+      "epoch": 1.0010016694490818,
+      "grad_norm": 0.45209682883206387,
+      "learning_rate": 5.859631149639449e-06,
+      "loss": 0.2569,
+      "step": 2998
+    },
+    {
+      "epoch": 1.0013355592654425,
+      "grad_norm": 0.48094465272460063,
+      "learning_rate": 5.856760651588882e-06,
+      "loss": 0.255,
+      "step": 2999
+    },
+    {
+      "epoch": 1.001669449081803,
+      "grad_norm": 0.45064556907976794,
+      "learning_rate": 5.853889862586829e-06,
+      "loss": 0.2553,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0020033388981635,
+      "grad_norm": 0.479651886448271,
+      "learning_rate": 5.851018783608195e-06,
+      "loss": 0.2601,
+      "step": 3001
+    },
+    {
+      "epoch": 1.0023372287145242,
+      "grad_norm": 0.5074697551843629,
+      "learning_rate": 5.848147415627981e-06,
+      "loss": 0.2577,
+      "step": 3002
+    },
+    {
+      "epoch": 1.0026711185308848,
+      "grad_norm": 0.5116802949196217,
+      "learning_rate": 5.84527575962129e-06,
+      "loss": 0.2647,
+      "step": 3003
+    },
+    {
+      "epoch": 1.0030050083472455,
+      "grad_norm": 0.45207555814636835,
+      "learning_rate": 5.842403816563319e-06,
+      "loss": 0.2445,
+      "step": 3004
+    },
+    {
+      "epoch": 1.003338898163606,
+      "grad_norm": 0.48551430162137105,
+      "learning_rate": 5.83953158742937e-06,
+      "loss": 0.2491,
+      "step": 3005
+    },
+    {
+      "epoch": 1.0036727879799665,
+      "grad_norm": 0.5265774545532096,
+      "learning_rate": 5.836659073194835e-06,
+      "loss": 0.2591,
+      "step": 3006
+    },
+    {
+      "epoch": 1.0040066777963272,
+      "grad_norm": 0.47722841521251597,
+      "learning_rate": 5.833786274835204e-06,
+      "loss": 0.257,
+      "step": 3007
+    },
+    {
+      "epoch": 1.0043405676126878,
+      "grad_norm": 0.44107818212939204,
+      "learning_rate": 5.830913193326064e-06,
+      "loss": 0.2413,
+      "step": 3008
+    },
+    {
+      "epoch": 1.0046744574290485,
+      "grad_norm": 0.4944625901500407,
+      "learning_rate": 5.828039829643101e-06,
+      "loss": 0.2673,
+      "step": 3009
+    },
+    {
+      "epoch": 1.005008347245409,
+      "grad_norm": 0.4746147806443608,
+      "learning_rate": 5.825166184762092e-06,
+      "loss": 0.2631,
+      "step": 3010
+    },
+    {
+      "epoch": 1.0053422370617695,
+      "grad_norm": 0.4567713814765493,
+      "learning_rate": 5.822292259658914e-06,
+      "loss": 0.2438,
+      "step": 3011
+    },
+    {
+      "epoch": 1.0056761268781302,
+      "grad_norm": 0.4645684382609121,
+      "learning_rate": 5.819418055309536e-06,
+      "loss": 0.2572,
+      "step": 3012
+    },
+    {
+      "epoch": 1.0060100166944907,
+      "grad_norm": 0.4718639489812278,
+      "learning_rate": 5.816543572690023e-06,
+      "loss": 0.2482,
+      "step": 3013
+    },
+    {
+      "epoch": 1.0063439065108515,
+      "grad_norm": 0.48110675275859527,
+      "learning_rate": 5.813668812776534e-06,
+      "loss": 0.2544,
+      "step": 3014
+    },
+    {
+      "epoch": 1.006677796327212,
+      "grad_norm": 0.47287275198368295,
+      "learning_rate": 5.810793776545325e-06,
+      "loss": 0.2601,
+      "step": 3015
+    },
+    {
+      "epoch": 1.0070116861435727,
+      "grad_norm": 0.49799449315710953,
+      "learning_rate": 5.807918464972739e-06,
+      "loss": 0.2724,
+      "step": 3016
+    },
+    {
+      "epoch": 1.0073455759599332,
+      "grad_norm": 0.4704218157882534,
+      "learning_rate": 5.805042879035221e-06,
+      "loss": 0.2577,
+      "step": 3017
+    },
+    {
+      "epoch": 1.0076794657762937,
+      "grad_norm": 0.4672123810356925,
+      "learning_rate": 5.802167019709303e-06,
+      "loss": 0.2499,
+      "step": 3018
+    },
+    {
+      "epoch": 1.0080133555926545,
+      "grad_norm": 0.49833817904925376,
+      "learning_rate": 5.799290887971614e-06,
+      "loss": 0.2588,
+      "step": 3019
+    },
+    {
+      "epoch": 1.008347245409015,
+      "grad_norm": 0.5037331848848845,
+      "learning_rate": 5.79641448479887e-06,
+      "loss": 0.2654,
+      "step": 3020
+    },
+    {
+      "epoch": 1.0086811352253757,
+      "grad_norm": 0.4833773148048413,
+      "learning_rate": 5.793537811167885e-06,
+      "loss": 0.2551,
+      "step": 3021
+    },
+    {
+      "epoch": 1.0090150250417362,
+      "grad_norm": 0.4698073627600636,
+      "learning_rate": 5.790660868055564e-06,
+      "loss": 0.2619,
+      "step": 3022
+    },
+    {
+      "epoch": 1.0093489148580967,
+      "grad_norm": 0.47178610834002005,
+      "learning_rate": 5.787783656438899e-06,
+      "loss": 0.2481,
+      "step": 3023
+    },
+    {
+      "epoch": 1.0096828046744575,
+      "grad_norm": 0.4672651634290253,
+      "learning_rate": 5.7849061772949785e-06,
+      "loss": 0.2476,
+      "step": 3024
+    },
+    {
+      "epoch": 1.010016694490818,
+      "grad_norm": 0.4751446135551542,
+      "learning_rate": 5.782028431600977e-06,
+      "loss": 0.2622,
+      "step": 3025
+    },
+    {
+      "epoch": 1.0103505843071787,
+      "grad_norm": 0.47842354169176554,
+      "learning_rate": 5.7791504203341645e-06,
+      "loss": 0.2573,
+      "step": 3026
+    },
+    {
+      "epoch": 1.0106844741235392,
+      "grad_norm": 0.47256582959523247,
+      "learning_rate": 5.776272144471896e-06,
+      "loss": 0.2508,
+      "step": 3027
+    },
+    {
+      "epoch": 1.0110183639398997,
+      "grad_norm": 0.4980163636977225,
+      "learning_rate": 5.773393604991623e-06,
+      "loss": 0.2605,
+      "step": 3028
+    },
+    {
+      "epoch": 1.0113522537562605,
+      "grad_norm": 0.4679509533101371,
+      "learning_rate": 5.770514802870879e-06,
+      "loss": 0.2544,
+      "step": 3029
+    },
+    {
+      "epoch": 1.011686143572621,
+      "grad_norm": 0.4688546484044532,
+      "learning_rate": 5.7676357390872925e-06,
+      "loss": 0.2582,
+      "step": 3030
+    },
+    {
+      "epoch": 1.0120200333889817,
+      "grad_norm": 0.47516968488058786,
+      "learning_rate": 5.764756414618579e-06,
+      "loss": 0.2489,
+      "step": 3031
+    },
+    {
+      "epoch": 1.0123539232053422,
+      "grad_norm": 0.507781268622218,
+      "learning_rate": 5.76187683044254e-06,
+      "loss": 0.2721,
+      "step": 3032
+    },
+    {
+      "epoch": 1.0126878130217027,
+      "grad_norm": 0.46467738600880437,
+      "learning_rate": 5.7589969875370685e-06,
+      "loss": 0.2511,
+      "step": 3033
+    },
+    {
+      "epoch": 1.0130217028380635,
+      "grad_norm": 0.4621023649903183,
+      "learning_rate": 5.756116886880143e-06,
+      "loss": 0.2514,
+      "step": 3034
+    },
+    {
+      "epoch": 1.013355592654424,
+      "grad_norm": 0.4991059016654631,
+      "learning_rate": 5.753236529449835e-06,
+      "loss": 0.2621,
+      "step": 3035
+    },
+    {
+      "epoch": 1.0136894824707847,
+      "grad_norm": 0.4690711140455441,
+      "learning_rate": 5.750355916224294e-06,
+      "loss": 0.2622,
+      "step": 3036
+    },
+    {
+      "epoch": 1.0140233722871452,
+      "grad_norm": 0.5085285574371332,
+      "learning_rate": 5.747475048181765e-06,
+      "loss": 0.2667,
+      "step": 3037
+    },
+    {
+      "epoch": 1.0143572621035057,
+      "grad_norm": 0.46725109012211596,
+      "learning_rate": 5.744593926300573e-06,
+      "loss": 0.2452,
+      "step": 3038
+    },
+    {
+      "epoch": 1.0146911519198665,
+      "grad_norm": 0.46468245697929483,
+      "learning_rate": 5.741712551559133e-06,
+      "loss": 0.2534,
+      "step": 3039
+    },
+    {
+      "epoch": 1.015025041736227,
+      "grad_norm": 0.4850343659691793,
+      "learning_rate": 5.738830924935948e-06,
+      "loss": 0.2703,
+      "step": 3040
+    },
+    {
+      "epoch": 1.0153589315525877,
+      "grad_norm": 0.4609372349130885,
+      "learning_rate": 5.735949047409599e-06,
+      "loss": 0.2565,
+      "step": 3041
+    },
+    {
+      "epoch": 1.0156928213689482,
+      "grad_norm": 0.4621706664015058,
+      "learning_rate": 5.7330669199587585e-06,
+      "loss": 0.2537,
+      "step": 3042
+    },
+    {
+      "epoch": 1.0160267111853087,
+      "grad_norm": 0.47276973362250896,
+      "learning_rate": 5.730184543562183e-06,
+      "loss": 0.2614,
+      "step": 3043
+    },
+    {
+      "epoch": 1.0163606010016695,
+      "grad_norm": 0.4956902686244128,
+      "learning_rate": 5.7273019191987115e-06,
+      "loss": 0.2615,
+      "step": 3044
+    },
+    {
+      "epoch": 1.01669449081803,
+      "grad_norm": 0.47516699055688427,
+      "learning_rate": 5.724419047847268e-06,
+      "loss": 0.2612,
+      "step": 3045
+    },
+    {
+      "epoch": 1.0170283806343907,
+      "grad_norm": 0.49607920763824637,
+      "learning_rate": 5.721535930486863e-06,
+      "loss": 0.2569,
+      "step": 3046
+    },
+    {
+      "epoch": 1.0173622704507512,
+      "grad_norm": 0.4735272686415401,
+      "learning_rate": 5.718652568096585e-06,
+      "loss": 0.2446,
+      "step": 3047
+    },
+    {
+      "epoch": 1.017696160267112,
+      "grad_norm": 0.46643734541983745,
+      "learning_rate": 5.715768961655611e-06,
+      "loss": 0.2507,
+      "step": 3048
+    },
+    {
+      "epoch": 1.0180300500834725,
+      "grad_norm": 0.5222425089246563,
+      "learning_rate": 5.7128851121431996e-06,
+      "loss": 0.2611,
+      "step": 3049
+    },
+    {
+      "epoch": 1.018363939899833,
+      "grad_norm": 0.49995043798304883,
+      "learning_rate": 5.710001020538688e-06,
+      "loss": 0.2598,
+      "step": 3050
+    },
+    {
+      "epoch": 1.0186978297161937,
+      "grad_norm": 0.49305925611406887,
+      "learning_rate": 5.707116687821501e-06,
+      "loss": 0.2604,
+      "step": 3051
+    },
+    {
+      "epoch": 1.0190317195325542,
+      "grad_norm": 0.502234674848021,
+      "learning_rate": 5.704232114971144e-06,
+      "loss": 0.2627,
+      "step": 3052
+    },
+    {
+      "epoch": 1.019365609348915,
+      "grad_norm": 0.4712924492458562,
+      "learning_rate": 5.701347302967203e-06,
+      "loss": 0.2548,
+      "step": 3053
+    },
+    {
+      "epoch": 1.0196994991652755,
+      "grad_norm": 0.4783849896604598,
+      "learning_rate": 5.698462252789342e-06,
+      "loss": 0.2461,
+      "step": 3054
+    },
+    {
+      "epoch": 1.020033388981636,
+      "grad_norm": 0.5035555725723966,
+      "learning_rate": 5.695576965417311e-06,
+      "loss": 0.2564,
+      "step": 3055
+    },
+    {
+      "epoch": 1.0203672787979967,
+      "grad_norm": 0.47144877377672234,
+      "learning_rate": 5.6926914418309405e-06,
+      "loss": 0.2565,
+      "step": 3056
+    },
+    {
+      "epoch": 1.0207011686143572,
+      "grad_norm": 0.4711260626018403,
+      "learning_rate": 5.689805683010137e-06,
+      "loss": 0.2551,
+      "step": 3057
+    },
+    {
+      "epoch": 1.021035058430718,
+      "grad_norm": 0.46571519286373214,
+      "learning_rate": 5.686919689934891e-06,
+      "loss": 0.2513,
+      "step": 3058
+    },
+    {
+      "epoch": 1.0213689482470785,
+      "grad_norm": 0.4758022819565318,
+      "learning_rate": 5.684033463585268e-06,
+      "loss": 0.2528,
+      "step": 3059
+    },
+    {
+      "epoch": 1.021702838063439,
+      "grad_norm": 0.4628082211806178,
+      "learning_rate": 5.68114700494142e-06,
+      "loss": 0.2567,
+      "step": 3060
+    },
+    {
+      "epoch": 1.0220367278797997,
+      "grad_norm": 0.4700983613413316,
+      "learning_rate": 5.678260314983569e-06,
+      "loss": 0.2576,
+      "step": 3061
+    },
+    {
+      "epoch": 1.0223706176961602,
+      "grad_norm": 0.472235554334349,
+      "learning_rate": 5.675373394692023e-06,
+      "loss": 0.2548,
+      "step": 3062
+    },
+    {
+      "epoch": 1.022704507512521,
+      "grad_norm": 0.4643225082647862,
+      "learning_rate": 5.672486245047163e-06,
+      "loss": 0.251,
+      "step": 3063
+    },
+    {
+      "epoch": 1.0230383973288815,
+      "grad_norm": 0.45045869472413,
+      "learning_rate": 5.6695988670294515e-06,
+      "loss": 0.2523,
+      "step": 3064
+    },
+    {
+      "epoch": 1.023372287145242,
+      "grad_norm": 0.47481495025101633,
+      "learning_rate": 5.666711261619429e-06,
+      "loss": 0.2588,
+      "step": 3065
+    },
+    {
+      "epoch": 1.0237061769616027,
+      "grad_norm": 0.4587027990674556,
+      "learning_rate": 5.663823429797706e-06,
+      "loss": 0.2567,
+      "step": 3066
+    },
+    {
+      "epoch": 1.0240400667779632,
+      "grad_norm": 0.4798165521772847,
+      "learning_rate": 5.6609353725449804e-06,
+      "loss": 0.2548,
+      "step": 3067
+    },
+    {
+      "epoch": 1.024373956594324,
+      "grad_norm": 0.49604298621689374,
+      "learning_rate": 5.658047090842017e-06,
+      "loss": 0.2679,
+      "step": 3068
+    },
+    {
+      "epoch": 1.0247078464106845,
+      "grad_norm": 0.471519631209844,
+      "learning_rate": 5.6551585856696665e-06,
+      "loss": 0.2541,
+      "step": 3069
+    },
+    {
+      "epoch": 1.025041736227045,
+      "grad_norm": 0.4753003663818607,
+      "learning_rate": 5.652269858008848e-06,
+      "loss": 0.2622,
+      "step": 3070
+    },
+    {
+      "epoch": 1.0253756260434057,
+      "grad_norm": 0.4816040198066664,
+      "learning_rate": 5.649380908840557e-06,
+      "loss": 0.2652,
+      "step": 3071
+    },
+    {
+      "epoch": 1.0257095158597662,
+      "grad_norm": 0.4812559352481771,
+      "learning_rate": 5.646491739145867e-06,
+      "loss": 0.259,
+      "step": 3072
+    },
+    {
+      "epoch": 1.026043405676127,
+      "grad_norm": 0.44981128215792077,
+      "learning_rate": 5.643602349905924e-06,
+      "loss": 0.2486,
+      "step": 3073
+    },
+    {
+      "epoch": 1.0263772954924875,
+      "grad_norm": 0.5010216661396969,
+      "learning_rate": 5.640712742101954e-06,
+      "loss": 0.2702,
+      "step": 3074
+    },
+    {
+      "epoch": 1.026711185308848,
+      "grad_norm": 0.5011982477593468,
+      "learning_rate": 5.637822916715247e-06,
+      "loss": 0.2713,
+      "step": 3075
+    },
+    {
+      "epoch": 1.0270450751252087,
+      "grad_norm": 0.4739440051120083,
+      "learning_rate": 5.634932874727175e-06,
+      "loss": 0.251,
+      "step": 3076
+    },
+    {
+      "epoch": 1.0273789649415692,
+      "grad_norm": 0.4765954875699686,
+      "learning_rate": 5.632042617119181e-06,
+      "loss": 0.2593,
+      "step": 3077
+    },
+    {
+      "epoch": 1.02771285475793,
+      "grad_norm": 0.4971964209562132,
+      "learning_rate": 5.629152144872784e-06,
+      "loss": 0.2589,
+      "step": 3078
+    },
+    {
+      "epoch": 1.0280467445742905,
+      "grad_norm": 0.4679571342501188,
+      "learning_rate": 5.626261458969571e-06,
+      "loss": 0.2546,
+      "step": 3079
+    },
+    {
+      "epoch": 1.0283806343906512,
+      "grad_norm": 0.49270939341467274,
+      "learning_rate": 5.623370560391205e-06,
+      "loss": 0.2551,
+      "step": 3080
+    },
+    {
+      "epoch": 1.0287145242070117,
+      "grad_norm": 0.4810778451026181,
+      "learning_rate": 5.6204794501194185e-06,
+      "loss": 0.249,
+      "step": 3081
+    },
+    {
+      "epoch": 1.0290484140233722,
+      "grad_norm": 0.47570127362816145,
+      "learning_rate": 5.617588129136019e-06,
+      "loss": 0.257,
+      "step": 3082
+    },
+    {
+      "epoch": 1.029382303839733,
+      "grad_norm": 0.4574869226464428,
+      "learning_rate": 5.614696598422885e-06,
+      "loss": 0.2416,
+      "step": 3083
+    },
+    {
+      "epoch": 1.0297161936560935,
+      "grad_norm": 0.49637801658635744,
+      "learning_rate": 5.611804858961963e-06,
+      "loss": 0.2696,
+      "step": 3084
+    },
+    {
+      "epoch": 1.0300500834724542,
+      "grad_norm": 0.49196198492841026,
+      "learning_rate": 5.608912911735275e-06,
+      "loss": 0.2609,
+      "step": 3085
+    },
+    {
+      "epoch": 1.0303839732888147,
+      "grad_norm": 0.4954001983722929,
+      "learning_rate": 5.6060207577249094e-06,
+      "loss": 0.2596,
+      "step": 3086
+    },
+    {
+      "epoch": 1.0307178631051752,
+      "grad_norm": 0.46039393215093555,
+      "learning_rate": 5.6031283979130305e-06,
+      "loss": 0.2528,
+      "step": 3087
+    },
+    {
+      "epoch": 1.031051752921536,
+      "grad_norm": 0.4825829481128531,
+      "learning_rate": 5.600235833281864e-06,
+      "loss": 0.2656,
+      "step": 3088
+    },
+    {
+      "epoch": 1.0313856427378965,
+      "grad_norm": 0.5151100410973768,
+      "learning_rate": 5.597343064813712e-06,
+      "loss": 0.2647,
+      "step": 3089
+    },
+    {
+      "epoch": 1.0317195325542572,
+      "grad_norm": 0.501426390349246,
+      "learning_rate": 5.594450093490944e-06,
+      "loss": 0.265,
+      "step": 3090
+    },
+    {
+      "epoch": 1.0320534223706177,
+      "grad_norm": 0.48349177367168567,
+      "learning_rate": 5.591556920295998e-06,
+      "loss": 0.2522,
+      "step": 3091
+    },
+    {
+      "epoch": 1.0323873121869782,
+      "grad_norm": 0.48553368345396164,
+      "learning_rate": 5.5886635462113805e-06,
+      "loss": 0.2683,
+      "step": 3092
+    },
+    {
+      "epoch": 1.032721202003339,
+      "grad_norm": 0.4674711987045283,
+      "learning_rate": 5.585769972219664e-06,
+      "loss": 0.2616,
+      "step": 3093
+    },
+    {
+      "epoch": 1.0330550918196995,
+      "grad_norm": 0.48155604465371543,
+      "learning_rate": 5.582876199303497e-06,
+      "loss": 0.2646,
+      "step": 3094
+    },
+    {
+      "epoch": 1.0333889816360602,
+      "grad_norm": 0.4901252260327605,
+      "learning_rate": 5.579982228445585e-06,
+      "loss": 0.2569,
+      "step": 3095
+    },
+    {
+      "epoch": 1.0337228714524207,
+      "grad_norm": 0.48537446625174424,
+      "learning_rate": 5.577088060628708e-06,
+      "loss": 0.2663,
+      "step": 3096
+    },
+    {
+      "epoch": 1.0340567612687812,
+      "grad_norm": 0.47036195584256313,
+      "learning_rate": 5.574193696835709e-06,
+      "loss": 0.2506,
+      "step": 3097
+    },
+    {
+      "epoch": 1.034390651085142,
+      "grad_norm": 0.47373199259866505,
+      "learning_rate": 5.571299138049498e-06,
+      "loss": 0.2579,
+      "step": 3098
+    },
+    {
+      "epoch": 1.0347245409015025,
+      "grad_norm": 0.4714176251784416,
+      "learning_rate": 5.568404385253054e-06,
+      "loss": 0.2602,
+      "step": 3099
+    },
+    {
+      "epoch": 1.0350584307178632,
+      "grad_norm": 0.4731098956037678,
+      "learning_rate": 5.56550943942942e-06,
+      "loss": 0.2564,
+      "step": 3100
+    },
+    {
+      "epoch": 1.0353923205342237,
+      "grad_norm": 0.49624053694318304,
+      "learning_rate": 5.562614301561704e-06,
+      "loss": 0.2596,
+      "step": 3101
+    },
+    {
+      "epoch": 1.0357262103505842,
+      "grad_norm": 0.48503201754188063,
+      "learning_rate": 5.559718972633079e-06,
+      "loss": 0.2571,
+      "step": 3102
+    },
+    {
+      "epoch": 1.036060100166945,
+      "grad_norm": 0.46940727104989804,
+      "learning_rate": 5.556823453626788e-06,
+      "loss": 0.2549,
+      "step": 3103
+    },
+    {
+      "epoch": 1.0363939899833055,
+      "grad_norm": 0.5036965949073305,
+      "learning_rate": 5.553927745526129e-06,
+      "loss": 0.2565,
+      "step": 3104
+    },
+    {
+      "epoch": 1.0367278797996662,
+      "grad_norm": 0.4861220859446145,
+      "learning_rate": 5.551031849314471e-06,
+      "loss": 0.2544,
+      "step": 3105
+    },
+    {
+      "epoch": 1.0370617696160267,
+      "grad_norm": 0.47799347109874574,
+      "learning_rate": 5.548135765975247e-06,
+      "loss": 0.2568,
+      "step": 3106
+    },
+    {
+      "epoch": 1.0373956594323872,
+      "grad_norm": 0.5100316537651639,
+      "learning_rate": 5.54523949649195e-06,
+      "loss": 0.2673,
+      "step": 3107
+    },
+    {
+      "epoch": 1.037729549248748,
+      "grad_norm": 0.5077834636122853,
+      "learning_rate": 5.542343041848141e-06,
+      "loss": 0.2475,
+      "step": 3108
+    },
+    {
+      "epoch": 1.0380634390651085,
+      "grad_norm": 0.5948496268201471,
+      "learning_rate": 5.539446403027438e-06,
+      "loss": 0.2409,
+      "step": 3109
+    },
+    {
+      "epoch": 1.0383973288814692,
+      "grad_norm": 0.47198664145996977,
+      "learning_rate": 5.536549581013525e-06,
+      "loss": 0.2616,
+      "step": 3110
+    },
+    {
+      "epoch": 1.0387312186978297,
+      "grad_norm": 0.4893390797784602,
+      "learning_rate": 5.53365257679015e-06,
+      "loss": 0.2589,
+      "step": 3111
+    },
+    {
+      "epoch": 1.0390651085141904,
+      "grad_norm": 0.5022839571222563,
+      "learning_rate": 5.530755391341117e-06,
+      "loss": 0.2642,
+      "step": 3112
+    },
+    {
+      "epoch": 1.039398998330551,
+      "grad_norm": 0.5091998835838087,
+      "learning_rate": 5.5278580256503e-06,
+      "loss": 0.2707,
+      "step": 3113
+    },
+    {
+      "epoch": 1.0397328881469114,
+      "grad_norm": 0.5012085581415622,
+      "learning_rate": 5.524960480701625e-06,
+      "loss": 0.2538,
+      "step": 3114
+    },
+    {
+      "epoch": 1.0400667779632722,
+      "grad_norm": 0.45646261771577124,
+      "learning_rate": 5.522062757479084e-06,
+      "loss": 0.2528,
+      "step": 3115
+    },
+    {
+      "epoch": 1.0404006677796327,
+      "grad_norm": 0.49062687253664755,
+      "learning_rate": 5.5191648569667314e-06,
+      "loss": 0.2581,
+      "step": 3116
+    },
+    {
+      "epoch": 1.0407345575959934,
+      "grad_norm": 0.4774329133087372,
+      "learning_rate": 5.516266780148678e-06,
+      "loss": 0.2616,
+      "step": 3117
+    },
+    {
+      "epoch": 1.041068447412354,
+      "grad_norm": 0.4954851112417069,
+      "learning_rate": 5.513368528009094e-06,
+      "loss": 0.2482,
+      "step": 3118
+    },
+    {
+      "epoch": 1.0414023372287144,
+      "grad_norm": 0.49293933197938683,
+      "learning_rate": 5.510470101532213e-06,
+      "loss": 0.2548,
+      "step": 3119
+    },
+    {
+      "epoch": 1.0417362270450752,
+      "grad_norm": 0.48086363259856635,
+      "learning_rate": 5.5075715017023244e-06,
+      "loss": 0.2597,
+      "step": 3120
+    },
+    {
+      "epoch": 1.0420701168614357,
+      "grad_norm": 0.4824075275699771,
+      "learning_rate": 5.504672729503778e-06,
+      "loss": 0.2672,
+      "step": 3121
+    },
+    {
+      "epoch": 1.0424040066777964,
+      "grad_norm": 0.5074385322127578,
+      "learning_rate": 5.5017737859209845e-06,
+      "loss": 0.2696,
+      "step": 3122
+    },
+    {
+      "epoch": 1.042737896494157,
+      "grad_norm": 0.4859322595230147,
+      "learning_rate": 5.498874671938403e-06,
+      "loss": 0.266,
+      "step": 3123
+    },
+    {
+      "epoch": 1.0430717863105174,
+      "grad_norm": 0.466129133218841,
+      "learning_rate": 5.495975388540567e-06,
+      "loss": 0.2604,
+      "step": 3124
+    },
+    {
+      "epoch": 1.0434056761268782,
+      "grad_norm": 0.5101817703334887,
+      "learning_rate": 5.49307593671205e-06,
+      "loss": 0.2544,
+      "step": 3125
+    },
+    {
+      "epoch": 1.0437395659432387,
+      "grad_norm": 0.4985163838799122,
+      "learning_rate": 5.490176317437497e-06,
+      "loss": 0.2564,
+      "step": 3126
+    },
+    {
+      "epoch": 1.0440734557595994,
+      "grad_norm": 0.46552821611999806,
+      "learning_rate": 5.487276531701599e-06,
+      "loss": 0.2555,
+      "step": 3127
+    },
+    {
+      "epoch": 1.04440734557596,
+      "grad_norm": 0.45647689224089,
+      "learning_rate": 5.48437658048911e-06,
+      "loss": 0.263,
+      "step": 3128
+    },
+    {
+      "epoch": 1.0447412353923204,
+      "grad_norm": 0.4722391027272058,
+      "learning_rate": 5.48147646478484e-06,
+      "loss": 0.258,
+      "step": 3129
+    },
+    {
+      "epoch": 1.0450751252086812,
+      "grad_norm": 0.5126197721727849,
+      "learning_rate": 5.47857618557365e-06,
+      "loss": 0.2606,
+      "step": 3130
+    },
+    {
+      "epoch": 1.0454090150250417,
+      "grad_norm": 0.5095317429463951,
+      "learning_rate": 5.475675743840465e-06,
+      "loss": 0.2413,
+      "step": 3131
+    },
+    {
+      "epoch": 1.0457429048414024,
+      "grad_norm": 0.47435865642475034,
+      "learning_rate": 5.4727751405702535e-06,
+      "loss": 0.2546,
+      "step": 3132
+    },
+    {
+      "epoch": 1.046076794657763,
+      "grad_norm": 0.47233705739259146,
+      "learning_rate": 5.469874376748051e-06,
+      "loss": 0.2571,
+      "step": 3133
+    },
+    {
+      "epoch": 1.0464106844741234,
+      "grad_norm": 0.4904866299925756,
+      "learning_rate": 5.466973453358938e-06,
+      "loss": 0.2626,
+      "step": 3134
+    },
+    {
+      "epoch": 1.0467445742904842,
+      "grad_norm": 0.46744997058123755,
+      "learning_rate": 5.464072371388056e-06,
+      "loss": 0.2441,
+      "step": 3135
+    },
+    {
+      "epoch": 1.0470784641068447,
+      "grad_norm": 0.4869732340296886,
+      "learning_rate": 5.461171131820596e-06,
+      "loss": 0.244,
+      "step": 3136
+    },
+    {
+      "epoch": 1.0474123539232054,
+      "grad_norm": 0.45408930185591273,
+      "learning_rate": 5.4582697356418036e-06,
+      "loss": 0.2419,
+      "step": 3137
+    },
+    {
+      "epoch": 1.047746243739566,
+      "grad_norm": 0.4755094565343282,
+      "learning_rate": 5.4553681838369784e-06,
+      "loss": 0.2565,
+      "step": 3138
+    },
+    {
+      "epoch": 1.0480801335559264,
+      "grad_norm": 0.5020673685983205,
+      "learning_rate": 5.4524664773914725e-06,
+      "loss": 0.2488,
+      "step": 3139
+    },
+    {
+      "epoch": 1.0484140233722872,
+      "grad_norm": 0.48800979378472237,
+      "learning_rate": 5.44956461729069e-06,
+      "loss": 0.2523,
+      "step": 3140
+    },
+    {
+      "epoch": 1.0487479131886477,
+      "grad_norm": 0.4876877593814704,
+      "learning_rate": 5.446662604520088e-06,
+      "loss": 0.2518,
+      "step": 3141
+    },
+    {
+      "epoch": 1.0490818030050084,
+      "grad_norm": 0.5021860892138628,
+      "learning_rate": 5.443760440065176e-06,
+      "loss": 0.2641,
+      "step": 3142
+    },
+    {
+      "epoch": 1.049415692821369,
+      "grad_norm": 0.4648138931761416,
+      "learning_rate": 5.440858124911511e-06,
+      "loss": 0.2576,
+      "step": 3143
+    },
+    {
+      "epoch": 1.0497495826377294,
+      "grad_norm": 0.48784452143822815,
+      "learning_rate": 5.437955660044707e-06,
+      "loss": 0.268,
+      "step": 3144
+    },
+    {
+      "epoch": 1.0500834724540902,
+      "grad_norm": 0.48058808569288686,
+      "learning_rate": 5.435053046450426e-06,
+      "loss": 0.2516,
+      "step": 3145
+    },
+    {
+      "epoch": 1.0504173622704507,
+      "grad_norm": 0.48074003325127496,
+      "learning_rate": 5.4321502851143785e-06,
+      "loss": 0.2613,
+      "step": 3146
+    },
+    {
+      "epoch": 1.0507512520868114,
+      "grad_norm": 0.49739470870841834,
+      "learning_rate": 5.429247377022329e-06,
+      "loss": 0.2546,
+      "step": 3147
+    },
+    {
+      "epoch": 1.051085141903172,
+      "grad_norm": 0.4870160000359012,
+      "learning_rate": 5.426344323160091e-06,
+      "loss": 0.2524,
+      "step": 3148
+    },
+    {
+      "epoch": 1.0514190317195327,
+      "grad_norm": 0.45633358012940795,
+      "learning_rate": 5.423441124513523e-06,
+      "loss": 0.2511,
+      "step": 3149
+    },
+    {
+      "epoch": 1.0517529215358932,
+      "grad_norm": 0.4897861924135719,
+      "learning_rate": 5.420537782068539e-06,
+      "loss": 0.2568,
+      "step": 3150
+    },
+    {
+      "epoch": 1.0520868113522537,
+      "grad_norm": 0.4828568914898983,
+      "learning_rate": 5.4176342968111005e-06,
+      "loss": 0.2648,
+      "step": 3151
+    },
+    {
+      "epoch": 1.0524207011686144,
+      "grad_norm": 0.5054715112671017,
+      "learning_rate": 5.414730669727212e-06,
+      "loss": 0.2582,
+      "step": 3152
+    },
+    {
+      "epoch": 1.052754590984975,
+      "grad_norm": 0.49212586042834794,
+      "learning_rate": 5.411826901802933e-06,
+      "loss": 0.2523,
+      "step": 3153
+    },
+    {
+      "epoch": 1.0530884808013357,
+      "grad_norm": 0.490974858729316,
+      "learning_rate": 5.4089229940243685e-06,
+      "loss": 0.2687,
+      "step": 3154
+    },
+    {
+      "epoch": 1.0534223706176962,
+      "grad_norm": 0.4814782833999093,
+      "learning_rate": 5.406018947377668e-06,
+      "loss": 0.2656,
+      "step": 3155
+    },
+    {
+      "epoch": 1.0537562604340567,
+      "grad_norm": 0.5247989964050128,
+      "learning_rate": 5.403114762849034e-06,
+      "loss": 0.2547,
+      "step": 3156
+    },
+    {
+      "epoch": 1.0540901502504174,
+      "grad_norm": 0.48500108661200253,
+      "learning_rate": 5.400210441424707e-06,
+      "loss": 0.2456,
+      "step": 3157
+    },
+    {
+      "epoch": 1.054424040066778,
+      "grad_norm": 0.5053509107525429,
+      "learning_rate": 5.397305984090987e-06,
+      "loss": 0.2671,
+      "step": 3158
+    },
+    {
+      "epoch": 1.0547579298831387,
+      "grad_norm": 0.46349620064116775,
+      "learning_rate": 5.394401391834207e-06,
+      "loss": 0.2398,
+      "step": 3159
+    },
+    {
+      "epoch": 1.0550918196994992,
+      "grad_norm": 0.4897516537084441,
+      "learning_rate": 5.391496665640757e-06,
+      "loss": 0.2524,
+      "step": 3160
+    },
+    {
+      "epoch": 1.0554257095158597,
+      "grad_norm": 0.4609513408703894,
+      "learning_rate": 5.38859180649706e-06,
+      "loss": 0.2541,
+      "step": 3161
+    },
+    {
+      "epoch": 1.0557595993322204,
+      "grad_norm": 0.48531891805879707,
+      "learning_rate": 5.385686815389595e-06,
+      "loss": 0.2522,
+      "step": 3162
+    },
+    {
+      "epoch": 1.056093489148581,
+      "grad_norm": 0.47872455633974786,
+      "learning_rate": 5.382781693304883e-06,
+      "loss": 0.2557,
+      "step": 3163
+    },
+    {
+      "epoch": 1.0564273789649417,
+      "grad_norm": 0.49171820667226773,
+      "learning_rate": 5.379876441229486e-06,
+      "loss": 0.2578,
+      "step": 3164
+    },
+    {
+      "epoch": 1.0567612687813022,
+      "grad_norm": 0.48695749993719817,
+      "learning_rate": 5.376971060150015e-06,
+      "loss": 0.2464,
+      "step": 3165
+    },
+    {
+      "epoch": 1.0570951585976627,
+      "grad_norm": 0.47139021382557966,
+      "learning_rate": 5.374065551053118e-06,
+      "loss": 0.2554,
+      "step": 3166
+    },
+    {
+      "epoch": 1.0574290484140234,
+      "grad_norm": 0.4576594271133858,
+      "learning_rate": 5.3711599149254966e-06,
+      "loss": 0.2584,
+      "step": 3167
+    },
+    {
+      "epoch": 1.057762938230384,
+      "grad_norm": 0.4801915597479104,
+      "learning_rate": 5.368254152753887e-06,
+      "loss": 0.2639,
+      "step": 3168
+    },
+    {
+      "epoch": 1.0580968280467447,
+      "grad_norm": 0.5067203343177439,
+      "learning_rate": 5.365348265525071e-06,
+      "loss": 0.2765,
+      "step": 3169
+    },
+    {
+      "epoch": 1.0584307178631052,
+      "grad_norm": 0.48993539578975254,
+      "learning_rate": 5.362442254225872e-06,
+      "loss": 0.2575,
+      "step": 3170
+    },
+    {
+      "epoch": 1.0587646076794657,
+      "grad_norm": 0.46878470787169646,
+      "learning_rate": 5.359536119843158e-06,
+      "loss": 0.25,
+      "step": 3171
+    },
+    {
+      "epoch": 1.0590984974958264,
+      "grad_norm": 0.4778336832582392,
+      "learning_rate": 5.356629863363838e-06,
+      "loss": 0.2565,
+      "step": 3172
+    },
+    {
+      "epoch": 1.059432387312187,
+      "grad_norm": 0.45945430098596857,
+      "learning_rate": 5.353723485774859e-06,
+      "loss": 0.2576,
+      "step": 3173
+    },
+    {
+      "epoch": 1.0597662771285477,
+      "grad_norm": 0.47535593434261963,
+      "learning_rate": 5.350816988063212e-06,
+      "loss": 0.2583,
+      "step": 3174
+    },
+    {
+      "epoch": 1.0601001669449082,
+      "grad_norm": 0.4649627148435848,
+      "learning_rate": 5.34791037121593e-06,
+      "loss": 0.2572,
+      "step": 3175
+    },
+    {
+      "epoch": 1.060434056761269,
+      "grad_norm": 0.4318821326726218,
+      "learning_rate": 5.345003636220088e-06,
+      "loss": 0.254,
+      "step": 3176
+    },
+    {
+      "epoch": 1.0607679465776294,
+      "grad_norm": 0.4908570811371904,
+      "learning_rate": 5.342096784062792e-06,
+      "loss": 0.2554,
+      "step": 3177
+    },
+    {
+      "epoch": 1.06110183639399,
+      "grad_norm": 0.4709757995990221,
+      "learning_rate": 5.339189815731196e-06,
+      "loss": 0.2589,
+      "step": 3178
+    },
+    {
+      "epoch": 1.0614357262103506,
+      "grad_norm": 0.4457516758192152,
+      "learning_rate": 5.336282732212494e-06,
+      "loss": 0.2508,
+      "step": 3179
+    },
+    {
+      "epoch": 1.0617696160267112,
+      "grad_norm": 0.4484170223112844,
+      "learning_rate": 5.333375534493915e-06,
+      "loss": 0.2495,
+      "step": 3180
+    },
+    {
+      "epoch": 1.062103505843072,
+      "grad_norm": 0.48031995621187984,
+      "learning_rate": 5.330468223562728e-06,
+      "loss": 0.2526,
+      "step": 3181
+    },
+    {
+      "epoch": 1.0624373956594324,
+      "grad_norm": 0.4671110520540753,
+      "learning_rate": 5.327560800406241e-06,
+      "loss": 0.2509,
+      "step": 3182
+    },
+    {
+      "epoch": 1.062771285475793,
+      "grad_norm": 0.48463778791324896,
+      "learning_rate": 5.324653266011799e-06,
+      "loss": 0.2584,
+      "step": 3183
+    },
+    {
+      "epoch": 1.0631051752921536,
+      "grad_norm": 0.45448510515914453,
+      "learning_rate": 5.321745621366786e-06,
+      "loss": 0.2528,
+      "step": 3184
+    },
+    {
+      "epoch": 1.0634390651085142,
+      "grad_norm": 0.5062361770532562,
+      "learning_rate": 5.318837867458626e-06,
+      "loss": 0.2594,
+      "step": 3185
+    },
+    {
+      "epoch": 1.063772954924875,
+      "grad_norm": 0.4853077554076567,
+      "learning_rate": 5.315930005274773e-06,
+      "loss": 0.259,
+      "step": 3186
+    },
+    {
+      "epoch": 1.0641068447412354,
+      "grad_norm": 0.4996892203754061,
+      "learning_rate": 5.313022035802723e-06,
+      "loss": 0.2654,
+      "step": 3187
+    },
+    {
+      "epoch": 1.064440734557596,
+      "grad_norm": 0.4669729141550617,
+      "learning_rate": 5.31011396003001e-06,
+      "loss": 0.26,
+      "step": 3188
+    },
+    {
+      "epoch": 1.0647746243739566,
+      "grad_norm": 0.4860963507306269,
+      "learning_rate": 5.307205778944198e-06,
+      "loss": 0.2645,
+      "step": 3189
+    },
+    {
+      "epoch": 1.0651085141903172,
+      "grad_norm": 0.4911443221550441,
+      "learning_rate": 5.3042974935328925e-06,
+      "loss": 0.2601,
+      "step": 3190
+    },
+    {
+      "epoch": 1.065442404006678,
+      "grad_norm": 0.478917287029953,
+      "learning_rate": 5.301389104783731e-06,
+      "loss": 0.2576,
+      "step": 3191
+    },
+    {
+      "epoch": 1.0657762938230384,
+      "grad_norm": 0.48041309962561946,
+      "learning_rate": 5.298480613684388e-06,
+      "loss": 0.2607,
+      "step": 3192
+    },
+    {
+      "epoch": 1.066110183639399,
+      "grad_norm": 0.4729097428466752,
+      "learning_rate": 5.295572021222573e-06,
+      "loss": 0.2464,
+      "step": 3193
+    },
+    {
+      "epoch": 1.0664440734557596,
+      "grad_norm": 0.4577785299743451,
+      "learning_rate": 5.292663328386028e-06,
+      "loss": 0.2496,
+      "step": 3194
+    },
+    {
+      "epoch": 1.0667779632721202,
+      "grad_norm": 0.495644342327035,
+      "learning_rate": 5.289754536162527e-06,
+      "loss": 0.2508,
+      "step": 3195
+    },
+    {
+      "epoch": 1.0671118530884809,
+      "grad_norm": 0.47123980551693617,
+      "learning_rate": 5.2868456455398855e-06,
+      "loss": 0.2547,
+      "step": 3196
+    },
+    {
+      "epoch": 1.0674457429048414,
+      "grad_norm": 0.4609115825218123,
+      "learning_rate": 5.283936657505946e-06,
+      "loss": 0.2507,
+      "step": 3197
+    },
+    {
+      "epoch": 1.067779632721202,
+      "grad_norm": 0.47312177696366803,
+      "learning_rate": 5.281027573048585e-06,
+      "loss": 0.2547,
+      "step": 3198
+    },
+    {
+      "epoch": 1.0681135225375626,
+      "grad_norm": 0.462744767927931,
+      "learning_rate": 5.278118393155713e-06,
+      "loss": 0.2531,
+      "step": 3199
+    },
+    {
+      "epoch": 1.0684474123539232,
+      "grad_norm": 0.47646833356681934,
+      "learning_rate": 5.275209118815274e-06,
+      "loss": 0.2584,
+      "step": 3200
+    },
+    {
+      "epoch": 1.0687813021702839,
+      "grad_norm": 0.46868955810541424,
+      "learning_rate": 5.272299751015241e-06,
+      "loss": 0.2522,
+      "step": 3201
+    },
+    {
+      "epoch": 1.0691151919866444,
+      "grad_norm": 0.48035196817036424,
+      "learning_rate": 5.2693902907436204e-06,
+      "loss": 0.259,
+      "step": 3202
+    },
+    {
+      "epoch": 1.069449081803005,
+      "grad_norm": 0.47111718328317653,
+      "learning_rate": 5.266480738988449e-06,
+      "loss": 0.2568,
+      "step": 3203
+    },
+    {
+      "epoch": 1.0697829716193656,
+      "grad_norm": 0.463242496576953,
+      "learning_rate": 5.263571096737796e-06,
+      "loss": 0.2608,
+      "step": 3204
+    },
+    {
+      "epoch": 1.0701168614357262,
+      "grad_norm": 0.47114205466203557,
+      "learning_rate": 5.2606613649797625e-06,
+      "loss": 0.2591,
+      "step": 3205
+    },
+    {
+      "epoch": 1.0704507512520869,
+      "grad_norm": 0.4564457208620905,
+      "learning_rate": 5.257751544702478e-06,
+      "loss": 0.2483,
+      "step": 3206
+    },
+    {
+      "epoch": 1.0707846410684474,
+      "grad_norm": 0.4847153936186009,
+      "learning_rate": 5.254841636894099e-06,
+      "loss": 0.2678,
+      "step": 3207
+    },
+    {
+      "epoch": 1.071118530884808,
+      "grad_norm": 0.4649782440783902,
+      "learning_rate": 5.251931642542819e-06,
+      "loss": 0.2596,
+      "step": 3208
+    },
+    {
+      "epoch": 1.0714524207011686,
+      "grad_norm": 0.4750946124144292,
+      "learning_rate": 5.249021562636857e-06,
+      "loss": 0.2628,
+      "step": 3209
+    },
+    {
+      "epoch": 1.0717863105175292,
+      "grad_norm": 0.4658919071060521,
+      "learning_rate": 5.246111398164459e-06,
+      "loss": 0.2606,
+      "step": 3210
+    },
+    {
+      "epoch": 1.0721202003338899,
+      "grad_norm": 0.4827757123942986,
+      "learning_rate": 5.243201150113902e-06,
+      "loss": 0.2603,
+      "step": 3211
+    },
+    {
+      "epoch": 1.0724540901502504,
+      "grad_norm": 0.4995393180674006,
+      "learning_rate": 5.240290819473492e-06,
+      "loss": 0.2526,
+      "step": 3212
+    },
+    {
+      "epoch": 1.072787979966611,
+      "grad_norm": 0.49062207351224135,
+      "learning_rate": 5.2373804072315625e-06,
+      "loss": 0.2746,
+      "step": 3213
+    },
+    {
+      "epoch": 1.0731218697829716,
+      "grad_norm": 0.4883060534998393,
+      "learning_rate": 5.2344699143764735e-06,
+      "loss": 0.2625,
+      "step": 3214
+    },
+    {
+      "epoch": 1.0734557595993321,
+      "grad_norm": 0.47085505949054923,
+      "learning_rate": 5.231559341896615e-06,
+      "loss": 0.2461,
+      "step": 3215
+    },
+    {
+      "epoch": 1.0737896494156929,
+      "grad_norm": 0.4757183458588549,
+      "learning_rate": 5.2286486907804e-06,
+      "loss": 0.2445,
+      "step": 3216
+    },
+    {
+      "epoch": 1.0741235392320534,
+      "grad_norm": 0.4709884837750735,
+      "learning_rate": 5.225737962016273e-06,
+      "loss": 0.2565,
+      "step": 3217
+    },
+    {
+      "epoch": 1.0744574290484141,
+      "grad_norm": 0.4680083953914647,
+      "learning_rate": 5.222827156592701e-06,
+      "loss": 0.2596,
+      "step": 3218
+    },
+    {
+      "epoch": 1.0747913188647746,
+      "grad_norm": 0.48952918063166884,
+      "learning_rate": 5.219916275498178e-06,
+      "loss": 0.249,
+      "step": 3219
+    },
+    {
+      "epoch": 1.0751252086811351,
+      "grad_norm": 0.49448266166554006,
+      "learning_rate": 5.217005319721225e-06,
+      "loss": 0.2657,
+      "step": 3220
+    },
+    {
+      "epoch": 1.0754590984974959,
+      "grad_norm": 0.4800844137057129,
+      "learning_rate": 5.214094290250386e-06,
+      "loss": 0.2627,
+      "step": 3221
+    },
+    {
+      "epoch": 1.0757929883138564,
+      "grad_norm": 0.4698310557838573,
+      "learning_rate": 5.211183188074237e-06,
+      "loss": 0.2472,
+      "step": 3222
+    },
+    {
+      "epoch": 1.0761268781302171,
+      "grad_norm": 0.4774972200790891,
+      "learning_rate": 5.208272014181366e-06,
+      "loss": 0.2607,
+      "step": 3223
+    },
+    {
+      "epoch": 1.0764607679465776,
+      "grad_norm": 0.499522338710151,
+      "learning_rate": 5.205360769560398e-06,
+      "loss": 0.2691,
+      "step": 3224
+    },
+    {
+      "epoch": 1.0767946577629381,
+      "grad_norm": 0.49745842673627644,
+      "learning_rate": 5.2024494551999725e-06,
+      "loss": 0.2591,
+      "step": 3225
+    },
+    {
+      "epoch": 1.0771285475792989,
+      "grad_norm": 0.47782783742912704,
+      "learning_rate": 5.199538072088759e-06,
+      "loss": 0.2648,
+      "step": 3226
+    },
+    {
+      "epoch": 1.0774624373956594,
+      "grad_norm": 0.4978686672391908,
+      "learning_rate": 5.196626621215449e-06,
+      "loss": 0.2535,
+      "step": 3227
+    },
+    {
+      "epoch": 1.0777963272120201,
+      "grad_norm": 0.4708559646310978,
+      "learning_rate": 5.193715103568754e-06,
+      "loss": 0.2517,
+      "step": 3228
+    },
+    {
+      "epoch": 1.0781302170283806,
+      "grad_norm": 0.4884200532727141,
+      "learning_rate": 5.190803520137412e-06,
+      "loss": 0.2636,
+      "step": 3229
+    },
+    {
+      "epoch": 1.0784641068447411,
+      "grad_norm": 0.47216324322555403,
+      "learning_rate": 5.18789187191018e-06,
+      "loss": 0.2566,
+      "step": 3230
+    },
+    {
+      "epoch": 1.0787979966611019,
+      "grad_norm": 0.4767306818111583,
+      "learning_rate": 5.184980159875841e-06,
+      "loss": 0.2526,
+      "step": 3231
+    },
+    {
+      "epoch": 1.0791318864774624,
+      "grad_norm": 0.4637406341729961,
+      "learning_rate": 5.182068385023196e-06,
+      "loss": 0.2447,
+      "step": 3232
+    },
+    {
+      "epoch": 1.0794657762938231,
+      "grad_norm": 0.48196123553887266,
+      "learning_rate": 5.179156548341068e-06,
+      "loss": 0.2616,
+      "step": 3233
+    },
+    {
+      "epoch": 1.0797996661101836,
+      "grad_norm": 0.4913195395698987,
+      "learning_rate": 5.176244650818304e-06,
+      "loss": 0.2598,
+      "step": 3234
+    },
+    {
+      "epoch": 1.0801335559265441,
+      "grad_norm": 0.47448932775354974,
+      "learning_rate": 5.1733326934437665e-06,
+      "loss": 0.2542,
+      "step": 3235
+    },
+    {
+      "epoch": 1.0804674457429049,
+      "grad_norm": 0.4482171683249098,
+      "learning_rate": 5.170420677206344e-06,
+      "loss": 0.2428,
+      "step": 3236
+    },
+    {
+      "epoch": 1.0808013355592654,
+      "grad_norm": 0.4630272127220971,
+      "learning_rate": 5.167508603094938e-06,
+      "loss": 0.2552,
+      "step": 3237
+    },
+    {
+      "epoch": 1.0811352253756261,
+      "grad_norm": 0.5010215353407538,
+      "learning_rate": 5.164596472098479e-06,
+      "loss": 0.2517,
+      "step": 3238
+    },
+    {
+      "epoch": 1.0814691151919866,
+      "grad_norm": 0.48744622660452824,
+      "learning_rate": 5.161684285205907e-06,
+      "loss": 0.2527,
+      "step": 3239
+    },
+    {
+      "epoch": 1.0818030050083474,
+      "grad_norm": 0.4729692946711275,
+      "learning_rate": 5.1587720434061895e-06,
+      "loss": 0.2391,
+      "step": 3240
+    },
+    {
+      "epoch": 1.0821368948247079,
+      "grad_norm": 0.47410617737202426,
+      "learning_rate": 5.155859747688306e-06,
+      "loss": 0.2562,
+      "step": 3241
+    },
+    {
+      "epoch": 1.0824707846410684,
+      "grad_norm": 0.5143791612768249,
+      "learning_rate": 5.152947399041258e-06,
+      "loss": 0.2576,
+      "step": 3242
+    },
+    {
+      "epoch": 1.0828046744574291,
+      "grad_norm": 0.4791286438666753,
+      "learning_rate": 5.1500349984540655e-06,
+      "loss": 0.2522,
+      "step": 3243
+    },
+    {
+      "epoch": 1.0831385642737896,
+      "grad_norm": 0.49196821028449766,
+      "learning_rate": 5.147122546915762e-06,
+      "loss": 0.2647,
+      "step": 3244
+    },
+    {
+      "epoch": 1.0834724540901504,
+      "grad_norm": 0.5018917704402174,
+      "learning_rate": 5.144210045415402e-06,
+      "loss": 0.2645,
+      "step": 3245
+    },
+    {
+      "epoch": 1.0838063439065109,
+      "grad_norm": 0.47478162571462834,
+      "learning_rate": 5.1412974949420555e-06,
+      "loss": 0.2541,
+      "step": 3246
+    },
+    {
+      "epoch": 1.0841402337228714,
+      "grad_norm": 0.4879345843118869,
+      "learning_rate": 5.138384896484813e-06,
+      "loss": 0.2619,
+      "step": 3247
+    },
+    {
+      "epoch": 1.0844741235392321,
+      "grad_norm": 0.5057389348093271,
+      "learning_rate": 5.135472251032772e-06,
+      "loss": 0.2527,
+      "step": 3248
+    },
+    {
+      "epoch": 1.0848080133555926,
+      "grad_norm": 0.47194679299707865,
+      "learning_rate": 5.132559559575058e-06,
+      "loss": 0.2525,
+      "step": 3249
+    },
+    {
+      "epoch": 1.0851419031719534,
+      "grad_norm": 0.5279650067996341,
+      "learning_rate": 5.129646823100801e-06,
+      "loss": 0.2745,
+      "step": 3250
+    },
+    {
+      "epoch": 1.0854757929883139,
+      "grad_norm": 0.48147895015650954,
+      "learning_rate": 5.126734042599153e-06,
+      "loss": 0.2617,
+      "step": 3251
+    },
+    {
+      "epoch": 1.0858096828046744,
+      "grad_norm": 0.46719185570759836,
+      "learning_rate": 5.12382121905928e-06,
+      "loss": 0.245,
+      "step": 3252
+    },
+    {
+      "epoch": 1.0861435726210351,
+      "grad_norm": 0.5039843837711871,
+      "learning_rate": 5.120908353470362e-06,
+      "loss": 0.2676,
+      "step": 3253
+    },
+    {
+      "epoch": 1.0864774624373956,
+      "grad_norm": 0.4920095296892038,
+      "learning_rate": 5.117995446821592e-06,
+      "loss": 0.2612,
+      "step": 3254
+    },
+    {
+      "epoch": 1.0868113522537564,
+      "grad_norm": 0.4955345312022653,
+      "learning_rate": 5.115082500102177e-06,
+      "loss": 0.256,
+      "step": 3255
+    },
+    {
+      "epoch": 1.0871452420701169,
+      "grad_norm": 0.4877180712143581,
+      "learning_rate": 5.112169514301341e-06,
+      "loss": 0.247,
+      "step": 3256
+    },
+    {
+      "epoch": 1.0874791318864774,
+      "grad_norm": 0.4655734883041201,
+      "learning_rate": 5.109256490408317e-06,
+      "loss": 0.251,
+      "step": 3257
+    },
+    {
+      "epoch": 1.0878130217028381,
+      "grad_norm": 0.4766659749608424,
+      "learning_rate": 5.1063434294123556e-06,
+      "loss": 0.2544,
+      "step": 3258
+    },
+    {
+      "epoch": 1.0881469115191986,
+      "grad_norm": 0.4921274900903924,
+      "learning_rate": 5.103430332302715e-06,
+      "loss": 0.2582,
+      "step": 3259
+    },
+    {
+      "epoch": 1.0884808013355594,
+      "grad_norm": 0.49570682606017635,
+      "learning_rate": 5.1005172000686675e-06,
+      "loss": 0.2475,
+      "step": 3260
+    },
+    {
+      "epoch": 1.0888146911519199,
+      "grad_norm": 0.4795494679085192,
+      "learning_rate": 5.097604033699499e-06,
+      "loss": 0.2605,
+      "step": 3261
+    },
+    {
+      "epoch": 1.0891485809682804,
+      "grad_norm": 0.4694218152948914,
+      "learning_rate": 5.094690834184505e-06,
+      "loss": 0.2507,
+      "step": 3262
+    },
+    {
+      "epoch": 1.089482470784641,
+      "grad_norm": 0.4510949970831343,
+      "learning_rate": 5.091777602512993e-06,
+      "loss": 0.2474,
+      "step": 3263
+    },
+    {
+      "epoch": 1.0898163606010016,
+      "grad_norm": 0.4592770588757295,
+      "learning_rate": 5.088864339674282e-06,
+      "loss": 0.2514,
+      "step": 3264
+    },
+    {
+      "epoch": 1.0901502504173624,
+      "grad_norm": 0.46403097850278985,
+      "learning_rate": 5.085951046657701e-06,
+      "loss": 0.2485,
+      "step": 3265
+    },
+    {
+      "epoch": 1.0904841402337229,
+      "grad_norm": 0.4743912965286824,
+      "learning_rate": 5.083037724452589e-06,
+      "loss": 0.2541,
+      "step": 3266
+    },
+    {
+      "epoch": 1.0908180300500834,
+      "grad_norm": 0.4905672325919222,
+      "learning_rate": 5.080124374048294e-06,
+      "loss": 0.2532,
+      "step": 3267
+    },
+    {
+      "epoch": 1.091151919866444,
+      "grad_norm": 0.4978884822194222,
+      "learning_rate": 5.077210996434177e-06,
+      "loss": 0.2718,
+      "step": 3268
+    },
+    {
+      "epoch": 1.0914858096828046,
+      "grad_norm": 0.4913228834438439,
+      "learning_rate": 5.074297592599604e-06,
+      "loss": 0.258,
+      "step": 3269
+    },
+    {
+      "epoch": 1.0918196994991654,
+      "grad_norm": 0.4758515456803783,
+      "learning_rate": 5.071384163533955e-06,
+      "loss": 0.2567,
+      "step": 3270
+    },
+    {
+      "epoch": 1.0921535893155259,
+      "grad_norm": 0.47254772969527864,
+      "learning_rate": 5.068470710226611e-06,
+      "loss": 0.2594,
+      "step": 3271
+    },
+    {
+      "epoch": 1.0924874791318864,
+      "grad_norm": 0.4861709846708059,
+      "learning_rate": 5.065557233666968e-06,
+      "loss": 0.2566,
+      "step": 3272
+    },
+    {
+      "epoch": 1.092821368948247,
+      "grad_norm": 0.4779148959627853,
+      "learning_rate": 5.062643734844427e-06,
+      "loss": 0.27,
+      "step": 3273
+    },
+    {
+      "epoch": 1.0931552587646076,
+      "grad_norm": 0.48531078794017424,
+      "learning_rate": 5.059730214748398e-06,
+      "loss": 0.2605,
+      "step": 3274
+    },
+    {
+      "epoch": 1.0934891485809684,
+      "grad_norm": 0.5013099712212802,
+      "learning_rate": 5.0568166743682964e-06,
+      "loss": 0.2623,
+      "step": 3275
+    },
+    {
+      "epoch": 1.0938230383973289,
+      "grad_norm": 0.4669743515657304,
+      "learning_rate": 5.053903114693547e-06,
+      "loss": 0.264,
+      "step": 3276
+    },
+    {
+      "epoch": 1.0941569282136894,
+      "grad_norm": 0.4716612249389492,
+      "learning_rate": 5.050989536713578e-06,
+      "loss": 0.2574,
+      "step": 3277
+    },
+    {
+      "epoch": 1.09449081803005,
+      "grad_norm": 0.4859150159213844,
+      "learning_rate": 5.048075941417827e-06,
+      "loss": 0.2485,
+      "step": 3278
+    },
+    {
+      "epoch": 1.0948247078464106,
+      "grad_norm": 0.482211469619731,
+      "learning_rate": 5.0451623297957345e-06,
+      "loss": 0.2528,
+      "step": 3279
+    },
+    {
+      "epoch": 1.0951585976627713,
+      "grad_norm": 0.5086632888574513,
+      "learning_rate": 5.042248702836746e-06,
+      "loss": 0.2679,
+      "step": 3280
+    },
+    {
+      "epoch": 1.0954924874791319,
+      "grad_norm": 0.4555537402005905,
+      "learning_rate": 5.039335061530319e-06,
+      "loss": 0.2525,
+      "step": 3281
+    },
+    {
+      "epoch": 1.0958263772954924,
+      "grad_norm": 0.4671370686668756,
+      "learning_rate": 5.036421406865908e-06,
+      "loss": 0.2551,
+      "step": 3282
+    },
+    {
+      "epoch": 1.096160267111853,
+      "grad_norm": 0.5195036249906094,
+      "learning_rate": 5.033507739832976e-06,
+      "loss": 0.255,
+      "step": 3283
+    },
+    {
+      "epoch": 1.0964941569282136,
+      "grad_norm": 0.5099391387948038,
+      "learning_rate": 5.030594061420987e-06,
+      "loss": 0.2644,
+      "step": 3284
+    },
+    {
+      "epoch": 1.0968280467445743,
+      "grad_norm": 0.48013654911099934,
+      "learning_rate": 5.027680372619414e-06,
+      "loss": 0.2518,
+      "step": 3285
+    },
+    {
+      "epoch": 1.0971619365609349,
+      "grad_norm": 0.46990133782477084,
+      "learning_rate": 5.02476667441773e-06,
+      "loss": 0.2558,
+      "step": 3286
+    },
+    {
+      "epoch": 1.0974958263772956,
+      "grad_norm": 0.475272895666666,
+      "learning_rate": 5.021852967805411e-06,
+      "loss": 0.2648,
+      "step": 3287
+    },
+    {
+      "epoch": 1.097829716193656,
+      "grad_norm": 0.4968860355795686,
+      "learning_rate": 5.018939253771937e-06,
+      "loss": 0.2555,
+      "step": 3288
+    },
+    {
+      "epoch": 1.0981636060100166,
+      "grad_norm": 0.4718478749886261,
+      "learning_rate": 5.0160255333067895e-06,
+      "loss": 0.2502,
+      "step": 3289
+    },
+    {
+      "epoch": 1.0984974958263773,
+      "grad_norm": 0.47935762166392604,
+      "learning_rate": 5.013111807399455e-06,
+      "loss": 0.2469,
+      "step": 3290
+    },
+    {
+      "epoch": 1.0988313856427379,
+      "grad_norm": 0.4742963803359443,
+      "learning_rate": 5.010198077039418e-06,
+      "loss": 0.2421,
+      "step": 3291
+    },
+    {
+      "epoch": 1.0991652754590986,
+      "grad_norm": 0.4739460911624742,
+      "learning_rate": 5.007284343216168e-06,
+      "loss": 0.2667,
+      "step": 3292
+    },
+    {
+      "epoch": 1.099499165275459,
+      "grad_norm": 0.48790157404547646,
+      "learning_rate": 5.004370606919191e-06,
+      "loss": 0.271,
+      "step": 3293
+    },
+    {
+      "epoch": 1.0998330550918196,
+      "grad_norm": 0.5131327733929237,
+      "learning_rate": 5.001456869137979e-06,
+      "loss": 0.262,
+      "step": 3294
+    },
+    {
+      "epoch": 1.1001669449081803,
+      "grad_norm": 0.469965233108866,
+      "learning_rate": 4.998543130862021e-06,
+      "loss": 0.2604,
+      "step": 3295
+    },
+    {
+      "epoch": 1.1005008347245409,
+      "grad_norm": 0.4818780628265589,
+      "learning_rate": 4.995629393080809e-06,
+      "loss": 0.2724,
+      "step": 3296
+    },
+    {
+      "epoch": 1.1008347245409016,
+      "grad_norm": 0.48604608669868893,
+      "learning_rate": 4.992715656783834e-06,
+      "loss": 0.2519,
+      "step": 3297
+    },
+    {
+      "epoch": 1.101168614357262,
+      "grad_norm": 0.5004678633035026,
+      "learning_rate": 4.989801922960583e-06,
+      "loss": 0.2655,
+      "step": 3298
+    },
+    {
+      "epoch": 1.1015025041736226,
+      "grad_norm": 0.45841645499248523,
+      "learning_rate": 4.986888192600546e-06,
+      "loss": 0.2564,
+      "step": 3299
+    },
+    {
+      "epoch": 1.1018363939899833,
+      "grad_norm": 0.4747327230533823,
+      "learning_rate": 4.983974466693212e-06,
+      "loss": 0.2611,
+      "step": 3300
+    },
+    {
+      "epoch": 1.1021702838063439,
+      "grad_norm": 0.4648805828123176,
+      "learning_rate": 4.9810607462280655e-06,
+      "loss": 0.2524,
+      "step": 3301
+    },
+    {
+      "epoch": 1.1025041736227046,
+      "grad_norm": 0.48765616081849056,
+      "learning_rate": 4.978147032194592e-06,
+      "loss": 0.2581,
+      "step": 3302
+    },
+    {
+      "epoch": 1.102838063439065,
+      "grad_norm": 0.511045364615192,
+      "learning_rate": 4.975233325582271e-06,
+      "loss": 0.2592,
+      "step": 3303
+    },
+    {
+      "epoch": 1.1031719532554256,
+      "grad_norm": 0.4949170773665287,
+      "learning_rate": 4.972319627380587e-06,
+      "loss": 0.2536,
+      "step": 3304
+    },
+    {
+      "epoch": 1.1035058430717863,
+      "grad_norm": 0.48002432073864537,
+      "learning_rate": 4.969405938579014e-06,
+      "loss": 0.2527,
+      "step": 3305
+    },
+    {
+      "epoch": 1.1038397328881469,
+      "grad_norm": 0.4907633242918611,
+      "learning_rate": 4.9664922601670265e-06,
+      "loss": 0.2705,
+      "step": 3306
+    },
+    {
+      "epoch": 1.1041736227045076,
+      "grad_norm": 0.48425062215651077,
+      "learning_rate": 4.963578593134094e-06,
+      "loss": 0.2627,
+      "step": 3307
+    },
+    {
+      "epoch": 1.104507512520868,
+      "grad_norm": 0.491141839891003,
+      "learning_rate": 4.960664938469683e-06,
+      "loss": 0.2679,
+      "step": 3308
+    },
+    {
+      "epoch": 1.1048414023372288,
+      "grad_norm": 0.456696689579314,
+      "learning_rate": 4.9577512971632555e-06,
+      "loss": 0.2451,
+      "step": 3309
+    },
+    {
+      "epoch": 1.1051752921535893,
+      "grad_norm": 0.501160536107327,
+      "learning_rate": 4.954837670204268e-06,
+      "loss": 0.261,
+      "step": 3310
+    },
+    {
+      "epoch": 1.1055091819699499,
+      "grad_norm": 0.46990903591222016,
+      "learning_rate": 4.951924058582176e-06,
+      "loss": 0.2515,
+      "step": 3311
+    },
+    {
+      "epoch": 1.1058430717863106,
+      "grad_norm": 0.5044218948761715,
+      "learning_rate": 4.949010463286422e-06,
+      "loss": 0.2629,
+      "step": 3312
+    },
+    {
+      "epoch": 1.106176961602671,
+      "grad_norm": 0.5062603471723732,
+      "learning_rate": 4.9460968853064535e-06,
+      "loss": 0.2639,
+      "step": 3313
+    },
+    {
+      "epoch": 1.1065108514190318,
+      "grad_norm": 0.4657580426684371,
+      "learning_rate": 4.943183325631704e-06,
+      "loss": 0.248,
+      "step": 3314
+    },
+    {
+      "epoch": 1.1068447412353923,
+      "grad_norm": 0.48098469377408215,
+      "learning_rate": 4.940269785251603e-06,
+      "loss": 0.2468,
+      "step": 3315
+    },
+    {
+      "epoch": 1.1071786310517528,
+      "grad_norm": 0.4682162153731721,
+      "learning_rate": 4.937356265155575e-06,
+      "loss": 0.2495,
+      "step": 3316
+    },
+    {
+      "epoch": 1.1075125208681136,
+      "grad_norm": 0.47436733593248315,
+      "learning_rate": 4.934442766333034e-06,
+      "loss": 0.2549,
+      "step": 3317
+    },
+    {
+      "epoch": 1.107846410684474,
+      "grad_norm": 0.48978585222070464,
+      "learning_rate": 4.931529289773392e-06,
+      "loss": 0.2655,
+      "step": 3318
+    },
+    {
+      "epoch": 1.1081803005008348,
+      "grad_norm": 0.5176613105080745,
+      "learning_rate": 4.928615836466048e-06,
+      "loss": 0.2643,
+      "step": 3319
+    },
+    {
+      "epoch": 1.1085141903171953,
+      "grad_norm": 0.4914604738925494,
+      "learning_rate": 4.925702407400396e-06,
+      "loss": 0.2467,
+      "step": 3320
+    },
+    {
+      "epoch": 1.1088480801335558,
+      "grad_norm": 0.47177783550530394,
+      "learning_rate": 4.922789003565824e-06,
+      "loss": 0.2559,
+      "step": 3321
+    },
+    {
+      "epoch": 1.1091819699499166,
+      "grad_norm": 0.4608549823117533,
+      "learning_rate": 4.919875625951707e-06,
+      "loss": 0.2565,
+      "step": 3322
+    },
+    {
+      "epoch": 1.109515859766277,
+      "grad_norm": 0.47219276241856967,
+      "learning_rate": 4.916962275547413e-06,
+      "loss": 0.2536,
+      "step": 3323
+    },
+    {
+      "epoch": 1.1098497495826378,
+      "grad_norm": 0.5237543763083652,
+      "learning_rate": 4.9140489533423e-06,
+      "loss": 0.2725,
+      "step": 3324
+    },
+    {
+      "epoch": 1.1101836393989983,
+      "grad_norm": 0.48356718681869415,
+      "learning_rate": 4.91113566032572e-06,
+      "loss": 0.2459,
+      "step": 3325
+    },
+    {
+      "epoch": 1.1105175292153588,
+      "grad_norm": 0.4905924238533504,
+      "learning_rate": 4.908222397487009e-06,
+      "loss": 0.2656,
+      "step": 3326
+    },
+    {
+      "epoch": 1.1108514190317196,
+      "grad_norm": 0.48500942984290396,
+      "learning_rate": 4.9053091658154975e-06,
+      "loss": 0.2592,
+      "step": 3327
+    },
+    {
+      "epoch": 1.11118530884808,
+      "grad_norm": 0.5069229306242481,
+      "learning_rate": 4.902395966300503e-06,
+      "loss": 0.2539,
+      "step": 3328
+    },
+    {
+      "epoch": 1.1115191986644408,
+      "grad_norm": 0.459483097526595,
+      "learning_rate": 4.899482799931333e-06,
+      "loss": 0.2475,
+      "step": 3329
+    },
+    {
+      "epoch": 1.1118530884808013,
+      "grad_norm": 0.5026420485624541,
+      "learning_rate": 4.896569667697287e-06,
+      "loss": 0.2639,
+      "step": 3330
+    },
+    {
+      "epoch": 1.1121869782971618,
+      "grad_norm": 0.4664149895902408,
+      "learning_rate": 4.893656570587645e-06,
+      "loss": 0.26,
+      "step": 3331
+    },
+    {
+      "epoch": 1.1125208681135226,
+      "grad_norm": 0.4596419459034059,
+      "learning_rate": 4.890743509591684e-06,
+      "loss": 0.255,
+      "step": 3332
+    },
+    {
+      "epoch": 1.112854757929883,
+      "grad_norm": 0.4810477843283111,
+      "learning_rate": 4.887830485698659e-06,
+      "loss": 0.253,
+      "step": 3333
+    },
+    {
+      "epoch": 1.1131886477462438,
+      "grad_norm": 0.4901040145369989,
+      "learning_rate": 4.884917499897824e-06,
+      "loss": 0.2665,
+      "step": 3334
+    },
+    {
+      "epoch": 1.1135225375626043,
+      "grad_norm": 0.45885422690084604,
+      "learning_rate": 4.88200455317841e-06,
+      "loss": 0.2611,
+      "step": 3335
+    },
+    {
+      "epoch": 1.1138564273789648,
+      "grad_norm": 0.480785703396849,
+      "learning_rate": 4.87909164652964e-06,
+      "loss": 0.2655,
+      "step": 3336
+    },
+    {
+      "epoch": 1.1141903171953256,
+      "grad_norm": 0.4910501978399561,
+      "learning_rate": 4.87617878094072e-06,
+      "loss": 0.2628,
+      "step": 3337
+    },
+    {
+      "epoch": 1.114524207011686,
+      "grad_norm": 0.48593955143283746,
+      "learning_rate": 4.873265957400847e-06,
+      "loss": 0.2527,
+      "step": 3338
+    },
+    {
+      "epoch": 1.1148580968280468,
+      "grad_norm": 0.4786314037051106,
+      "learning_rate": 4.8703531768992e-06,
+      "loss": 0.2568,
+      "step": 3339
+    },
+    {
+      "epoch": 1.1151919866444073,
+      "grad_norm": 0.46409758075806445,
+      "learning_rate": 4.867440440424944e-06,
+      "loss": 0.2469,
+      "step": 3340
+    },
+    {
+      "epoch": 1.1155258764607678,
+      "grad_norm": 0.4578210978494995,
+      "learning_rate": 4.864527748967229e-06,
+      "loss": 0.2604,
+      "step": 3341
+    },
+    {
+      "epoch": 1.1158597662771286,
+      "grad_norm": 0.48124665569684744,
+      "learning_rate": 4.86161510351519e-06,
+      "loss": 0.2587,
+      "step": 3342
+    },
+    {
+      "epoch": 1.116193656093489,
+      "grad_norm": 0.493094574153891,
+      "learning_rate": 4.858702505057945e-06,
+      "loss": 0.2545,
+      "step": 3343
+    },
+    {
+      "epoch": 1.1165275459098498,
+      "grad_norm": 0.47867822797754944,
+      "learning_rate": 4.8557899545846e-06,
+      "loss": 0.247,
+      "step": 3344
+    },
+    {
+      "epoch": 1.1168614357262103,
+      "grad_norm": 0.4822455074382586,
+      "learning_rate": 4.852877453084241e-06,
+      "loss": 0.2639,
+      "step": 3345
+    },
+    {
+      "epoch": 1.1171953255425708,
+      "grad_norm": 0.4765824815109976,
+      "learning_rate": 4.849965001545936e-06,
+      "loss": 0.2626,
+      "step": 3346
+    },
+    {
+      "epoch": 1.1175292153589316,
+      "grad_norm": 0.4681179654002761,
+      "learning_rate": 4.847052600958743e-06,
+      "loss": 0.2549,
+      "step": 3347
+    },
+    {
+      "epoch": 1.117863105175292,
+      "grad_norm": 0.4663663620409239,
+      "learning_rate": 4.844140252311695e-06,
+      "loss": 0.2426,
+      "step": 3348
+    },
+    {
+      "epoch": 1.1181969949916528,
+      "grad_norm": 0.4755229762246299,
+      "learning_rate": 4.841227956593812e-06,
+      "loss": 0.2516,
+      "step": 3349
+    },
+    {
+      "epoch": 1.1185308848080133,
+      "grad_norm": 0.4925523643410613,
+      "learning_rate": 4.838315714794094e-06,
+      "loss": 0.262,
+      "step": 3350
+    },
+    {
+      "epoch": 1.118864774624374,
+      "grad_norm": 0.49413475014094815,
+      "learning_rate": 4.8354035279015235e-06,
+      "loss": 0.2666,
+      "step": 3351
+    },
+    {
+      "epoch": 1.1191986644407346,
+      "grad_norm": 0.48697328572649823,
+      "learning_rate": 4.8324913969050645e-06,
+      "loss": 0.2512,
+      "step": 3352
+    },
+    {
+      "epoch": 1.119532554257095,
+      "grad_norm": 0.4643723432186017,
+      "learning_rate": 4.829579322793659e-06,
+      "loss": 0.2462,
+      "step": 3353
+    },
+    {
+      "epoch": 1.1198664440734558,
+      "grad_norm": 0.4850952566363859,
+      "learning_rate": 4.826667306556234e-06,
+      "loss": 0.2658,
+      "step": 3354
+    },
+    {
+      "epoch": 1.1202003338898163,
+      "grad_norm": 0.5007245806460484,
+      "learning_rate": 4.823755349181698e-06,
+      "loss": 0.2534,
+      "step": 3355
+    },
+    {
+      "epoch": 1.120534223706177,
+      "grad_norm": 0.50828805288149,
+      "learning_rate": 4.820843451658932e-06,
+      "loss": 0.276,
+      "step": 3356
+    },
+    {
+      "epoch": 1.1208681135225376,
+      "grad_norm": 0.4770967612220026,
+      "learning_rate": 4.817931614976806e-06,
+      "loss": 0.2499,
+      "step": 3357
+    },
+    {
+      "epoch": 1.121202003338898,
+      "grad_norm": 0.4620785185546986,
+      "learning_rate": 4.81501984012416e-06,
+      "loss": 0.2598,
+      "step": 3358
+    },
+    {
+      "epoch": 1.1215358931552588,
+      "grad_norm": 0.4670476176453921,
+      "learning_rate": 4.812108128089821e-06,
+      "loss": 0.2548,
+      "step": 3359
+    },
+    {
+      "epoch": 1.1218697829716193,
+      "grad_norm": 0.47632848287555607,
+      "learning_rate": 4.8091964798625905e-06,
+      "loss": 0.2567,
+      "step": 3360
+    },
+    {
+      "epoch": 1.12220367278798,
+      "grad_norm": 0.48055834067508496,
+      "learning_rate": 4.8062848964312485e-06,
+      "loss": 0.2613,
+      "step": 3361
+    },
+    {
+      "epoch": 1.1225375626043406,
+      "grad_norm": 0.4807161012160384,
+      "learning_rate": 4.8033733787845535e-06,
+      "loss": 0.2469,
+      "step": 3362
+    },
+    {
+      "epoch": 1.122871452420701,
+      "grad_norm": 0.48413032292121805,
+      "learning_rate": 4.800461927911241e-06,
+      "loss": 0.2599,
+      "step": 3363
+    },
+    {
+      "epoch": 1.1232053422370618,
+      "grad_norm": 0.46406736675585186,
+      "learning_rate": 4.797550544800029e-06,
+      "loss": 0.2588,
+      "step": 3364
+    },
+    {
+      "epoch": 1.1235392320534223,
+      "grad_norm": 0.4419629727027401,
+      "learning_rate": 4.794639230439604e-06,
+      "loss": 0.243,
+      "step": 3365
+    },
+    {
+      "epoch": 1.123873121869783,
+      "grad_norm": 0.5045405803679711,
+      "learning_rate": 4.791727985818636e-06,
+      "loss": 0.2518,
+      "step": 3366
+    },
+    {
+      "epoch": 1.1242070116861436,
+      "grad_norm": 0.49589152318810237,
+      "learning_rate": 4.788816811925765e-06,
+      "loss": 0.2542,
+      "step": 3367
+    },
+    {
+      "epoch": 1.124540901502504,
+      "grad_norm": 0.4876225715296857,
+      "learning_rate": 4.7859057097496145e-06,
+      "loss": 0.2487,
+      "step": 3368
+    },
+    {
+      "epoch": 1.1248747913188648,
+      "grad_norm": 0.4692616404123496,
+      "learning_rate": 4.782994680278776e-06,
+      "loss": 0.2489,
+      "step": 3369
+    },
+    {
+      "epoch": 1.1252086811352253,
+      "grad_norm": 0.47610298193563994,
+      "learning_rate": 4.780083724501825e-06,
+      "loss": 0.2554,
+      "step": 3370
+    },
+    {
+      "epoch": 1.125542570951586,
+      "grad_norm": 0.4805057684548899,
+      "learning_rate": 4.7771728434073005e-06,
+      "loss": 0.2614,
+      "step": 3371
+    },
+    {
+      "epoch": 1.1258764607679466,
+      "grad_norm": 0.47330605700559797,
+      "learning_rate": 4.7742620379837276e-06,
+      "loss": 0.2603,
+      "step": 3372
+    },
+    {
+      "epoch": 1.1262103505843073,
+      "grad_norm": 0.48550495154849865,
+      "learning_rate": 4.771351309219601e-06,
+      "loss": 0.2614,
+      "step": 3373
+    },
+    {
+      "epoch": 1.1265442404006678,
+      "grad_norm": 0.48653992147283703,
+      "learning_rate": 4.768440658103386e-06,
+      "loss": 0.2595,
+      "step": 3374
+    },
+    {
+      "epoch": 1.1268781302170283,
+      "grad_norm": 0.5202888246231431,
+      "learning_rate": 4.765530085623527e-06,
+      "loss": 0.2559,
+      "step": 3375
+    },
+    {
+      "epoch": 1.127212020033389,
+      "grad_norm": 0.47181833884933116,
+      "learning_rate": 4.762619592768439e-06,
+      "loss": 0.2517,
+      "step": 3376
+    },
+    {
+      "epoch": 1.1275459098497496,
+      "grad_norm": 0.4703284944592757,
+      "learning_rate": 4.75970918052651e-06,
+      "loss": 0.2456,
+      "step": 3377
+    },
+    {
+      "epoch": 1.1278797996661103,
+      "grad_norm": 0.4830649294835701,
+      "learning_rate": 4.7567988498861e-06,
+      "loss": 0.2641,
+      "step": 3378
+    },
+    {
+      "epoch": 1.1282136894824708,
+      "grad_norm": 0.4636866321661554,
+      "learning_rate": 4.753888601835544e-06,
+      "loss": 0.249,
+      "step": 3379
+    },
+    {
+      "epoch": 1.1285475792988313,
+      "grad_norm": 0.46296835910178785,
+      "learning_rate": 4.7509784373631446e-06,
+      "loss": 0.2559,
+      "step": 3380
+    },
+    {
+      "epoch": 1.128881469115192,
+      "grad_norm": 0.47001908089785516,
+      "learning_rate": 4.748068357457182e-06,
+      "loss": 0.2568,
+      "step": 3381
+    },
+    {
+      "epoch": 1.1292153589315526,
+      "grad_norm": 0.47751808080856645,
+      "learning_rate": 4.7451583631059026e-06,
+      "loss": 0.2589,
+      "step": 3382
+    },
+    {
+      "epoch": 1.1295492487479133,
+      "grad_norm": 0.4787271582822124,
+      "learning_rate": 4.742248455297524e-06,
+      "loss": 0.263,
+      "step": 3383
+    },
+    {
+      "epoch": 1.1298831385642738,
+      "grad_norm": 0.4923450872491639,
+      "learning_rate": 4.739338635020239e-06,
+      "loss": 0.2581,
+      "step": 3384
+    },
+    {
+      "epoch": 1.1302170283806343,
+      "grad_norm": 0.4872992680144528,
+      "learning_rate": 4.736428903262205e-06,
+      "loss": 0.2624,
+      "step": 3385
+    },
+    {
+      "epoch": 1.130550918196995,
+      "grad_norm": 0.4434160478725096,
+      "learning_rate": 4.7335192610115535e-06,
+      "loss": 0.2425,
+      "step": 3386
+    },
+    {
+      "epoch": 1.1308848080133556,
+      "grad_norm": 0.4588020589213422,
+      "learning_rate": 4.730609709256382e-06,
+      "loss": 0.2492,
+      "step": 3387
+    },
+    {
+      "epoch": 1.1312186978297163,
+      "grad_norm": 0.46923133080769724,
+      "learning_rate": 4.727700248984759e-06,
+      "loss": 0.2536,
+      "step": 3388
+    },
+    {
+      "epoch": 1.1315525876460768,
+      "grad_norm": 0.4662961962058868,
+      "learning_rate": 4.724790881184727e-06,
+      "loss": 0.2657,
+      "step": 3389
+    },
+    {
+      "epoch": 1.1318864774624373,
+      "grad_norm": 0.46380433224155077,
+      "learning_rate": 4.721881606844287e-06,
+      "loss": 0.2556,
+      "step": 3390
+    },
+    {
+      "epoch": 1.132220367278798,
+      "grad_norm": 0.47753855803050693,
+      "learning_rate": 4.718972426951417e-06,
+      "loss": 0.2677,
+      "step": 3391
+    },
+    {
+      "epoch": 1.1325542570951586,
+      "grad_norm": 0.4628103625240149,
+      "learning_rate": 4.716063342494055e-06,
+      "loss": 0.2403,
+      "step": 3392
+    },
+    {
+      "epoch": 1.1328881469115193,
+      "grad_norm": 0.492804679465433,
+      "learning_rate": 4.713154354460116e-06,
+      "loss": 0.258,
+      "step": 3393
+    },
+    {
+      "epoch": 1.1332220367278798,
+      "grad_norm": 0.4755871558576671,
+      "learning_rate": 4.710245463837475e-06,
+      "loss": 0.2631,
+      "step": 3394
+    },
+    {
+      "epoch": 1.1335559265442403,
+      "grad_norm": 0.4738809814015301,
+      "learning_rate": 4.707336671613976e-06,
+      "loss": 0.258,
+      "step": 3395
+    },
+    {
+      "epoch": 1.133889816360601,
+      "grad_norm": 0.4803552802140428,
+      "learning_rate": 4.7044279787774296e-06,
+      "loss": 0.2632,
+      "step": 3396
+    },
+    {
+      "epoch": 1.1342237061769616,
+      "grad_norm": 0.4893559748684362,
+      "learning_rate": 4.701519386315612e-06,
+      "loss": 0.2603,
+      "step": 3397
+    },
+    {
+      "epoch": 1.1345575959933223,
+      "grad_norm": 0.5060894578358864,
+      "learning_rate": 4.69861089521627e-06,
+      "loss": 0.2656,
+      "step": 3398
+    },
+    {
+      "epoch": 1.1348914858096828,
+      "grad_norm": 0.48314623898736725,
+      "learning_rate": 4.695702506467108e-06,
+      "loss": 0.2618,
+      "step": 3399
+    },
+    {
+      "epoch": 1.1352253756260433,
+      "grad_norm": 0.4640056219565378,
+      "learning_rate": 4.692794221055804e-06,
+      "loss": 0.26,
+      "step": 3400
+    },
+    {
+      "epoch": 1.135559265442404,
+      "grad_norm": 0.4682498603876342,
+      "learning_rate": 4.689886039969991e-06,
+      "loss": 0.2439,
+      "step": 3401
+    },
+    {
+      "epoch": 1.1358931552587646,
+      "grad_norm": 0.45749726217208936,
+      "learning_rate": 4.6869779641972785e-06,
+      "loss": 0.2466,
+      "step": 3402
+    },
+    {
+      "epoch": 1.1362270450751253,
+      "grad_norm": 0.46790222131580256,
+      "learning_rate": 4.684069994725229e-06,
+      "loss": 0.2652,
+      "step": 3403
+    },
+    {
+      "epoch": 1.1365609348914858,
+      "grad_norm": 0.4694721284673073,
+      "learning_rate": 4.681162132541377e-06,
+      "loss": 0.2565,
+      "step": 3404
+    },
+    {
+      "epoch": 1.1368948247078463,
+      "grad_norm": 0.48771117727250024,
+      "learning_rate": 4.678254378633214e-06,
+      "loss": 0.2578,
+      "step": 3405
+    },
+    {
+      "epoch": 1.137228714524207,
+      "grad_norm": 0.4647063665412998,
+      "learning_rate": 4.675346733988201e-06,
+      "loss": 0.2601,
+      "step": 3406
+    },
+    {
+      "epoch": 1.1375626043405676,
+      "grad_norm": 0.4812187595572356,
+      "learning_rate": 4.672439199593761e-06,
+      "loss": 0.2622,
+      "step": 3407
+    },
+    {
+      "epoch": 1.1378964941569283,
+      "grad_norm": 0.48872382450446084,
+      "learning_rate": 4.669531776437273e-06,
+      "loss": 0.2467,
+      "step": 3408
+    },
+    {
+      "epoch": 1.1382303839732888,
+      "grad_norm": 0.5029779827242856,
+      "learning_rate": 4.6666244655060866e-06,
+      "loss": 0.2671,
+      "step": 3409
+    },
+    {
+      "epoch": 1.1385642737896493,
+      "grad_norm": 0.4672205328485424,
+      "learning_rate": 4.663717267787507e-06,
+      "loss": 0.2461,
+      "step": 3410
+    },
+    {
+      "epoch": 1.13889816360601,
+      "grad_norm": 0.48323715627685626,
+      "learning_rate": 4.660810184268805e-06,
+      "loss": 0.2654,
+      "step": 3411
+    },
+    {
+      "epoch": 1.1392320534223705,
+      "grad_norm": 0.45528088923250254,
+      "learning_rate": 4.65790321593721e-06,
+      "loss": 0.2483,
+      "step": 3412
+    },
+    {
+      "epoch": 1.1395659432387313,
+      "grad_norm": 0.4711841289995006,
+      "learning_rate": 4.654996363779916e-06,
+      "loss": 0.243,
+      "step": 3413
+    },
+    {
+      "epoch": 1.1398998330550918,
+      "grad_norm": 0.504574645686041,
+      "learning_rate": 4.65208962878407e-06,
+      "loss": 0.2638,
+      "step": 3414
+    },
+    {
+      "epoch": 1.1402337228714523,
+      "grad_norm": 0.49354861155073687,
+      "learning_rate": 4.649183011936788e-06,
+      "loss": 0.2552,
+      "step": 3415
+    },
+    {
+      "epoch": 1.140567612687813,
+      "grad_norm": 0.492164296197576,
+      "learning_rate": 4.646276514225143e-06,
+      "loss": 0.2602,
+      "step": 3416
+    },
+    {
+      "epoch": 1.1409015025041735,
+      "grad_norm": 0.4939295895043665,
+      "learning_rate": 4.643370136636164e-06,
+      "loss": 0.2528,
+      "step": 3417
+    },
+    {
+      "epoch": 1.1412353923205343,
+      "grad_norm": 0.45112085356953563,
+      "learning_rate": 4.6404638801568435e-06,
+      "loss": 0.2529,
+      "step": 3418
+    },
+    {
+      "epoch": 1.1415692821368948,
+      "grad_norm": 0.46281809205660246,
+      "learning_rate": 4.637557745774129e-06,
+      "loss": 0.2377,
+      "step": 3419
+    },
+    {
+      "epoch": 1.1419031719532553,
+      "grad_norm": 0.4791200697264613,
+      "learning_rate": 4.634651734474932e-06,
+      "loss": 0.2516,
+      "step": 3420
+    },
+    {
+      "epoch": 1.142237061769616,
+      "grad_norm": 0.4852322250339329,
+      "learning_rate": 4.631745847246115e-06,
+      "loss": 0.2601,
+      "step": 3421
+    },
+    {
+      "epoch": 1.1425709515859765,
+      "grad_norm": 0.47190097779958856,
+      "learning_rate": 4.628840085074503e-06,
+      "loss": 0.251,
+      "step": 3422
+    },
+    {
+      "epoch": 1.1429048414023373,
+      "grad_norm": 0.48127570305688655,
+      "learning_rate": 4.6259344489468825e-06,
+      "loss": 0.2532,
+      "step": 3423
+    },
+    {
+      "epoch": 1.1432387312186978,
+      "grad_norm": 0.4651774089327851,
+      "learning_rate": 4.623028939849987e-06,
+      "loss": 0.2545,
+      "step": 3424
+    },
+    {
+      "epoch": 1.1435726210350585,
+      "grad_norm": 0.4695803001400044,
+      "learning_rate": 4.6201235587705155e-06,
+      "loss": 0.2458,
+      "step": 3425
+    },
+    {
+      "epoch": 1.143906510851419,
+      "grad_norm": 0.4842174025493916,
+      "learning_rate": 4.617218306695119e-06,
+      "loss": 0.26,
+      "step": 3426
+    },
+    {
+      "epoch": 1.1442404006677795,
+      "grad_norm": 0.4570493256231313,
+      "learning_rate": 4.614313184610407e-06,
+      "loss": 0.2435,
+      "step": 3427
+    },
+    {
+      "epoch": 1.1445742904841403,
+      "grad_norm": 0.4605835710752473,
+      "learning_rate": 4.611408193502942e-06,
+      "loss": 0.2516,
+      "step": 3428
+    },
+    {
+      "epoch": 1.1449081803005008,
+      "grad_norm": 0.48247395924640896,
+      "learning_rate": 4.608503334359247e-06,
+      "loss": 0.2661,
+      "step": 3429
+    },
+    {
+      "epoch": 1.1452420701168615,
+      "grad_norm": 0.485542207917262,
+      "learning_rate": 4.605598608165794e-06,
+      "loss": 0.2569,
+      "step": 3430
+    },
+    {
+      "epoch": 1.145575959933222,
+      "grad_norm": 0.4874839460283488,
+      "learning_rate": 4.602694015909014e-06,
+      "loss": 0.2633,
+      "step": 3431
+    },
+    {
+      "epoch": 1.1459098497495825,
+      "grad_norm": 0.4758032519824041,
+      "learning_rate": 4.599789558575293e-06,
+      "loss": 0.2627,
+      "step": 3432
+    },
+    {
+      "epoch": 1.1462437395659433,
+      "grad_norm": 0.45574429863246313,
+      "learning_rate": 4.596885237150968e-06,
+      "loss": 0.2464,
+      "step": 3433
+    },
+    {
+      "epoch": 1.1465776293823038,
+      "grad_norm": 0.460202557964237,
+      "learning_rate": 4.593981052622334e-06,
+      "loss": 0.2577,
+      "step": 3434
+    },
+    {
+      "epoch": 1.1469115191986645,
+      "grad_norm": 0.48881710153894803,
+      "learning_rate": 4.591077005975633e-06,
+      "loss": 0.2495,
+      "step": 3435
+    },
+    {
+      "epoch": 1.147245409015025,
+      "grad_norm": 0.48093890359251346,
+      "learning_rate": 4.588173098197068e-06,
+      "loss": 0.2591,
+      "step": 3436
+    },
+    {
+      "epoch": 1.1475792988313858,
+      "grad_norm": 0.45885231597487663,
+      "learning_rate": 4.585269330272789e-06,
+      "loss": 0.2468,
+      "step": 3437
+    },
+    {
+      "epoch": 1.1479131886477463,
+      "grad_norm": 0.49825486691399423,
+      "learning_rate": 4.582365703188902e-06,
+      "loss": 0.2538,
+      "step": 3438
+    },
+    {
+      "epoch": 1.1482470784641068,
+      "grad_norm": 0.5209641258849883,
+      "learning_rate": 4.5794622179314605e-06,
+      "loss": 0.2437,
+      "step": 3439
+    },
+    {
+      "epoch": 1.1485809682804675,
+      "grad_norm": 0.4792304140240776,
+      "learning_rate": 4.576558875486477e-06,
+      "loss": 0.2514,
+      "step": 3440
+    },
+    {
+      "epoch": 1.148914858096828,
+      "grad_norm": 0.47006680275425583,
+      "learning_rate": 4.573655676839911e-06,
+      "loss": 0.2602,
+      "step": 3441
+    },
+    {
+      "epoch": 1.1492487479131888,
+      "grad_norm": 0.4968726339949825,
+      "learning_rate": 4.5707526229776715e-06,
+      "loss": 0.2583,
+      "step": 3442
+    },
+    {
+      "epoch": 1.1495826377295493,
+      "grad_norm": 0.5114608563476335,
+      "learning_rate": 4.567849714885622e-06,
+      "loss": 0.2591,
+      "step": 3443
+    },
+    {
+      "epoch": 1.1499165275459098,
+      "grad_norm": 0.5026719564051744,
+      "learning_rate": 4.564946953549576e-06,
+      "loss": 0.259,
+      "step": 3444
+    },
+    {
+      "epoch": 1.1502504173622705,
+      "grad_norm": 0.5045197912614322,
+      "learning_rate": 4.562044339955295e-06,
+      "loss": 0.2663,
+      "step": 3445
+    },
+    {
+      "epoch": 1.150584307178631,
+      "grad_norm": 0.46574757526494864,
+      "learning_rate": 4.559141875088491e-06,
+      "loss": 0.2561,
+      "step": 3446
+    },
+    {
+      "epoch": 1.1509181969949918,
+      "grad_norm": 0.4753922852219625,
+      "learning_rate": 4.556239559934828e-06,
+      "loss": 0.2527,
+      "step": 3447
+    },
+    {
+      "epoch": 1.1512520868113523,
+      "grad_norm": 0.4458902961016588,
+      "learning_rate": 4.553337395479913e-06,
+      "loss": 0.2336,
+      "step": 3448
+    },
+    {
+      "epoch": 1.1515859766277128,
+      "grad_norm": 0.48231930546110324,
+      "learning_rate": 4.550435382709311e-06,
+      "loss": 0.2609,
+      "step": 3449
+    },
+    {
+      "epoch": 1.1519198664440735,
+      "grad_norm": 0.4943348582256607,
+      "learning_rate": 4.547533522608529e-06,
+      "loss": 0.2584,
+      "step": 3450
+    },
+    {
+      "epoch": 1.152253756260434,
+      "grad_norm": 0.47634416171686617,
+      "learning_rate": 4.544631816163022e-06,
+      "loss": 0.2607,
+      "step": 3451
+    },
+    {
+      "epoch": 1.1525876460767948,
+      "grad_norm": 0.49369952553358815,
+      "learning_rate": 4.541730264358198e-06,
+      "loss": 0.2501,
+      "step": 3452
+    },
+    {
+      "epoch": 1.1529215358931553,
+      "grad_norm": 0.49847832785402146,
+      "learning_rate": 4.5388288681794056e-06,
+      "loss": 0.249,
+      "step": 3453
+    },
+    {
+      "epoch": 1.1532554257095158,
+      "grad_norm": 0.468018946458216,
+      "learning_rate": 4.535927628611946e-06,
+      "loss": 0.2468,
+      "step": 3454
+    },
+    {
+      "epoch": 1.1535893155258765,
+      "grad_norm": 0.47085316976405056,
+      "learning_rate": 4.5330265466410625e-06,
+      "loss": 0.249,
+      "step": 3455
+    },
+    {
+      "epoch": 1.153923205342237,
+      "grad_norm": 0.4786753803315987,
+      "learning_rate": 4.5301256232519495e-06,
+      "loss": 0.2588,
+      "step": 3456
+    },
+    {
+      "epoch": 1.1542570951585978,
+      "grad_norm": 0.4636415063170965,
+      "learning_rate": 4.5272248594297465e-06,
+      "loss": 0.257,
+      "step": 3457
+    },
+    {
+      "epoch": 1.1545909849749583,
+      "grad_norm": 0.47524339494188417,
+      "learning_rate": 4.524324256159536e-06,
+      "loss": 0.2638,
+      "step": 3458
+    },
+    {
+      "epoch": 1.1549248747913188,
+      "grad_norm": 0.47353915058271845,
+      "learning_rate": 4.5214238144263504e-06,
+      "loss": 0.2538,
+      "step": 3459
+    },
+    {
+      "epoch": 1.1552587646076795,
+      "grad_norm": 0.4937284790283414,
+      "learning_rate": 4.518523535215161e-06,
+      "loss": 0.248,
+      "step": 3460
+    },
+    {
+      "epoch": 1.15559265442404,
+      "grad_norm": 0.4881316832292819,
+      "learning_rate": 4.515623419510891e-06,
+      "loss": 0.2449,
+      "step": 3461
+    },
+    {
+      "epoch": 1.1559265442404008,
+      "grad_norm": 0.4657675308479749,
+      "learning_rate": 4.512723468298403e-06,
+      "loss": 0.2621,
+      "step": 3462
+    },
+    {
+      "epoch": 1.1562604340567613,
+      "grad_norm": 0.4855881975380206,
+      "learning_rate": 4.509823682562507e-06,
+      "loss": 0.2578,
+      "step": 3463
+    },
+    {
+      "epoch": 1.1565943238731218,
+      "grad_norm": 0.4819617034341652,
+      "learning_rate": 4.506924063287951e-06,
+      "loss": 0.2558,
+      "step": 3464
+    },
+    {
+      "epoch": 1.1569282136894825,
+      "grad_norm": 0.4834067139400789,
+      "learning_rate": 4.504024611459434e-06,
+      "loss": 0.2614,
+      "step": 3465
+    },
+    {
+      "epoch": 1.157262103505843,
+      "grad_norm": 0.46584092850446146,
+      "learning_rate": 4.501125328061597e-06,
+      "loss": 0.2519,
+      "step": 3466
+    },
+    {
+      "epoch": 1.1575959933222038,
+      "grad_norm": 0.49322034564584494,
+      "learning_rate": 4.498226214079018e-06,
+      "loss": 0.2665,
+      "step": 3467
+    },
+    {
+      "epoch": 1.1579298831385643,
+      "grad_norm": 0.48636648744469796,
+      "learning_rate": 4.495327270496222e-06,
+      "loss": 0.2465,
+      "step": 3468
+    },
+    {
+      "epoch": 1.1582637729549248,
+      "grad_norm": 0.505294020232348,
+      "learning_rate": 4.492428498297676e-06,
+      "loss": 0.2702,
+      "step": 3469
+    },
+    {
+      "epoch": 1.1585976627712855,
+      "grad_norm": 0.4766606863757607,
+      "learning_rate": 4.489529898467789e-06,
+      "loss": 0.2562,
+      "step": 3470
+    },
+    {
+      "epoch": 1.158931552587646,
+      "grad_norm": 0.4743996048458284,
+      "learning_rate": 4.486631471990907e-06,
+      "loss": 0.2523,
+      "step": 3471
+    },
+    {
+      "epoch": 1.1592654424040068,
+      "grad_norm": 0.49074832309865424,
+      "learning_rate": 4.483733219851324e-06,
+      "loss": 0.2658,
+      "step": 3472
+    },
+    {
+      "epoch": 1.1595993322203673,
+      "grad_norm": 0.4865677602606859,
+      "learning_rate": 4.480835143033269e-06,
+      "loss": 0.2668,
+      "step": 3473
+    },
+    {
+      "epoch": 1.1599332220367278,
+      "grad_norm": 0.48262747810386697,
+      "learning_rate": 4.477937242520915e-06,
+      "loss": 0.259,
+      "step": 3474
+    },
+    {
+      "epoch": 1.1602671118530885,
+      "grad_norm": 0.4672756744565033,
+      "learning_rate": 4.475039519298376e-06,
+      "loss": 0.259,
+      "step": 3475
+    },
+    {
+      "epoch": 1.160601001669449,
+      "grad_norm": 0.5155767109746476,
+      "learning_rate": 4.472141974349702e-06,
+      "loss": 0.2672,
+      "step": 3476
+    },
+    {
+      "epoch": 1.1609348914858098,
+      "grad_norm": 0.45810077333459864,
+      "learning_rate": 4.469244608658884e-06,
+      "loss": 0.2574,
+      "step": 3477
+    },
+    {
+      "epoch": 1.1612687813021703,
+      "grad_norm": 0.48879351975377217,
+      "learning_rate": 4.466347423209852e-06,
+      "loss": 0.2547,
+      "step": 3478
+    },
+    {
+      "epoch": 1.1616026711185308,
+      "grad_norm": 0.4860580083134971,
+      "learning_rate": 4.463450418986477e-06,
+      "loss": 0.2664,
+      "step": 3479
+    },
+    {
+      "epoch": 1.1619365609348915,
+      "grad_norm": 0.4735014726351396,
+      "learning_rate": 4.460553596972565e-06,
+      "loss": 0.2517,
+      "step": 3480
+    },
+    {
+      "epoch": 1.162270450751252,
+      "grad_norm": 0.4623926740515699,
+      "learning_rate": 4.457656958151862e-06,
+      "loss": 0.254,
+      "step": 3481
+    },
+    {
+      "epoch": 1.1626043405676127,
+      "grad_norm": 0.4472464714383808,
+      "learning_rate": 4.45476050350805e-06,
+      "loss": 0.2407,
+      "step": 3482
+    },
+    {
+      "epoch": 1.1629382303839733,
+      "grad_norm": 0.4857714437239861,
+      "learning_rate": 4.451864234024754e-06,
+      "loss": 0.2635,
+      "step": 3483
+    },
+    {
+      "epoch": 1.1632721202003338,
+      "grad_norm": 0.4759630879742391,
+      "learning_rate": 4.44896815068553e-06,
+      "loss": 0.2558,
+      "step": 3484
+    },
+    {
+      "epoch": 1.1636060100166945,
+      "grad_norm": 0.46614926089157926,
+      "learning_rate": 4.446072254473873e-06,
+      "loss": 0.2582,
+      "step": 3485
+    },
+    {
+      "epoch": 1.163939899833055,
+      "grad_norm": 0.47246588811222084,
+      "learning_rate": 4.443176546373215e-06,
+      "loss": 0.2523,
+      "step": 3486
+    },
+    {
+      "epoch": 1.1642737896494157,
+      "grad_norm": 0.46717998736623373,
+      "learning_rate": 4.440281027366922e-06,
+      "loss": 0.2574,
+      "step": 3487
+    },
+    {
+      "epoch": 1.1646076794657763,
+      "grad_norm": 0.4665283350518405,
+      "learning_rate": 4.4373856984382985e-06,
+      "loss": 0.2559,
+      "step": 3488
+    },
+    {
+      "epoch": 1.164941569282137,
+      "grad_norm": 0.47319497186739645,
+      "learning_rate": 4.434490560570583e-06,
+      "loss": 0.2597,
+      "step": 3489
+    },
+    {
+      "epoch": 1.1652754590984975,
+      "grad_norm": 0.4742288054666444,
+      "learning_rate": 4.431595614746946e-06,
+      "loss": 0.2525,
+      "step": 3490
+    },
+    {
+      "epoch": 1.165609348914858,
+      "grad_norm": 0.4711168776536424,
+      "learning_rate": 4.428700861950503e-06,
+      "loss": 0.251,
+      "step": 3491
+    },
+    {
+      "epoch": 1.1659432387312187,
+      "grad_norm": 0.4664410345922968,
+      "learning_rate": 4.425806303164293e-06,
+      "loss": 0.2618,
+      "step": 3492
+    },
+    {
+      "epoch": 1.1662771285475793,
+      "grad_norm": 0.47386994384623854,
+      "learning_rate": 4.422911939371294e-06,
+      "loss": 0.2509,
+      "step": 3493
+    },
+    {
+      "epoch": 1.16661101836394,
+      "grad_norm": 0.48911672100885184,
+      "learning_rate": 4.420017771554417e-06,
+      "loss": 0.2528,
+      "step": 3494
+    },
+    {
+      "epoch": 1.1669449081803005,
+      "grad_norm": 0.4848486889176519,
+      "learning_rate": 4.4171238006965046e-06,
+      "loss": 0.2617,
+      "step": 3495
+    },
+    {
+      "epoch": 1.167278797996661,
+      "grad_norm": 0.477501790875134,
+      "learning_rate": 4.4142300277803365e-06,
+      "loss": 0.2552,
+      "step": 3496
+    },
+    {
+      "epoch": 1.1676126878130217,
+      "grad_norm": 0.47102001956227585,
+      "learning_rate": 4.411336453788622e-06,
+      "loss": 0.2533,
+      "step": 3497
+    },
+    {
+      "epoch": 1.1679465776293823,
+      "grad_norm": 0.48717115672643757,
+      "learning_rate": 4.408443079704005e-06,
+      "loss": 0.2572,
+      "step": 3498
+    },
+    {
+      "epoch": 1.168280467445743,
+      "grad_norm": 0.45593138670494165,
+      "learning_rate": 4.4055499065090575e-06,
+      "loss": 0.2635,
+      "step": 3499
+    },
+    {
+      "epoch": 1.1686143572621035,
+      "grad_norm": 0.5009152216054562,
+      "learning_rate": 4.402656935186289e-06,
+      "loss": 0.2611,
+      "step": 3500
+    },
+    {
+      "epoch": 1.1689482470784642,
+      "grad_norm": 0.45855156815179526,
+      "learning_rate": 4.399764166718138e-06,
+      "loss": 0.252,
+      "step": 3501
+    },
+    {
+      "epoch": 1.1692821368948247,
+      "grad_norm": 0.49244469094797705,
+      "learning_rate": 4.396871602086972e-06,
+      "loss": 0.2584,
+      "step": 3502
+    },
+    {
+      "epoch": 1.1696160267111853,
+      "grad_norm": 0.4864777603556692,
+      "learning_rate": 4.393979242275091e-06,
+      "loss": 0.2672,
+      "step": 3503
+    },
+    {
+      "epoch": 1.169949916527546,
+      "grad_norm": 0.5241389691884824,
+      "learning_rate": 4.3910870882647265e-06,
+      "loss": 0.2499,
+      "step": 3504
+    },
+    {
+      "epoch": 1.1702838063439065,
+      "grad_norm": 0.5079456495919704,
+      "learning_rate": 4.38819514103804e-06,
+      "loss": 0.2546,
+      "step": 3505
+    },
+    {
+      "epoch": 1.1706176961602672,
+      "grad_norm": 0.4716002909106285,
+      "learning_rate": 4.3853034015771176e-06,
+      "loss": 0.2542,
+      "step": 3506
+    },
+    {
+      "epoch": 1.1709515859766277,
+      "grad_norm": 0.48454826054947664,
+      "learning_rate": 4.382411870863982e-06,
+      "loss": 0.2651,
+      "step": 3507
+    },
+    {
+      "epoch": 1.1712854757929883,
+      "grad_norm": 0.5019164557002449,
+      "learning_rate": 4.379520549880583e-06,
+      "loss": 0.2551,
+      "step": 3508
+    },
+    {
+      "epoch": 1.171619365609349,
+      "grad_norm": 0.5000904630667264,
+      "learning_rate": 4.376629439608797e-06,
+      "loss": 0.2686,
+      "step": 3509
+    },
+    {
+      "epoch": 1.1719532554257095,
+      "grad_norm": 0.4786419922536363,
+      "learning_rate": 4.3737385410304314e-06,
+      "loss": 0.2551,
+      "step": 3510
+    },
+    {
+      "epoch": 1.1722871452420702,
+      "grad_norm": 0.47355555635829893,
+      "learning_rate": 4.370847855127218e-06,
+      "loss": 0.2616,
+      "step": 3511
+    },
+    {
+      "epoch": 1.1726210350584307,
+      "grad_norm": 0.4839159518300518,
+      "learning_rate": 4.36795738288082e-06,
+      "loss": 0.2517,
+      "step": 3512
+    },
+    {
+      "epoch": 1.1729549248747912,
+      "grad_norm": 0.4799255465899718,
+      "learning_rate": 4.365067125272827e-06,
+      "loss": 0.2582,
+      "step": 3513
+    },
+    {
+      "epoch": 1.173288814691152,
+      "grad_norm": 0.46424645506166096,
+      "learning_rate": 4.362177083284757e-06,
+      "loss": 0.2488,
+      "step": 3514
+    },
+    {
+      "epoch": 1.1736227045075125,
+      "grad_norm": 0.48747757430836697,
+      "learning_rate": 4.3592872578980495e-06,
+      "loss": 0.2597,
+      "step": 3515
+    },
+    {
+      "epoch": 1.1739565943238732,
+      "grad_norm": 0.456862948770951,
+      "learning_rate": 4.356397650094076e-06,
+      "loss": 0.248,
+      "step": 3516
+    },
+    {
+      "epoch": 1.1742904841402337,
+      "grad_norm": 0.48183157496733514,
+      "learning_rate": 4.353508260854134e-06,
+      "loss": 0.2495,
+      "step": 3517
+    },
+    {
+      "epoch": 1.1746243739565942,
+      "grad_norm": 0.4921478424388822,
+      "learning_rate": 4.350619091159444e-06,
+      "loss": 0.2572,
+      "step": 3518
+    },
+    {
+      "epoch": 1.174958263772955,
+      "grad_norm": 0.47003770489623325,
+      "learning_rate": 4.347730141991155e-06,
+      "loss": 0.2639,
+      "step": 3519
+    },
+    {
+      "epoch": 1.1752921535893155,
+      "grad_norm": 0.4881824291812725,
+      "learning_rate": 4.344841414330334e-06,
+      "loss": 0.2618,
+      "step": 3520
+    },
+    {
+      "epoch": 1.1756260434056762,
+      "grad_norm": 0.45436185822877323,
+      "learning_rate": 4.341952909157984e-06,
+      "loss": 0.2507,
+      "step": 3521
+    },
+    {
+      "epoch": 1.1759599332220367,
+      "grad_norm": 0.4899991943101413,
+      "learning_rate": 4.339064627455022e-06,
+      "loss": 0.2559,
+      "step": 3522
+    },
+    {
+      "epoch": 1.1762938230383972,
+      "grad_norm": 0.4410383242951134,
+      "learning_rate": 4.3361765702022965e-06,
+      "loss": 0.2412,
+      "step": 3523
+    },
+    {
+      "epoch": 1.176627712854758,
+      "grad_norm": 0.47571750740011465,
+      "learning_rate": 4.333288738380574e-06,
+      "loss": 0.2661,
+      "step": 3524
+    },
+    {
+      "epoch": 1.1769616026711185,
+      "grad_norm": 0.4774301977159378,
+      "learning_rate": 4.3304011329705485e-06,
+      "loss": 0.2508,
+      "step": 3525
+    },
+    {
+      "epoch": 1.1772954924874792,
+      "grad_norm": 0.5099263665995097,
+      "learning_rate": 4.327513754952838e-06,
+      "loss": 0.2505,
+      "step": 3526
+    },
+    {
+      "epoch": 1.1776293823038397,
+      "grad_norm": 0.4814503722508517,
+      "learning_rate": 4.3246266053079785e-06,
+      "loss": 0.2712,
+      "step": 3527
+    },
+    {
+      "epoch": 1.1779632721202002,
+      "grad_norm": 0.4739135278356501,
+      "learning_rate": 4.321739685016433e-06,
+      "loss": 0.2528,
+      "step": 3528
+    },
+    {
+      "epoch": 1.178297161936561,
+      "grad_norm": 0.47374159674201455,
+      "learning_rate": 4.318852995058582e-06,
+      "loss": 0.2574,
+      "step": 3529
+    },
+    {
+      "epoch": 1.1786310517529215,
+      "grad_norm": 0.4696422743160708,
+      "learning_rate": 4.315966536414733e-06,
+      "loss": 0.2607,
+      "step": 3530
+    },
+    {
+      "epoch": 1.1789649415692822,
+      "grad_norm": 0.46663690813878184,
+      "learning_rate": 4.313080310065111e-06,
+      "loss": 0.2426,
+      "step": 3531
+    },
+    {
+      "epoch": 1.1792988313856427,
+      "grad_norm": 0.4730314227203527,
+      "learning_rate": 4.3101943169898655e-06,
+      "loss": 0.259,
+      "step": 3532
+    },
+    {
+      "epoch": 1.1796327212020032,
+      "grad_norm": 0.4667274396473063,
+      "learning_rate": 4.30730855816906e-06,
+      "loss": 0.2473,
+      "step": 3533
+    },
+    {
+      "epoch": 1.179966611018364,
+      "grad_norm": 0.48978843061901417,
+      "learning_rate": 4.3044230345826885e-06,
+      "loss": 0.2501,
+      "step": 3534
+    },
+    {
+      "epoch": 1.1803005008347245,
+      "grad_norm": 0.4503101554106291,
+      "learning_rate": 4.30153774721066e-06,
+      "loss": 0.2542,
+      "step": 3535
+    },
+    {
+      "epoch": 1.1806343906510852,
+      "grad_norm": 0.4636812836131682,
+      "learning_rate": 4.298652697032799e-06,
+      "loss": 0.2487,
+      "step": 3536
+    },
+    {
+      "epoch": 1.1809682804674457,
+      "grad_norm": 0.5022421387252975,
+      "learning_rate": 4.2957678850288574e-06,
+      "loss": 0.27,
+      "step": 3537
+    },
+    {
+      "epoch": 1.1813021702838062,
+      "grad_norm": 0.4855188745403909,
+      "learning_rate": 4.2928833121785e-06,
+      "loss": 0.2476,
+      "step": 3538
+    },
+    {
+      "epoch": 1.181636060100167,
+      "grad_norm": 0.5024105424123585,
+      "learning_rate": 4.289998979461314e-06,
+      "loss": 0.2696,
+      "step": 3539
+    },
+    {
+      "epoch": 1.1819699499165275,
+      "grad_norm": 0.4897147815967363,
+      "learning_rate": 4.287114887856803e-06,
+      "loss": 0.2587,
+      "step": 3540
+    },
+    {
+      "epoch": 1.1823038397328882,
+      "grad_norm": 0.4688388743286566,
+      "learning_rate": 4.284231038344389e-06,
+      "loss": 0.2485,
+      "step": 3541
+    },
+    {
+      "epoch": 1.1826377295492487,
+      "grad_norm": 0.48810015078630753,
+      "learning_rate": 4.281347431903416e-06,
+      "loss": 0.2689,
+      "step": 3542
+    },
+    {
+      "epoch": 1.1829716193656092,
+      "grad_norm": 0.4907150982179452,
+      "learning_rate": 4.278464069513138e-06,
+      "loss": 0.247,
+      "step": 3543
+    },
+    {
+      "epoch": 1.18330550918197,
+      "grad_norm": 0.4749536840946802,
+      "learning_rate": 4.275580952152733e-06,
+      "loss": 0.2619,
+      "step": 3544
+    },
+    {
+      "epoch": 1.1836393989983305,
+      "grad_norm": 0.47876549174866534,
+      "learning_rate": 4.272698080801289e-06,
+      "loss": 0.2579,
+      "step": 3545
+    },
+    {
+      "epoch": 1.1839732888146912,
+      "grad_norm": 0.4823002090073175,
+      "learning_rate": 4.269815456437819e-06,
+      "loss": 0.2535,
+      "step": 3546
+    },
+    {
+      "epoch": 1.1843071786310517,
+      "grad_norm": 0.47850807556405756,
+      "learning_rate": 4.266933080041243e-06,
+      "loss": 0.2631,
+      "step": 3547
+    },
+    {
+      "epoch": 1.1846410684474122,
+      "grad_norm": 0.4763315708661121,
+      "learning_rate": 4.264050952590404e-06,
+      "loss": 0.2524,
+      "step": 3548
+    },
+    {
+      "epoch": 1.184974958263773,
+      "grad_norm": 0.49398834964920285,
+      "learning_rate": 4.261169075064054e-06,
+      "loss": 0.2523,
+      "step": 3549
+    },
+    {
+      "epoch": 1.1853088480801335,
+      "grad_norm": 0.5018914458068438,
+      "learning_rate": 4.258287448440867e-06,
+      "loss": 0.2765,
+      "step": 3550
+    },
+    {
+      "epoch": 1.1856427378964942,
+      "grad_norm": 0.517364937308809,
+      "learning_rate": 4.255406073699428e-06,
+      "loss": 0.2513,
+      "step": 3551
+    },
+    {
+      "epoch": 1.1859766277128547,
+      "grad_norm": 0.48730319411416956,
+      "learning_rate": 4.252524951818236e-06,
+      "loss": 0.2683,
+      "step": 3552
+    },
+    {
+      "epoch": 1.1863105175292152,
+      "grad_norm": 0.48597243617524993,
+      "learning_rate": 4.2496440837757076e-06,
+      "loss": 0.2599,
+      "step": 3553
+    },
+    {
+      "epoch": 1.186644407345576,
+      "grad_norm": 0.48516024704444966,
+      "learning_rate": 4.246763470550167e-06,
+      "loss": 0.2587,
+      "step": 3554
+    },
+    {
+      "epoch": 1.1869782971619365,
+      "grad_norm": 0.4806958033204075,
+      "learning_rate": 4.243883113119858e-06,
+      "loss": 0.2568,
+      "step": 3555
+    },
+    {
+      "epoch": 1.1873121869782972,
+      "grad_norm": 0.4852087609669724,
+      "learning_rate": 4.241003012462934e-06,
+      "loss": 0.2626,
+      "step": 3556
+    },
+    {
+      "epoch": 1.1876460767946577,
+      "grad_norm": 0.46049914924100016,
+      "learning_rate": 4.2381231695574635e-06,
+      "loss": 0.2525,
+      "step": 3557
+    },
+    {
+      "epoch": 1.1879799666110185,
+      "grad_norm": 0.4768933833134438,
+      "learning_rate": 4.235243585381423e-06,
+      "loss": 0.2458,
+      "step": 3558
+    },
+    {
+      "epoch": 1.188313856427379,
+      "grad_norm": 0.46209068043329504,
+      "learning_rate": 4.2323642609127075e-06,
+      "loss": 0.2537,
+      "step": 3559
+    },
+    {
+      "epoch": 1.1886477462437395,
+      "grad_norm": 0.4648449330397029,
+      "learning_rate": 4.229485197129122e-06,
+      "loss": 0.2496,
+      "step": 3560
+    },
+    {
+      "epoch": 1.1889816360601002,
+      "grad_norm": 0.45843889488890005,
+      "learning_rate": 4.226606395008378e-06,
+      "loss": 0.2619,
+      "step": 3561
+    },
+    {
+      "epoch": 1.1893155258764607,
+      "grad_norm": 0.4754941973864502,
+      "learning_rate": 4.2237278555281045e-06,
+      "loss": 0.2668,
+      "step": 3562
+    },
+    {
+      "epoch": 1.1896494156928215,
+      "grad_norm": 0.4836575004198821,
+      "learning_rate": 4.220849579665838e-06,
+      "loss": 0.2601,
+      "step": 3563
+    },
+    {
+      "epoch": 1.189983305509182,
+      "grad_norm": 0.4614390742601844,
+      "learning_rate": 4.217971568399025e-06,
+      "loss": 0.2567,
+      "step": 3564
+    },
+    {
+      "epoch": 1.1903171953255425,
+      "grad_norm": 0.49222658467874836,
+      "learning_rate": 4.215093822705024e-06,
+      "loss": 0.2578,
+      "step": 3565
+    },
+    {
+      "epoch": 1.1906510851419032,
+      "grad_norm": 0.47841183020391936,
+      "learning_rate": 4.212216343561103e-06,
+      "loss": 0.2616,
+      "step": 3566
+    },
+    {
+      "epoch": 1.1909849749582637,
+      "grad_norm": 0.46501235466231183,
+      "learning_rate": 4.209339131944437e-06,
+      "loss": 0.2567,
+      "step": 3567
+    },
+    {
+      "epoch": 1.1913188647746245,
+      "grad_norm": 0.4561815247285503,
+      "learning_rate": 4.206462188832114e-06,
+      "loss": 0.2566,
+      "step": 3568
+    },
+    {
+      "epoch": 1.191652754590985,
+      "grad_norm": 0.4765502870010848,
+      "learning_rate": 4.203585515201131e-06,
+      "loss": 0.2485,
+      "step": 3569
+    },
+    {
+      "epoch": 1.1919866444073457,
+      "grad_norm": 0.47932369908424866,
+      "learning_rate": 4.200709112028389e-06,
+      "loss": 0.2588,
+      "step": 3570
+    },
+    {
+      "epoch": 1.1923205342237062,
+      "grad_norm": 0.4825609459238394,
+      "learning_rate": 4.197832980290699e-06,
+      "loss": 0.2493,
+      "step": 3571
+    },
+    {
+      "epoch": 1.1926544240400667,
+      "grad_norm": 0.508127897766414,
+      "learning_rate": 4.194957120964781e-06,
+      "loss": 0.2672,
+      "step": 3572
+    },
+    {
+      "epoch": 1.1929883138564275,
+      "grad_norm": 0.4610601619742866,
+      "learning_rate": 4.192081535027263e-06,
+      "loss": 0.2447,
+      "step": 3573
+    },
+    {
+      "epoch": 1.193322203672788,
+      "grad_norm": 0.4758123220260201,
+      "learning_rate": 4.189206223454678e-06,
+      "loss": 0.2528,
+      "step": 3574
+    },
+    {
+      "epoch": 1.1936560934891487,
+      "grad_norm": 0.4700046473284258,
+      "learning_rate": 4.1863311872234666e-06,
+      "loss": 0.2605,
+      "step": 3575
+    },
+    {
+      "epoch": 1.1939899833055092,
+      "grad_norm": 0.4792447126430432,
+      "learning_rate": 4.183456427309979e-06,
+      "loss": 0.2541,
+      "step": 3576
+    },
+    {
+      "epoch": 1.1943238731218697,
+      "grad_norm": 0.4755706804553937,
+      "learning_rate": 4.180581944690465e-06,
+      "loss": 0.2496,
+      "step": 3577
+    },
+    {
+      "epoch": 1.1946577629382304,
+      "grad_norm": 0.4586086928163983,
+      "learning_rate": 4.177707740341088e-06,
+      "loss": 0.2545,
+      "step": 3578
+    },
+    {
+      "epoch": 1.194991652754591,
+      "grad_norm": 0.4757086899649477,
+      "learning_rate": 4.1748338152379085e-06,
+      "loss": 0.2575,
+      "step": 3579
+    },
+    {
+      "epoch": 1.1953255425709517,
+      "grad_norm": 0.4380140294765915,
+      "learning_rate": 4.171960170356901e-06,
+      "loss": 0.245,
+      "step": 3580
+    },
+    {
+      "epoch": 1.1956594323873122,
+      "grad_norm": 0.47450434456393065,
+      "learning_rate": 4.169086806673938e-06,
+      "loss": 0.2622,
+      "step": 3581
+    },
+    {
+      "epoch": 1.1959933222036727,
+      "grad_norm": 0.47229644812372223,
+      "learning_rate": 4.1662137251647985e-06,
+      "loss": 0.259,
+      "step": 3582
+    },
+    {
+      "epoch": 1.1963272120200334,
+      "grad_norm": 0.46305431221882776,
+      "learning_rate": 4.1633409268051675e-06,
+      "loss": 0.2666,
+      "step": 3583
+    },
+    {
+      "epoch": 1.196661101836394,
+      "grad_norm": 0.4718273605004935,
+      "learning_rate": 4.1604684125706296e-06,
+      "loss": 0.2552,
+      "step": 3584
+    },
+    {
+      "epoch": 1.1969949916527547,
+      "grad_norm": 0.4620870463887997,
+      "learning_rate": 4.157596183436681e-06,
+      "loss": 0.2594,
+      "step": 3585
+    },
+    {
+      "epoch": 1.1973288814691152,
+      "grad_norm": 0.46218667751440995,
+      "learning_rate": 4.154724240378712e-06,
+      "loss": 0.2563,
+      "step": 3586
+    },
+    {
+      "epoch": 1.1976627712854757,
+      "grad_norm": 0.47130824080017675,
+      "learning_rate": 4.151852584372021e-06,
+      "loss": 0.2568,
+      "step": 3587
+    },
+    {
+      "epoch": 1.1979966611018364,
+      "grad_norm": 0.48804510773529425,
+      "learning_rate": 4.148981216391807e-06,
+      "loss": 0.2657,
+      "step": 3588
+    },
+    {
+      "epoch": 1.198330550918197,
+      "grad_norm": 0.48066396390395894,
+      "learning_rate": 4.146110137413172e-06,
+      "loss": 0.2505,
+      "step": 3589
+    },
+    {
+      "epoch": 1.1986644407345577,
+      "grad_norm": 0.4880645238828345,
+      "learning_rate": 4.143239348411119e-06,
+      "loss": 0.2559,
+      "step": 3590
+    },
+    {
+      "epoch": 1.1989983305509182,
+      "grad_norm": 0.4728617780985077,
+      "learning_rate": 4.1403688503605545e-06,
+      "loss": 0.2532,
+      "step": 3591
+    },
+    {
+      "epoch": 1.1993322203672787,
+      "grad_norm": 0.4816684241469875,
+      "learning_rate": 4.137498644236281e-06,
+      "loss": 0.2539,
+      "step": 3592
+    },
+    {
+      "epoch": 1.1996661101836394,
+      "grad_norm": 0.4552101175501875,
+      "learning_rate": 4.13462873101301e-06,
+      "loss": 0.2463,
+      "step": 3593
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.534203825069781,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.2618,
+      "step": 3594
+    },
+    {
+      "epoch": 1.2003338898163607,
+      "grad_norm": 0.4657417298228479,
+      "learning_rate": 4.128889787167803e-06,
+      "loss": 0.2385,
+      "step": 3595
+    },
+    {
+      "epoch": 1.2006677796327212,
+      "grad_norm": 0.4802575028409594,
+      "learning_rate": 4.1260207584947825e-06,
+      "loss": 0.2568,
+      "step": 3596
+    },
+    {
+      "epoch": 1.2010016694490817,
+      "grad_norm": 0.44055430188175276,
+      "learning_rate": 4.123152026620592e-06,
+      "loss": 0.2378,
+      "step": 3597
+    },
+    {
+      "epoch": 1.2013355592654424,
+      "grad_norm": 0.44905373641286084,
+      "learning_rate": 4.120283592519442e-06,
+      "loss": 0.2359,
+      "step": 3598
+    },
+    {
+      "epoch": 1.201669449081803,
+      "grad_norm": 0.4894973649350659,
+      "learning_rate": 4.117415457165434e-06,
+      "loss": 0.2484,
+      "step": 3599
+    },
+    {
+      "epoch": 1.2020033388981637,
+      "grad_norm": 0.48276195603876465,
+      "learning_rate": 4.1145476215325745e-06,
+      "loss": 0.2453,
+      "step": 3600
+    },
+    {
+      "epoch": 1.2023372287145242,
+      "grad_norm": 0.47805011812249776,
+      "learning_rate": 4.111680086594762e-06,
+      "loss": 0.2535,
+      "step": 3601
+    },
+    {
+      "epoch": 1.2026711185308847,
+      "grad_norm": 0.48040796515300277,
+      "learning_rate": 4.1088128533258005e-06,
+      "loss": 0.2546,
+      "step": 3602
+    },
+    {
+      "epoch": 1.2030050083472454,
+      "grad_norm": 0.4660682167550782,
+      "learning_rate": 4.105945922699387e-06,
+      "loss": 0.2646,
+      "step": 3603
+    },
+    {
+      "epoch": 1.203338898163606,
+      "grad_norm": 0.4751980151189258,
+      "learning_rate": 4.103079295689115e-06,
+      "loss": 0.2615,
+      "step": 3604
+    },
+    {
+      "epoch": 1.2036727879799667,
+      "grad_norm": 0.48260280462291555,
+      "learning_rate": 4.100212973268478e-06,
+      "loss": 0.2601,
+      "step": 3605
+    },
+    {
+      "epoch": 1.2040066777963272,
+      "grad_norm": 0.474680010725752,
+      "learning_rate": 4.097346956410861e-06,
+      "loss": 0.2424,
+      "step": 3606
+    },
+    {
+      "epoch": 1.2043405676126877,
+      "grad_norm": 0.4616201524141293,
+      "learning_rate": 4.094481246089553e-06,
+      "loss": 0.2435,
+      "step": 3607
+    },
+    {
+      "epoch": 1.2046744574290484,
+      "grad_norm": 0.48107813124840965,
+      "learning_rate": 4.09161584327773e-06,
+      "loss": 0.2501,
+      "step": 3608
+    },
+    {
+      "epoch": 1.205008347245409,
+      "grad_norm": 0.47965594383430143,
+      "learning_rate": 4.088750748948469e-06,
+      "loss": 0.266,
+      "step": 3609
+    },
+    {
+      "epoch": 1.2053422370617697,
+      "grad_norm": 0.4864129915363767,
+      "learning_rate": 4.0858859640747435e-06,
+      "loss": 0.2619,
+      "step": 3610
+    },
+    {
+      "epoch": 1.2056761268781302,
+      "grad_norm": 0.4729174851706973,
+      "learning_rate": 4.083021489629418e-06,
+      "loss": 0.249,
+      "step": 3611
+    },
+    {
+      "epoch": 1.2060100166944907,
+      "grad_norm": 0.4619385078313871,
+      "learning_rate": 4.080157326585256e-06,
+      "loss": 0.2637,
+      "step": 3612
+    },
+    {
+      "epoch": 1.2063439065108514,
+      "grad_norm": 0.4576977862128267,
+      "learning_rate": 4.077293475914907e-06,
+      "loss": 0.2543,
+      "step": 3613
+    },
+    {
+      "epoch": 1.206677796327212,
+      "grad_norm": 0.45873252059908426,
+      "learning_rate": 4.074429938590924e-06,
+      "loss": 0.2497,
+      "step": 3614
+    },
+    {
+      "epoch": 1.2070116861435727,
+      "grad_norm": 0.48672006861244377,
+      "learning_rate": 4.071566715585748e-06,
+      "loss": 0.2604,
+      "step": 3615
+    },
+    {
+      "epoch": 1.2073455759599332,
+      "grad_norm": 0.496955851780668,
+      "learning_rate": 4.068703807871716e-06,
+      "loss": 0.2661,
+      "step": 3616
+    },
+    {
+      "epoch": 1.2076794657762937,
+      "grad_norm": 0.45837824600930605,
+      "learning_rate": 4.065841216421054e-06,
+      "loss": 0.2518,
+      "step": 3617
+    },
+    {
+      "epoch": 1.2080133555926544,
+      "grad_norm": 0.47011291805846767,
+      "learning_rate": 4.062978942205885e-06,
+      "loss": 0.2624,
+      "step": 3618
+    },
+    {
+      "epoch": 1.208347245409015,
+      "grad_norm": 0.47674536433805975,
+      "learning_rate": 4.060116986198223e-06,
+      "loss": 0.2491,
+      "step": 3619
+    },
+    {
+      "epoch": 1.2086811352253757,
+      "grad_norm": 0.4306380872542934,
+      "learning_rate": 4.057255349369974e-06,
+      "loss": 0.2371,
+      "step": 3620
+    },
+    {
+      "epoch": 1.2090150250417362,
+      "grad_norm": 0.4649101585573663,
+      "learning_rate": 4.054394032692933e-06,
+      "loss": 0.2427,
+      "step": 3621
+    },
+    {
+      "epoch": 1.209348914858097,
+      "grad_norm": 0.4767867075615579,
+      "learning_rate": 4.051533037138789e-06,
+      "loss": 0.2617,
+      "step": 3622
+    },
+    {
+      "epoch": 1.2096828046744574,
+      "grad_norm": 0.4682841397803673,
+      "learning_rate": 4.048672363679124e-06,
+      "loss": 0.2507,
+      "step": 3623
+    },
+    {
+      "epoch": 1.210016694490818,
+      "grad_norm": 0.46116940460738737,
+      "learning_rate": 4.045812013285403e-06,
+      "loss": 0.2456,
+      "step": 3624
+    },
+    {
+      "epoch": 1.2103505843071787,
+      "grad_norm": 0.4943352620622597,
+      "learning_rate": 4.04295198692899e-06,
+      "loss": 0.2621,
+      "step": 3625
+    },
+    {
+      "epoch": 1.2106844741235392,
+      "grad_norm": 0.4896744876042902,
+      "learning_rate": 4.040092285581132e-06,
+      "loss": 0.2648,
+      "step": 3626
+    },
+    {
+      "epoch": 1.2110183639399,
+      "grad_norm": 0.4795342737650419,
+      "learning_rate": 4.037232910212972e-06,
+      "loss": 0.2661,
+      "step": 3627
+    },
+    {
+      "epoch": 1.2113522537562604,
+      "grad_norm": 0.49843452667608645,
+      "learning_rate": 4.034373861795538e-06,
+      "loss": 0.2741,
+      "step": 3628
+    },
+    {
+      "epoch": 1.211686143572621,
+      "grad_norm": 0.4599908168268277,
+      "learning_rate": 4.031515141299748e-06,
+      "loss": 0.2489,
+      "step": 3629
+    },
+    {
+      "epoch": 1.2120200333889817,
+      "grad_norm": 0.46144400232972455,
+      "learning_rate": 4.02865674969641e-06,
+      "loss": 0.2553,
+      "step": 3630
+    },
+    {
+      "epoch": 1.2123539232053422,
+      "grad_norm": 0.4751290207650356,
+      "learning_rate": 4.025798687956215e-06,
+      "loss": 0.2601,
+      "step": 3631
+    },
+    {
+      "epoch": 1.212687813021703,
+      "grad_norm": 0.48360504037174834,
+      "learning_rate": 4.022940957049752e-06,
+      "loss": 0.2507,
+      "step": 3632
+    },
+    {
+      "epoch": 1.2130217028380634,
+      "grad_norm": 0.4797308858005336,
+      "learning_rate": 4.020083557947485e-06,
+      "loss": 0.2672,
+      "step": 3633
+    },
+    {
+      "epoch": 1.2133555926544242,
+      "grad_norm": 0.4716054060532315,
+      "learning_rate": 4.0172264916197765e-06,
+      "loss": 0.2661,
+      "step": 3634
+    },
+    {
+      "epoch": 1.2136894824707847,
+      "grad_norm": 0.5094697609555799,
+      "learning_rate": 4.014369759036869e-06,
+      "loss": 0.2555,
+      "step": 3635
+    },
+    {
+      "epoch": 1.2140233722871452,
+      "grad_norm": 0.4796501215757895,
+      "learning_rate": 4.011513361168895e-06,
+      "loss": 0.2541,
+      "step": 3636
+    },
+    {
+      "epoch": 1.214357262103506,
+      "grad_norm": 0.4702393171134311,
+      "learning_rate": 4.0086572989858745e-06,
+      "loss": 0.2497,
+      "step": 3637
+    },
+    {
+      "epoch": 1.2146911519198664,
+      "grad_norm": 0.4805136304112716,
+      "learning_rate": 4.005801573457709e-06,
+      "loss": 0.2505,
+      "step": 3638
+    },
+    {
+      "epoch": 1.2150250417362272,
+      "grad_norm": 0.45554258939623604,
+      "learning_rate": 4.0029461855541895e-06,
+      "loss": 0.2481,
+      "step": 3639
+    },
+    {
+      "epoch": 1.2153589315525877,
+      "grad_norm": 0.47483449161585783,
+      "learning_rate": 4.000091136244989e-06,
+      "loss": 0.2591,
+      "step": 3640
+    },
+    {
+      "epoch": 1.2156928213689482,
+      "grad_norm": 0.47532043008835506,
+      "learning_rate": 3.99723642649967e-06,
+      "loss": 0.2493,
+      "step": 3641
+    },
+    {
+      "epoch": 1.216026711185309,
+      "grad_norm": 0.4874941778333078,
+      "learning_rate": 3.994382057287674e-06,
+      "loss": 0.2531,
+      "step": 3642
+    },
+    {
+      "epoch": 1.2163606010016694,
+      "grad_norm": 0.4811989812134524,
+      "learning_rate": 3.991528029578332e-06,
+      "loss": 0.2581,
+      "step": 3643
+    },
+    {
+      "epoch": 1.2166944908180302,
+      "grad_norm": 0.49180530555194873,
+      "learning_rate": 3.988674344340859e-06,
+      "loss": 0.2528,
+      "step": 3644
+    },
+    {
+      "epoch": 1.2170283806343907,
+      "grad_norm": 0.5088022297191149,
+      "learning_rate": 3.985821002544348e-06,
+      "loss": 0.2642,
+      "step": 3645
+    },
+    {
+      "epoch": 1.2173622704507512,
+      "grad_norm": 0.4678194457672616,
+      "learning_rate": 3.982968005157783e-06,
+      "loss": 0.2493,
+      "step": 3646
+    },
+    {
+      "epoch": 1.217696160267112,
+      "grad_norm": 0.45957368740111687,
+      "learning_rate": 3.9801153531500234e-06,
+      "loss": 0.2511,
+      "step": 3647
+    },
+    {
+      "epoch": 1.2180300500834724,
+      "grad_norm": 0.4857632034768414,
+      "learning_rate": 3.9772630474898175e-06,
+      "loss": 0.2568,
+      "step": 3648
+    },
+    {
+      "epoch": 1.2183639398998332,
+      "grad_norm": 0.4597925608384798,
+      "learning_rate": 3.974411089145792e-06,
+      "loss": 0.2436,
+      "step": 3649
+    },
+    {
+      "epoch": 1.2186978297161937,
+      "grad_norm": 0.476455232321933,
+      "learning_rate": 3.971559479086459e-06,
+      "loss": 0.2572,
+      "step": 3650
+    },
+    {
+      "epoch": 1.2190317195325542,
+      "grad_norm": 0.461259512853758,
+      "learning_rate": 3.9687082182802076e-06,
+      "loss": 0.2553,
+      "step": 3651
+    },
+    {
+      "epoch": 1.219365609348915,
+      "grad_norm": 0.4618911260975359,
+      "learning_rate": 3.965857307695313e-06,
+      "loss": 0.2513,
+      "step": 3652
+    },
+    {
+      "epoch": 1.2196994991652754,
+      "grad_norm": 0.4843229810768652,
+      "learning_rate": 3.963006748299932e-06,
+      "loss": 0.2583,
+      "step": 3653
+    },
+    {
+      "epoch": 1.2200333889816362,
+      "grad_norm": 0.45661722846085745,
+      "learning_rate": 3.9601565410620956e-06,
+      "loss": 0.2548,
+      "step": 3654
+    },
+    {
+      "epoch": 1.2203672787979967,
+      "grad_norm": 0.49074304380574724,
+      "learning_rate": 3.957306686949723e-06,
+      "loss": 0.2582,
+      "step": 3655
+    },
+    {
+      "epoch": 1.2207011686143572,
+      "grad_norm": 0.4789195453332366,
+      "learning_rate": 3.954457186930608e-06,
+      "loss": 0.2542,
+      "step": 3656
+    },
+    {
+      "epoch": 1.221035058430718,
+      "grad_norm": 0.5141220855893336,
+      "learning_rate": 3.951608041972427e-06,
+      "loss": 0.2643,
+      "step": 3657
+    },
+    {
+      "epoch": 1.2213689482470784,
+      "grad_norm": 0.5065563915780893,
+      "learning_rate": 3.948759253042733e-06,
+      "loss": 0.2682,
+      "step": 3658
+    },
+    {
+      "epoch": 1.2217028380634392,
+      "grad_norm": 0.467374917482788,
+      "learning_rate": 3.945910821108963e-06,
+      "loss": 0.2582,
+      "step": 3659
+    },
+    {
+      "epoch": 1.2220367278797997,
+      "grad_norm": 0.4793764829632708,
+      "learning_rate": 3.9430627471384256e-06,
+      "loss": 0.2629,
+      "step": 3660
+    },
+    {
+      "epoch": 1.2223706176961602,
+      "grad_norm": 0.48743610905229234,
+      "learning_rate": 3.940215032098316e-06,
+      "loss": 0.2623,
+      "step": 3661
+    },
+    {
+      "epoch": 1.222704507512521,
+      "grad_norm": 0.4683392850826351,
+      "learning_rate": 3.937367676955703e-06,
+      "loss": 0.2635,
+      "step": 3662
+    },
+    {
+      "epoch": 1.2230383973288814,
+      "grad_norm": 0.48682637123928735,
+      "learning_rate": 3.934520682677532e-06,
+      "loss": 0.2572,
+      "step": 3663
+    },
+    {
+      "epoch": 1.2233722871452422,
+      "grad_norm": 0.4927688512179272,
+      "learning_rate": 3.931674050230628e-06,
+      "loss": 0.2584,
+      "step": 3664
+    },
+    {
+      "epoch": 1.2237061769616027,
+      "grad_norm": 0.49274864927139395,
+      "learning_rate": 3.928827780581693e-06,
+      "loss": 0.2628,
+      "step": 3665
+    },
+    {
+      "epoch": 1.2240400667779632,
+      "grad_norm": 0.5028360706722189,
+      "learning_rate": 3.925981874697306e-06,
+      "loss": 0.2643,
+      "step": 3666
+    },
+    {
+      "epoch": 1.224373956594324,
+      "grad_norm": 0.4864981535074828,
+      "learning_rate": 3.92313633354392e-06,
+      "loss": 0.2582,
+      "step": 3667
+    },
+    {
+      "epoch": 1.2247078464106844,
+      "grad_norm": 0.4879423339237642,
+      "learning_rate": 3.9202911580878685e-06,
+      "loss": 0.2549,
+      "step": 3668
+    },
+    {
+      "epoch": 1.2250417362270452,
+      "grad_norm": 0.46306791601752995,
+      "learning_rate": 3.917446349295356e-06,
+      "loss": 0.249,
+      "step": 3669
+    },
+    {
+      "epoch": 1.2253756260434057,
+      "grad_norm": 0.48296576079883535,
+      "learning_rate": 3.914601908132466e-06,
+      "loss": 0.2628,
+      "step": 3670
+    },
+    {
+      "epoch": 1.2257095158597662,
+      "grad_norm": 0.4729072215721062,
+      "learning_rate": 3.911757835565157e-06,
+      "loss": 0.2602,
+      "step": 3671
+    },
+    {
+      "epoch": 1.226043405676127,
+      "grad_norm": 0.4645915593897604,
+      "learning_rate": 3.90891413255926e-06,
+      "loss": 0.2477,
+      "step": 3672
+    },
+    {
+      "epoch": 1.2263772954924874,
+      "grad_norm": 0.4581634174837437,
+      "learning_rate": 3.906070800080484e-06,
+      "loss": 0.2459,
+      "step": 3673
+    },
+    {
+      "epoch": 1.2267111853088482,
+      "grad_norm": 0.4890338542308791,
+      "learning_rate": 3.903227839094406e-06,
+      "loss": 0.2552,
+      "step": 3674
+    },
+    {
+      "epoch": 1.2270450751252087,
+      "grad_norm": 0.48659434164307647,
+      "learning_rate": 3.900385250566485e-06,
+      "loss": 0.2461,
+      "step": 3675
+    },
+    {
+      "epoch": 1.2273789649415692,
+      "grad_norm": 0.4807354231187734,
+      "learning_rate": 3.897543035462047e-06,
+      "loss": 0.2679,
+      "step": 3676
+    },
+    {
+      "epoch": 1.22771285475793,
+      "grad_norm": 0.4684486785688895,
+      "learning_rate": 3.894701194746291e-06,
+      "loss": 0.2516,
+      "step": 3677
+    },
+    {
+      "epoch": 1.2280467445742904,
+      "grad_norm": 0.4719677716390586,
+      "learning_rate": 3.891859729384298e-06,
+      "loss": 0.2553,
+      "step": 3678
+    },
+    {
+      "epoch": 1.2283806343906511,
+      "grad_norm": 0.49045861272243246,
+      "learning_rate": 3.88901864034101e-06,
+      "loss": 0.2503,
+      "step": 3679
+    },
+    {
+      "epoch": 1.2287145242070117,
+      "grad_norm": 0.4635534715508029,
+      "learning_rate": 3.886177928581248e-06,
+      "loss": 0.2562,
+      "step": 3680
+    },
+    {
+      "epoch": 1.2290484140233722,
+      "grad_norm": 0.46275308103655693,
+      "learning_rate": 3.883337595069704e-06,
+      "loss": 0.245,
+      "step": 3681
+    },
+    {
+      "epoch": 1.229382303839733,
+      "grad_norm": 0.5007779383384727,
+      "learning_rate": 3.8804976407709375e-06,
+      "loss": 0.2645,
+      "step": 3682
+    },
+    {
+      "epoch": 1.2297161936560934,
+      "grad_norm": 0.5109035467398982,
+      "learning_rate": 3.877658066649385e-06,
+      "loss": 0.2555,
+      "step": 3683
+    },
+    {
+      "epoch": 1.2300500834724541,
+      "grad_norm": 0.4496152439534712,
+      "learning_rate": 3.874818873669348e-06,
+      "loss": 0.2432,
+      "step": 3684
+    },
+    {
+      "epoch": 1.2303839732888147,
+      "grad_norm": 0.4867377133616017,
+      "learning_rate": 3.871980062795006e-06,
+      "loss": 0.2531,
+      "step": 3685
+    },
+    {
+      "epoch": 1.2307178631051752,
+      "grad_norm": 0.4908452988396808,
+      "learning_rate": 3.869141634990399e-06,
+      "loss": 0.245,
+      "step": 3686
+    },
+    {
+      "epoch": 1.231051752921536,
+      "grad_norm": 0.48752219150640025,
+      "learning_rate": 3.866303591219449e-06,
+      "loss": 0.2501,
+      "step": 3687
+    },
+    {
+      "epoch": 1.2313856427378964,
+      "grad_norm": 0.48989329453930575,
+      "learning_rate": 3.8634659324459375e-06,
+      "loss": 0.2436,
+      "step": 3688
+    },
+    {
+      "epoch": 1.2317195325542571,
+      "grad_norm": 0.5078554041857629,
+      "learning_rate": 3.8606286596335195e-06,
+      "loss": 0.2664,
+      "step": 3689
+    },
+    {
+      "epoch": 1.2320534223706177,
+      "grad_norm": 0.5039975190422826,
+      "learning_rate": 3.857791773745717e-06,
+      "loss": 0.2596,
+      "step": 3690
+    },
+    {
+      "epoch": 1.2323873121869784,
+      "grad_norm": 0.48732844520108587,
+      "learning_rate": 3.854955275745922e-06,
+      "loss": 0.2543,
+      "step": 3691
+    },
+    {
+      "epoch": 1.232721202003339,
+      "grad_norm": 0.48902798123629637,
+      "learning_rate": 3.8521191665973965e-06,
+      "loss": 0.2607,
+      "step": 3692
+    },
+    {
+      "epoch": 1.2330550918196994,
+      "grad_norm": 0.5159922055332969,
+      "learning_rate": 3.849283447263266e-06,
+      "loss": 0.2633,
+      "step": 3693
+    },
+    {
+      "epoch": 1.2333889816360601,
+      "grad_norm": 0.47617880917467653,
+      "learning_rate": 3.846448118706525e-06,
+      "loss": 0.2549,
+      "step": 3694
+    },
+    {
+      "epoch": 1.2337228714524207,
+      "grad_norm": 0.46466789770829764,
+      "learning_rate": 3.843613181890042e-06,
+      "loss": 0.2644,
+      "step": 3695
+    },
+    {
+      "epoch": 1.2340567612687814,
+      "grad_norm": 0.4687410657185262,
+      "learning_rate": 3.840778637776541e-06,
+      "loss": 0.2533,
+      "step": 3696
+    },
+    {
+      "epoch": 1.234390651085142,
+      "grad_norm": 0.48336803877706347,
+      "learning_rate": 3.837944487328624e-06,
+      "loss": 0.2573,
+      "step": 3697
+    },
+    {
+      "epoch": 1.2347245409015024,
+      "grad_norm": 0.47613808306091354,
+      "learning_rate": 3.835110731508749e-06,
+      "loss": 0.2688,
+      "step": 3698
+    },
+    {
+      "epoch": 1.2350584307178631,
+      "grad_norm": 0.4634873033581562,
+      "learning_rate": 3.832277371279247e-06,
+      "loss": 0.2608,
+      "step": 3699
+    },
+    {
+      "epoch": 1.2353923205342237,
+      "grad_norm": 0.4635834234557756,
+      "learning_rate": 3.8294444076023115e-06,
+      "loss": 0.2637,
+      "step": 3700
+    },
+    {
+      "epoch": 1.2357262103505844,
+      "grad_norm": 0.4685714709529008,
+      "learning_rate": 3.826611841440005e-06,
+      "loss": 0.2646,
+      "step": 3701
+    },
+    {
+      "epoch": 1.236060100166945,
+      "grad_norm": 0.48483185040961485,
+      "learning_rate": 3.823779673754247e-06,
+      "loss": 0.2592,
+      "step": 3702
+    },
+    {
+      "epoch": 1.2363939899833056,
+      "grad_norm": 0.4919283781279512,
+      "learning_rate": 3.820947905506831e-06,
+      "loss": 0.2551,
+      "step": 3703
+    },
+    {
+      "epoch": 1.2367278797996661,
+      "grad_norm": 0.4567777007634364,
+      "learning_rate": 3.818116537659412e-06,
+      "loss": 0.2493,
+      "step": 3704
+    },
+    {
+      "epoch": 1.2370617696160267,
+      "grad_norm": 0.47426084636607374,
+      "learning_rate": 3.8152855711735045e-06,
+      "loss": 0.2535,
+      "step": 3705
+    },
+    {
+      "epoch": 1.2373956594323874,
+      "grad_norm": 0.47129170531389997,
+      "learning_rate": 3.8124550070104938e-06,
+      "loss": 0.252,
+      "step": 3706
+    },
+    {
+      "epoch": 1.237729549248748,
+      "grad_norm": 0.5094966829944446,
+      "learning_rate": 3.809624846131621e-06,
+      "loss": 0.2679,
+      "step": 3707
+    },
+    {
+      "epoch": 1.2380634390651086,
+      "grad_norm": 0.47831952684880147,
+      "learning_rate": 3.8067950894979965e-06,
+      "loss": 0.2536,
+      "step": 3708
+    },
+    {
+      "epoch": 1.2383973288814691,
+      "grad_norm": 0.46323241487592437,
+      "learning_rate": 3.8039657380705895e-06,
+      "loss": 0.2421,
+      "step": 3709
+    },
+    {
+      "epoch": 1.2387312186978297,
+      "grad_norm": 0.5001404269594205,
+      "learning_rate": 3.801136792810235e-06,
+      "loss": 0.2584,
+      "step": 3710
+    },
+    {
+      "epoch": 1.2390651085141904,
+      "grad_norm": 0.46631652265781004,
+      "learning_rate": 3.7983082546776244e-06,
+      "loss": 0.2481,
+      "step": 3711
+    },
+    {
+      "epoch": 1.239398998330551,
+      "grad_norm": 0.5108204180183292,
+      "learning_rate": 3.7954801246333196e-06,
+      "loss": 0.2613,
+      "step": 3712
+    },
+    {
+      "epoch": 1.2397328881469116,
+      "grad_norm": 0.4550844442739663,
+      "learning_rate": 3.7926524036377366e-06,
+      "loss": 0.2449,
+      "step": 3713
+    },
+    {
+      "epoch": 1.2400667779632721,
+      "grad_norm": 0.5077921240031825,
+      "learning_rate": 3.789825092651154e-06,
+      "loss": 0.2524,
+      "step": 3714
+    },
+    {
+      "epoch": 1.2404006677796326,
+      "grad_norm": 0.46296886178337027,
+      "learning_rate": 3.7869981926337146e-06,
+      "loss": 0.2553,
+      "step": 3715
+    },
+    {
+      "epoch": 1.2407345575959934,
+      "grad_norm": 0.46801849572183435,
+      "learning_rate": 3.784171704545416e-06,
+      "loss": 0.2541,
+      "step": 3716
+    },
+    {
+      "epoch": 1.241068447412354,
+      "grad_norm": 0.5137306098581467,
+      "learning_rate": 3.781345629346121e-06,
+      "loss": 0.2599,
+      "step": 3717
+    },
+    {
+      "epoch": 1.2414023372287146,
+      "grad_norm": 0.4792578434102303,
+      "learning_rate": 3.778519967995548e-06,
+      "loss": 0.2518,
+      "step": 3718
+    },
+    {
+      "epoch": 1.2417362270450751,
+      "grad_norm": 0.47923765816390734,
+      "learning_rate": 3.775694721453278e-06,
+      "loss": 0.2553,
+      "step": 3719
+    },
+    {
+      "epoch": 1.2420701168614356,
+      "grad_norm": 0.4598482098645983,
+      "learning_rate": 3.7728698906787527e-06,
+      "loss": 0.2507,
+      "step": 3720
+    },
+    {
+      "epoch": 1.2424040066777964,
+      "grad_norm": 0.48280276667184674,
+      "learning_rate": 3.7700454766312662e-06,
+      "loss": 0.2556,
+      "step": 3721
+    },
+    {
+      "epoch": 1.242737896494157,
+      "grad_norm": 0.4645003796561382,
+      "learning_rate": 3.767221480269978e-06,
+      "loss": 0.2574,
+      "step": 3722
+    },
+    {
+      "epoch": 1.2430717863105176,
+      "grad_norm": 0.5085309673563764,
+      "learning_rate": 3.7643979025538995e-06,
+      "loss": 0.2765,
+      "step": 3723
+    },
+    {
+      "epoch": 1.2434056761268781,
+      "grad_norm": 0.49606444439484115,
+      "learning_rate": 3.7615747444419064e-06,
+      "loss": 0.264,
+      "step": 3724
+    },
+    {
+      "epoch": 1.2437395659432386,
+      "grad_norm": 0.4686552716172875,
+      "learning_rate": 3.758752006892726e-06,
+      "loss": 0.2604,
+      "step": 3725
+    },
+    {
+      "epoch": 1.2440734557595994,
+      "grad_norm": 0.44565378958851043,
+      "learning_rate": 3.7559296908649467e-06,
+      "loss": 0.2445,
+      "step": 3726
+    },
+    {
+      "epoch": 1.2444073455759599,
+      "grad_norm": 0.463531686825008,
+      "learning_rate": 3.7531077973170117e-06,
+      "loss": 0.263,
+      "step": 3727
+    },
+    {
+      "epoch": 1.2447412353923206,
+      "grad_norm": 0.47447247267272186,
+      "learning_rate": 3.750286327207221e-06,
+      "loss": 0.257,
+      "step": 3728
+    },
+    {
+      "epoch": 1.2450751252086811,
+      "grad_norm": 0.4698037246641178,
+      "learning_rate": 3.7474652814937332e-06,
+      "loss": 0.2682,
+      "step": 3729
+    },
+    {
+      "epoch": 1.2454090150250416,
+      "grad_norm": 0.46342390468763117,
+      "learning_rate": 3.74464466113456e-06,
+      "loss": 0.2574,
+      "step": 3730
+    },
+    {
+      "epoch": 1.2457429048414024,
+      "grad_norm": 0.4637615791346367,
+      "learning_rate": 3.741824467087569e-06,
+      "loss": 0.2513,
+      "step": 3731
+    },
+    {
+      "epoch": 1.2460767946577629,
+      "grad_norm": 0.4827087842383037,
+      "learning_rate": 3.7390047003104835e-06,
+      "loss": 0.2552,
+      "step": 3732
+    },
+    {
+      "epoch": 1.2464106844741236,
+      "grad_norm": 0.48500123490695307,
+      "learning_rate": 3.736185361760882e-06,
+      "loss": 0.2587,
+      "step": 3733
+    },
+    {
+      "epoch": 1.2467445742904841,
+      "grad_norm": 0.4653584888722768,
+      "learning_rate": 3.7333664523961975e-06,
+      "loss": 0.247,
+      "step": 3734
+    },
+    {
+      "epoch": 1.2470784641068446,
+      "grad_norm": 0.4782359895967455,
+      "learning_rate": 3.730547973173717e-06,
+      "loss": 0.2537,
+      "step": 3735
+    },
+    {
+      "epoch": 1.2474123539232054,
+      "grad_norm": 0.46572037202735517,
+      "learning_rate": 3.72772992505058e-06,
+      "loss": 0.2477,
+      "step": 3736
+    },
+    {
+      "epoch": 1.2477462437395659,
+      "grad_norm": 0.48264297123035715,
+      "learning_rate": 3.724912308983783e-06,
+      "loss": 0.2516,
+      "step": 3737
+    },
+    {
+      "epoch": 1.2480801335559266,
+      "grad_norm": 0.49384545967047555,
+      "learning_rate": 3.7220951259301742e-06,
+      "loss": 0.2554,
+      "step": 3738
+    },
+    {
+      "epoch": 1.2484140233722871,
+      "grad_norm": 0.4508161671593082,
+      "learning_rate": 3.7192783768464523e-06,
+      "loss": 0.2457,
+      "step": 3739
+    },
+    {
+      "epoch": 1.2487479131886476,
+      "grad_norm": 0.5045807873671156,
+      "learning_rate": 3.7164620626891724e-06,
+      "loss": 0.2632,
+      "step": 3740
+    },
+    {
+      "epoch": 1.2490818030050084,
+      "grad_norm": 0.4677630583329784,
+      "learning_rate": 3.7136461844147392e-06,
+      "loss": 0.2471,
+      "step": 3741
+    },
+    {
+      "epoch": 1.2494156928213689,
+      "grad_norm": 0.4664220574321503,
+      "learning_rate": 3.710830742979412e-06,
+      "loss": 0.2548,
+      "step": 3742
+    },
+    {
+      "epoch": 1.2497495826377296,
+      "grad_norm": 0.44887323103034327,
+      "learning_rate": 3.708015739339298e-06,
+      "loss": 0.2374,
+      "step": 3743
+    },
+    {
+      "epoch": 1.2500834724540901,
+      "grad_norm": 0.4747471190564493,
+      "learning_rate": 3.7052011744503593e-06,
+      "loss": 0.2532,
+      "step": 3744
+    },
+    {
+      "epoch": 1.2504173622704506,
+      "grad_norm": 0.49238753495751597,
+      "learning_rate": 3.7023870492684055e-06,
+      "loss": 0.2625,
+      "step": 3745
+    },
+    {
+      "epoch": 1.2507512520868114,
+      "grad_norm": 0.4957559228031111,
+      "learning_rate": 3.6995733647491016e-06,
+      "loss": 0.2482,
+      "step": 3746
+    },
+    {
+      "epoch": 1.2510851419031719,
+      "grad_norm": 0.4668245989658274,
+      "learning_rate": 3.6967601218479604e-06,
+      "loss": 0.2567,
+      "step": 3747
+    },
+    {
+      "epoch": 1.2514190317195326,
+      "grad_norm": 0.45259997242360783,
+      "learning_rate": 3.6939473215203424e-06,
+      "loss": 0.2451,
+      "step": 3748
+    },
+    {
+      "epoch": 1.2517529215358931,
+      "grad_norm": 0.47024700543180625,
+      "learning_rate": 3.6911349647214623e-06,
+      "loss": 0.2538,
+      "step": 3749
+    },
+    {
+      "epoch": 1.2520868113522536,
+      "grad_norm": 0.46333131652481696,
+      "learning_rate": 3.6883230524063802e-06,
+      "loss": 0.252,
+      "step": 3750
+    },
+    {
+      "epoch": 1.2524207011686144,
+      "grad_norm": 0.4863973468707675,
+      "learning_rate": 3.685511585530008e-06,
+      "loss": 0.2642,
+      "step": 3751
+    },
+    {
+      "epoch": 1.2527545909849749,
+      "grad_norm": 0.5055209175329437,
+      "learning_rate": 3.682700565047105e-06,
+      "loss": 0.2586,
+      "step": 3752
+    },
+    {
+      "epoch": 1.2530884808013356,
+      "grad_norm": 0.4776594147925309,
+      "learning_rate": 3.6798899919122773e-06,
+      "loss": 0.2609,
+      "step": 3753
+    },
+    {
+      "epoch": 1.2534223706176961,
+      "grad_norm": 0.49083947895253144,
+      "learning_rate": 3.677079867079985e-06,
+      "loss": 0.248,
+      "step": 3754
+    },
+    {
+      "epoch": 1.2537562604340566,
+      "grad_norm": 0.4978219451787454,
+      "learning_rate": 3.6742701915045285e-06,
+      "loss": 0.259,
+      "step": 3755
+    },
+    {
+      "epoch": 1.2540901502504174,
+      "grad_norm": 0.4904675584810712,
+      "learning_rate": 3.6714609661400624e-06,
+      "loss": 0.2615,
+      "step": 3756
+    },
+    {
+      "epoch": 1.254424040066778,
+      "grad_norm": 0.48630744784053886,
+      "learning_rate": 3.6686521919405814e-06,
+      "loss": 0.2448,
+      "step": 3757
+    },
+    {
+      "epoch": 1.2547579298831386,
+      "grad_norm": 0.4881711131617577,
+      "learning_rate": 3.665843869859934e-06,
+      "loss": 0.2497,
+      "step": 3758
+    },
+    {
+      "epoch": 1.2550918196994991,
+      "grad_norm": 0.5210699242291336,
+      "learning_rate": 3.6630360008518094e-06,
+      "loss": 0.2709,
+      "step": 3759
+    },
+    {
+      "epoch": 1.2554257095158596,
+      "grad_norm": 0.48860786392370525,
+      "learning_rate": 3.6602285858697474e-06,
+      "loss": 0.2529,
+      "step": 3760
+    },
+    {
+      "epoch": 1.2557595993322204,
+      "grad_norm": 0.4640297254773273,
+      "learning_rate": 3.6574216258671287e-06,
+      "loss": 0.2486,
+      "step": 3761
+    },
+    {
+      "epoch": 1.256093489148581,
+      "grad_norm": 0.4804353979281327,
+      "learning_rate": 3.6546151217971836e-06,
+      "loss": 0.2652,
+      "step": 3762
+    },
+    {
+      "epoch": 1.2564273789649416,
+      "grad_norm": 0.499388576492226,
+      "learning_rate": 3.6518090746129884e-06,
+      "loss": 0.2532,
+      "step": 3763
+    },
+    {
+      "epoch": 1.2567612687813021,
+      "grad_norm": 0.4667244932750343,
+      "learning_rate": 3.649003485267461e-06,
+      "loss": 0.2562,
+      "step": 3764
+    },
+    {
+      "epoch": 1.2570951585976629,
+      "grad_norm": 0.46947893457973366,
+      "learning_rate": 3.646198354713365e-06,
+      "loss": 0.2552,
+      "step": 3765
+    },
+    {
+      "epoch": 1.2574290484140234,
+      "grad_norm": 0.4704153851536768,
+      "learning_rate": 3.643393683903308e-06,
+      "loss": 0.2564,
+      "step": 3766
+    },
+    {
+      "epoch": 1.257762938230384,
+      "grad_norm": 0.46744867165643705,
+      "learning_rate": 3.640589473789742e-06,
+      "loss": 0.2521,
+      "step": 3767
+    },
+    {
+      "epoch": 1.2580968280467446,
+      "grad_norm": 0.4765780036389642,
+      "learning_rate": 3.6377857253249616e-06,
+      "loss": 0.2575,
+      "step": 3768
+    },
+    {
+      "epoch": 1.2584307178631051,
+      "grad_norm": 0.4818870341694187,
+      "learning_rate": 3.634982439461106e-06,
+      "loss": 0.2563,
+      "step": 3769
+    },
+    {
+      "epoch": 1.2587646076794659,
+      "grad_norm": 0.4755175895991557,
+      "learning_rate": 3.6321796171501546e-06,
+      "loss": 0.258,
+      "step": 3770
+    },
+    {
+      "epoch": 1.2590984974958264,
+      "grad_norm": 0.5101778552459092,
+      "learning_rate": 3.629377259343934e-06,
+      "loss": 0.2556,
+      "step": 3771
+    },
+    {
+      "epoch": 1.259432387312187,
+      "grad_norm": 0.468149124733761,
+      "learning_rate": 3.6265753669941104e-06,
+      "loss": 0.254,
+      "step": 3772
+    },
+    {
+      "epoch": 1.2597662771285476,
+      "grad_norm": 0.4772590125954835,
+      "learning_rate": 3.623773941052191e-06,
+      "loss": 0.2621,
+      "step": 3773
+    },
+    {
+      "epoch": 1.2601001669449081,
+      "grad_norm": 0.49637655113035284,
+      "learning_rate": 3.6209729824695262e-06,
+      "loss": 0.2597,
+      "step": 3774
+    },
+    {
+      "epoch": 1.2604340567612689,
+      "grad_norm": 0.4731908875417793,
+      "learning_rate": 3.618172492197305e-06,
+      "loss": 0.2575,
+      "step": 3775
+    },
+    {
+      "epoch": 1.2607679465776294,
+      "grad_norm": 0.47861953004079943,
+      "learning_rate": 3.6153724711865623e-06,
+      "loss": 0.2444,
+      "step": 3776
+    },
+    {
+      "epoch": 1.26110183639399,
+      "grad_norm": 0.47362053128045006,
+      "learning_rate": 3.6125729203881676e-06,
+      "loss": 0.2589,
+      "step": 3777
+    },
+    {
+      "epoch": 1.2614357262103506,
+      "grad_norm": 0.4920726650859917,
+      "learning_rate": 3.609773840752837e-06,
+      "loss": 0.2443,
+      "step": 3778
+    },
+    {
+      "epoch": 1.2617696160267111,
+      "grad_norm": 0.5115916541065607,
+      "learning_rate": 3.6069752332311203e-06,
+      "loss": 0.2565,
+      "step": 3779
+    },
+    {
+      "epoch": 1.2621035058430718,
+      "grad_norm": 0.48518662781506294,
+      "learning_rate": 3.604177098773414e-06,
+      "loss": 0.2559,
+      "step": 3780
+    },
+    {
+      "epoch": 1.2624373956594324,
+      "grad_norm": 0.49207793712306785,
+      "learning_rate": 3.6013794383299474e-06,
+      "loss": 0.2464,
+      "step": 3781
+    },
+    {
+      "epoch": 1.262771285475793,
+      "grad_norm": 0.49285953554778805,
+      "learning_rate": 3.598582252850792e-06,
+      "loss": 0.2662,
+      "step": 3782
+    },
+    {
+      "epoch": 1.2631051752921536,
+      "grad_norm": 0.4742022592910343,
+      "learning_rate": 3.595785543285858e-06,
+      "loss": 0.2574,
+      "step": 3783
+    },
+    {
+      "epoch": 1.2634390651085141,
+      "grad_norm": 0.4843316198145805,
+      "learning_rate": 3.592989310584893e-06,
+      "loss": 0.2546,
+      "step": 3784
+    },
+    {
+      "epoch": 1.2637729549248748,
+      "grad_norm": 0.5006424298152526,
+      "learning_rate": 3.5901935556974837e-06,
+      "loss": 0.2558,
+      "step": 3785
+    },
+    {
+      "epoch": 1.2641068447412354,
+      "grad_norm": 0.5115192241851619,
+      "learning_rate": 3.5873982795730533e-06,
+      "loss": 0.2546,
+      "step": 3786
+    },
+    {
+      "epoch": 1.264440734557596,
+      "grad_norm": 0.4690845144604071,
+      "learning_rate": 3.5846034831608617e-06,
+      "loss": 0.2585,
+      "step": 3787
+    },
+    {
+      "epoch": 1.2647746243739566,
+      "grad_norm": 0.46874029134483675,
+      "learning_rate": 3.581809167410012e-06,
+      "loss": 0.2587,
+      "step": 3788
+    },
+    {
+      "epoch": 1.2651085141903171,
+      "grad_norm": 0.4821928011566208,
+      "learning_rate": 3.5790153332694344e-06,
+      "loss": 0.2557,
+      "step": 3789
+    },
+    {
+      "epoch": 1.2654424040066778,
+      "grad_norm": 0.4821872906402347,
+      "learning_rate": 3.576221981687904e-06,
+      "loss": 0.2528,
+      "step": 3790
+    },
+    {
+      "epoch": 1.2657762938230384,
+      "grad_norm": 0.47802516773594705,
+      "learning_rate": 3.5734291136140265e-06,
+      "loss": 0.245,
+      "step": 3791
+    },
+    {
+      "epoch": 1.266110183639399,
+      "grad_norm": 0.48433580177679125,
+      "learning_rate": 3.5706367299962474e-06,
+      "loss": 0.2593,
+      "step": 3792
+    },
+    {
+      "epoch": 1.2664440734557596,
+      "grad_norm": 0.4817756413908495,
+      "learning_rate": 3.5678448317828427e-06,
+      "loss": 0.2544,
+      "step": 3793
+    },
+    {
+      "epoch": 1.2667779632721201,
+      "grad_norm": 0.46097006408792046,
+      "learning_rate": 3.56505341992193e-06,
+      "loss": 0.2606,
+      "step": 3794
+    },
+    {
+      "epoch": 1.2671118530884808,
+      "grad_norm": 0.48395014391311614,
+      "learning_rate": 3.5622624953614554e-06,
+      "loss": 0.2513,
+      "step": 3795
+    },
+    {
+      "epoch": 1.2674457429048414,
+      "grad_norm": 0.46395577457447396,
+      "learning_rate": 3.559472059049203e-06,
+      "loss": 0.259,
+      "step": 3796
+    },
+    {
+      "epoch": 1.267779632721202,
+      "grad_norm": 0.4761899340367128,
+      "learning_rate": 3.5566821119327927e-06,
+      "loss": 0.2594,
+      "step": 3797
+    },
+    {
+      "epoch": 1.2681135225375626,
+      "grad_norm": 0.47662500639113065,
+      "learning_rate": 3.553892654959674e-06,
+      "loss": 0.2556,
+      "step": 3798
+    },
+    {
+      "epoch": 1.268447412353923,
+      "grad_norm": 0.48114001613873425,
+      "learning_rate": 3.5511036890771337e-06,
+      "loss": 0.2525,
+      "step": 3799
+    },
+    {
+      "epoch": 1.2687813021702838,
+      "grad_norm": 0.47445745613046064,
+      "learning_rate": 3.5483152152322884e-06,
+      "loss": 0.2497,
+      "step": 3800
+    },
+    {
+      "epoch": 1.2691151919866444,
+      "grad_norm": 0.4642481614762914,
+      "learning_rate": 3.5455272343720915e-06,
+      "loss": 0.2492,
+      "step": 3801
+    },
+    {
+      "epoch": 1.269449081803005,
+      "grad_norm": 0.4886275621892356,
+      "learning_rate": 3.5427397474433238e-06,
+      "loss": 0.2622,
+      "step": 3802
+    },
+    {
+      "epoch": 1.2697829716193656,
+      "grad_norm": 0.4814943994329635,
+      "learning_rate": 3.539952755392605e-06,
+      "loss": 0.2611,
+      "step": 3803
+    },
+    {
+      "epoch": 1.270116861435726,
+      "grad_norm": 0.5124373092528955,
+      "learning_rate": 3.537166259166379e-06,
+      "loss": 0.2672,
+      "step": 3804
+    },
+    {
+      "epoch": 1.2704507512520868,
+      "grad_norm": 0.48220752650821735,
+      "learning_rate": 3.53438025971093e-06,
+      "loss": 0.2518,
+      "step": 3805
+    },
+    {
+      "epoch": 1.2707846410684474,
+      "grad_norm": 0.5061395380975122,
+      "learning_rate": 3.5315947579723677e-06,
+      "loss": 0.2565,
+      "step": 3806
+    },
+    {
+      "epoch": 1.271118530884808,
+      "grad_norm": 0.4736997358392899,
+      "learning_rate": 3.5288097548966314e-06,
+      "loss": 0.2499,
+      "step": 3807
+    },
+    {
+      "epoch": 1.2714524207011686,
+      "grad_norm": 0.4528023278436743,
+      "learning_rate": 3.5260252514294974e-06,
+      "loss": 0.2483,
+      "step": 3808
+    },
+    {
+      "epoch": 1.271786310517529,
+      "grad_norm": 0.48927238142774493,
+      "learning_rate": 3.5232412485165657e-06,
+      "loss": 0.2469,
+      "step": 3809
+    },
+    {
+      "epoch": 1.2721202003338898,
+      "grad_norm": 0.48474314032671006,
+      "learning_rate": 3.520457747103272e-06,
+      "loss": 0.2582,
+      "step": 3810
+    },
+    {
+      "epoch": 1.2724540901502503,
+      "grad_norm": 0.4928304403646798,
+      "learning_rate": 3.5176747481348774e-06,
+      "loss": 0.2536,
+      "step": 3811
+    },
+    {
+      "epoch": 1.272787979966611,
+      "grad_norm": 0.4837666972368579,
+      "learning_rate": 3.514892252556474e-06,
+      "loss": 0.2507,
+      "step": 3812
+    },
+    {
+      "epoch": 1.2731218697829716,
+      "grad_norm": 0.4958396190955158,
+      "learning_rate": 3.512110261312981e-06,
+      "loss": 0.2468,
+      "step": 3813
+    },
+    {
+      "epoch": 1.273455759599332,
+      "grad_norm": 0.5058280442476211,
+      "learning_rate": 3.5093287753491515e-06,
+      "loss": 0.2502,
+      "step": 3814
+    },
+    {
+      "epoch": 1.2737896494156928,
+      "grad_norm": 0.485549474214428,
+      "learning_rate": 3.506547795609564e-06,
+      "loss": 0.2586,
+      "step": 3815
+    },
+    {
+      "epoch": 1.2741235392320533,
+      "grad_norm": 0.48880726862090473,
+      "learning_rate": 3.503767323038622e-06,
+      "loss": 0.2653,
+      "step": 3816
+    },
+    {
+      "epoch": 1.274457429048414,
+      "grad_norm": 0.4945805332996736,
+      "learning_rate": 3.500987358580562e-06,
+      "loss": 0.2555,
+      "step": 3817
+    },
+    {
+      "epoch": 1.2747913188647746,
+      "grad_norm": 0.5051361028319796,
+      "learning_rate": 3.498207903179444e-06,
+      "loss": 0.2512,
+      "step": 3818
+    },
+    {
+      "epoch": 1.275125208681135,
+      "grad_norm": 0.4999966113740608,
+      "learning_rate": 3.495428957779158e-06,
+      "loss": 0.2536,
+      "step": 3819
+    },
+    {
+      "epoch": 1.2754590984974958,
+      "grad_norm": 0.4911157870636642,
+      "learning_rate": 3.4926505233234177e-06,
+      "loss": 0.2377,
+      "step": 3820
+    },
+    {
+      "epoch": 1.2757929883138563,
+      "grad_norm": 0.4876051914444362,
+      "learning_rate": 3.4898726007557655e-06,
+      "loss": 0.2412,
+      "step": 3821
+    },
+    {
+      "epoch": 1.276126878130217,
+      "grad_norm": 0.46704270304446516,
+      "learning_rate": 3.4870951910195716e-06,
+      "loss": 0.2487,
+      "step": 3822
+    },
+    {
+      "epoch": 1.2764607679465776,
+      "grad_norm": 0.4987264502641984,
+      "learning_rate": 3.484318295058028e-06,
+      "loss": 0.2683,
+      "step": 3823
+    },
+    {
+      "epoch": 1.276794657762938,
+      "grad_norm": 0.4680191752398623,
+      "learning_rate": 3.481541913814156e-06,
+      "loss": 0.2532,
+      "step": 3824
+    },
+    {
+      "epoch": 1.2771285475792988,
+      "grad_norm": 0.5030523340855039,
+      "learning_rate": 3.478766048230797e-06,
+      "loss": 0.2623,
+      "step": 3825
+    },
+    {
+      "epoch": 1.2774624373956596,
+      "grad_norm": 0.5043257068826584,
+      "learning_rate": 3.4759906992506244e-06,
+      "loss": 0.2525,
+      "step": 3826
+    },
+    {
+      "epoch": 1.27779632721202,
+      "grad_norm": 0.48507863179567795,
+      "learning_rate": 3.47321586781613e-06,
+      "loss": 0.2423,
+      "step": 3827
+    },
+    {
+      "epoch": 1.2781302170283806,
+      "grad_norm": 0.46702018179111776,
+      "learning_rate": 3.4704415548696335e-06,
+      "loss": 0.2542,
+      "step": 3828
+    },
+    {
+      "epoch": 1.278464106844741,
+      "grad_norm": 0.49936065251490036,
+      "learning_rate": 3.467667761353275e-06,
+      "loss": 0.2575,
+      "step": 3829
+    },
+    {
+      "epoch": 1.2787979966611018,
+      "grad_norm": 0.512338567622577,
+      "learning_rate": 3.464894488209022e-06,
+      "loss": 0.2697,
+      "step": 3830
+    },
+    {
+      "epoch": 1.2791318864774626,
+      "grad_norm": 0.5131462066967339,
+      "learning_rate": 3.462121736378665e-06,
+      "loss": 0.258,
+      "step": 3831
+    },
+    {
+      "epoch": 1.279465776293823,
+      "grad_norm": 0.46351360203488806,
+      "learning_rate": 3.4593495068038142e-06,
+      "loss": 0.2486,
+      "step": 3832
+    },
+    {
+      "epoch": 1.2797996661101836,
+      "grad_norm": 0.5211722380183532,
+      "learning_rate": 3.4565778004259064e-06,
+      "loss": 0.2586,
+      "step": 3833
+    },
+    {
+      "epoch": 1.2801335559265443,
+      "grad_norm": 0.4632371998435844,
+      "learning_rate": 3.4538066181861964e-06,
+      "loss": 0.2548,
+      "step": 3834
+    },
+    {
+      "epoch": 1.2804674457429048,
+      "grad_norm": 0.4656699202628051,
+      "learning_rate": 3.451035961025765e-06,
+      "loss": 0.2565,
+      "step": 3835
+    },
+    {
+      "epoch": 1.2808013355592656,
+      "grad_norm": 0.470991662126532,
+      "learning_rate": 3.448265829885512e-06,
+      "loss": 0.2486,
+      "step": 3836
+    },
+    {
+      "epoch": 1.281135225375626,
+      "grad_norm": 0.48334235313753865,
+      "learning_rate": 3.445496225706161e-06,
+      "loss": 0.2589,
+      "step": 3837
+    },
+    {
+      "epoch": 1.2814691151919866,
+      "grad_norm": 0.5173613115389184,
+      "learning_rate": 3.442727149428252e-06,
+      "loss": 0.2576,
+      "step": 3838
+    },
+    {
+      "epoch": 1.2818030050083473,
+      "grad_norm": 0.4760812509609349,
+      "learning_rate": 3.439958601992153e-06,
+      "loss": 0.2441,
+      "step": 3839
+    },
+    {
+      "epoch": 1.2821368948247078,
+      "grad_norm": 0.4941461315916647,
+      "learning_rate": 3.4371905843380483e-06,
+      "loss": 0.2575,
+      "step": 3840
+    },
+    {
+      "epoch": 1.2824707846410686,
+      "grad_norm": 0.4789829364383051,
+      "learning_rate": 3.4344230974059388e-06,
+      "loss": 0.2482,
+      "step": 3841
+    },
+    {
+      "epoch": 1.282804674457429,
+      "grad_norm": 0.48395489067099784,
+      "learning_rate": 3.431656142135652e-06,
+      "loss": 0.2444,
+      "step": 3842
+    },
+    {
+      "epoch": 1.2831385642737896,
+      "grad_norm": 0.4956790796537504,
+      "learning_rate": 3.428889719466829e-06,
+      "loss": 0.2488,
+      "step": 3843
+    },
+    {
+      "epoch": 1.2834724540901503,
+      "grad_norm": 0.5275068111601888,
+      "learning_rate": 3.426123830338936e-06,
+      "loss": 0.2631,
+      "step": 3844
+    },
+    {
+      "epoch": 1.2838063439065108,
+      "grad_norm": 0.45159366470034157,
+      "learning_rate": 3.4233584756912513e-06,
+      "loss": 0.2528,
+      "step": 3845
+    },
+    {
+      "epoch": 1.2841402337228716,
+      "grad_norm": 0.47057795518480633,
+      "learning_rate": 3.420593656462877e-06,
+      "loss": 0.2366,
+      "step": 3846
+    },
+    {
+      "epoch": 1.284474123539232,
+      "grad_norm": 0.5048488708158583,
+      "learning_rate": 3.4178293735927298e-06,
+      "loss": 0.2602,
+      "step": 3847
+    },
+    {
+      "epoch": 1.2848080133555926,
+      "grad_norm": 0.5212158059569421,
+      "learning_rate": 3.415065628019547e-06,
+      "loss": 0.2703,
+      "step": 3848
+    },
+    {
+      "epoch": 1.2851419031719533,
+      "grad_norm": 0.4790696435752834,
+      "learning_rate": 3.4123024206818846e-06,
+      "loss": 0.2504,
+      "step": 3849
+    },
+    {
+      "epoch": 1.2854757929883138,
+      "grad_norm": 0.49559656507324196,
+      "learning_rate": 3.4095397525181084e-06,
+      "loss": 0.2484,
+      "step": 3850
+    },
+    {
+      "epoch": 1.2858096828046746,
+      "grad_norm": 0.47237647653819304,
+      "learning_rate": 3.406777624466411e-06,
+      "loss": 0.2608,
+      "step": 3851
+    },
+    {
+      "epoch": 1.286143572621035,
+      "grad_norm": 0.47879779775233733,
+      "learning_rate": 3.4040160374647936e-06,
+      "loss": 0.2627,
+      "step": 3852
+    },
+    {
+      "epoch": 1.2864774624373956,
+      "grad_norm": 0.48187259093331325,
+      "learning_rate": 3.401254992451079e-06,
+      "loss": 0.2476,
+      "step": 3853
+    },
+    {
+      "epoch": 1.2868113522537563,
+      "grad_norm": 0.49385142889482286,
+      "learning_rate": 3.398494490362902e-06,
+      "loss": 0.2474,
+      "step": 3854
+    },
+    {
+      "epoch": 1.2871452420701168,
+      "grad_norm": 0.47956880425852233,
+      "learning_rate": 3.3957345321377143e-06,
+      "loss": 0.2577,
+      "step": 3855
+    },
+    {
+      "epoch": 1.2874791318864776,
+      "grad_norm": 0.4865752082642456,
+      "learning_rate": 3.392975118712787e-06,
+      "loss": 0.2627,
+      "step": 3856
+    },
+    {
+      "epoch": 1.287813021702838,
+      "grad_norm": 0.48176129449033905,
+      "learning_rate": 3.3902162510252e-06,
+      "loss": 0.2413,
+      "step": 3857
+    },
+    {
+      "epoch": 1.2881469115191986,
+      "grad_norm": 0.47976468588711807,
+      "learning_rate": 3.3874579300118515e-06,
+      "loss": 0.2486,
+      "step": 3858
+    },
+    {
+      "epoch": 1.2884808013355593,
+      "grad_norm": 0.44440422955559816,
+      "learning_rate": 3.3847001566094518e-06,
+      "loss": 0.2424,
+      "step": 3859
+    },
+    {
+      "epoch": 1.2888146911519198,
+      "grad_norm": 0.4583537505776499,
+      "learning_rate": 3.381942931754528e-06,
+      "loss": 0.2392,
+      "step": 3860
+    },
+    {
+      "epoch": 1.2891485809682806,
+      "grad_norm": 0.47923620352019136,
+      "learning_rate": 3.379186256383418e-06,
+      "loss": 0.2615,
+      "step": 3861
+    },
+    {
+      "epoch": 1.289482470784641,
+      "grad_norm": 0.49947675659557456,
+      "learning_rate": 3.376430131432275e-06,
+      "loss": 0.2504,
+      "step": 3862
+    },
+    {
+      "epoch": 1.2898163606010016,
+      "grad_norm": 0.4882981459139035,
+      "learning_rate": 3.3736745578370654e-06,
+      "loss": 0.2478,
+      "step": 3863
+    },
+    {
+      "epoch": 1.2901502504173623,
+      "grad_norm": 0.5051258459021996,
+      "learning_rate": 3.3709195365335646e-06,
+      "loss": 0.2593,
+      "step": 3864
+    },
+    {
+      "epoch": 1.2904841402337228,
+      "grad_norm": 0.49039158421652934,
+      "learning_rate": 3.3681650684573695e-06,
+      "loss": 0.2611,
+      "step": 3865
+    },
+    {
+      "epoch": 1.2908180300500836,
+      "grad_norm": 0.47584700183499345,
+      "learning_rate": 3.365411154543878e-06,
+      "loss": 0.2533,
+      "step": 3866
+    },
+    {
+      "epoch": 1.291151919866444,
+      "grad_norm": 0.4645555662188108,
+      "learning_rate": 3.362657795728309e-06,
+      "loss": 0.2495,
+      "step": 3867
+    },
+    {
+      "epoch": 1.2914858096828046,
+      "grad_norm": 0.4663966174057095,
+      "learning_rate": 3.3599049929456857e-06,
+      "loss": 0.2531,
+      "step": 3868
+    },
+    {
+      "epoch": 1.2918196994991653,
+      "grad_norm": 0.4823892231701605,
+      "learning_rate": 3.3571527471308476e-06,
+      "loss": 0.2523,
+      "step": 3869
+    },
+    {
+      "epoch": 1.2921535893155258,
+      "grad_norm": 0.5136292029807318,
+      "learning_rate": 3.3544010592184428e-06,
+      "loss": 0.262,
+      "step": 3870
+    },
+    {
+      "epoch": 1.2924874791318866,
+      "grad_norm": 0.4613429109010515,
+      "learning_rate": 3.3516499301429282e-06,
+      "loss": 0.246,
+      "step": 3871
+    },
+    {
+      "epoch": 1.292821368948247,
+      "grad_norm": 0.4899723159896895,
+      "learning_rate": 3.3488993608385745e-06,
+      "loss": 0.2561,
+      "step": 3872
+    },
+    {
+      "epoch": 1.2931552587646076,
+      "grad_norm": 0.4518898569138849,
+      "learning_rate": 3.3461493522394616e-06,
+      "loss": 0.2345,
+      "step": 3873
+    },
+    {
+      "epoch": 1.2934891485809683,
+      "grad_norm": 0.47703396344528676,
+      "learning_rate": 3.343399905279479e-06,
+      "loss": 0.2493,
+      "step": 3874
+    },
+    {
+      "epoch": 1.2938230383973288,
+      "grad_norm": 0.47137610789907136,
+      "learning_rate": 3.3406510208923228e-06,
+      "loss": 0.2544,
+      "step": 3875
+    },
+    {
+      "epoch": 1.2941569282136896,
+      "grad_norm": 0.4637687879292807,
+      "learning_rate": 3.3379027000115005e-06,
+      "loss": 0.2499,
+      "step": 3876
+    },
+    {
+      "epoch": 1.29449081803005,
+      "grad_norm": 0.4465289261760884,
+      "learning_rate": 3.335154943570328e-06,
+      "loss": 0.2511,
+      "step": 3877
+    },
+    {
+      "epoch": 1.2948247078464106,
+      "grad_norm": 0.46635202255391245,
+      "learning_rate": 3.332407752501928e-06,
+      "loss": 0.2428,
+      "step": 3878
+    },
+    {
+      "epoch": 1.2951585976627713,
+      "grad_norm": 0.5240288904197932,
+      "learning_rate": 3.329661127739234e-06,
+      "loss": 0.2543,
+      "step": 3879
+    },
+    {
+      "epoch": 1.2954924874791318,
+      "grad_norm": 0.4690282642874593,
+      "learning_rate": 3.3269150702149834e-06,
+      "loss": 0.2576,
+      "step": 3880
+    },
+    {
+      "epoch": 1.2958263772954925,
+      "grad_norm": 0.46629640872351674,
+      "learning_rate": 3.3241695808617235e-06,
+      "loss": 0.2548,
+      "step": 3881
+    },
+    {
+      "epoch": 1.296160267111853,
+      "grad_norm": 0.47131079628186284,
+      "learning_rate": 3.3214246606118116e-06,
+      "loss": 0.2472,
+      "step": 3882
+    },
+    {
+      "epoch": 1.2964941569282136,
+      "grad_norm": 0.49635524612498916,
+      "learning_rate": 3.3186803103974054e-06,
+      "loss": 0.257,
+      "step": 3883
+    },
+    {
+      "epoch": 1.2968280467445743,
+      "grad_norm": 0.5138736753301437,
+      "learning_rate": 3.3159365311504732e-06,
+      "loss": 0.2554,
+      "step": 3884
+    },
+    {
+      "epoch": 1.2971619365609348,
+      "grad_norm": 0.4801637361971029,
+      "learning_rate": 3.3131933238027877e-06,
+      "loss": 0.2439,
+      "step": 3885
+    },
+    {
+      "epoch": 1.2974958263772955,
+      "grad_norm": 0.4668282174865204,
+      "learning_rate": 3.310450689285928e-06,
+      "loss": 0.2508,
+      "step": 3886
+    },
+    {
+      "epoch": 1.297829716193656,
+      "grad_norm": 0.4852728039566259,
+      "learning_rate": 3.307708628531278e-06,
+      "loss": 0.2613,
+      "step": 3887
+    },
+    {
+      "epoch": 1.2981636060100166,
+      "grad_norm": 0.5139727945699935,
+      "learning_rate": 3.3049671424700293e-06,
+      "loss": 0.262,
+      "step": 3888
+    },
+    {
+      "epoch": 1.2984974958263773,
+      "grad_norm": 0.48046872202285024,
+      "learning_rate": 3.302226232033172e-06,
+      "loss": 0.2603,
+      "step": 3889
+    },
+    {
+      "epoch": 1.298831385642738,
+      "grad_norm": 0.47851098282797494,
+      "learning_rate": 3.2994858981515103e-06,
+      "loss": 0.2441,
+      "step": 3890
+    },
+    {
+      "epoch": 1.2991652754590985,
+      "grad_norm": 0.48104135340526405,
+      "learning_rate": 3.296746141755646e-06,
+      "loss": 0.2597,
+      "step": 3891
+    },
+    {
+      "epoch": 1.299499165275459,
+      "grad_norm": 0.45983572396166966,
+      "learning_rate": 3.2940069637759843e-06,
+      "loss": 0.2573,
+      "step": 3892
+    },
+    {
+      "epoch": 1.2998330550918196,
+      "grad_norm": 0.49455826296277255,
+      "learning_rate": 3.291268365142738e-06,
+      "loss": 0.259,
+      "step": 3893
+    },
+    {
+      "epoch": 1.3001669449081803,
+      "grad_norm": 0.46986807744358855,
+      "learning_rate": 3.2885303467859196e-06,
+      "loss": 0.2534,
+      "step": 3894
+    },
+    {
+      "epoch": 1.300500834724541,
+      "grad_norm": 0.47249222555656584,
+      "learning_rate": 3.285792909635347e-06,
+      "loss": 0.2416,
+      "step": 3895
+    },
+    {
+      "epoch": 1.3008347245409015,
+      "grad_norm": 0.4794160781367273,
+      "learning_rate": 3.2830560546206392e-06,
+      "loss": 0.2594,
+      "step": 3896
+    },
+    {
+      "epoch": 1.301168614357262,
+      "grad_norm": 0.46437915278487474,
+      "learning_rate": 3.280319782671218e-06,
+      "loss": 0.2526,
+      "step": 3897
+    },
+    {
+      "epoch": 1.3015025041736228,
+      "grad_norm": 0.49980292815477134,
+      "learning_rate": 3.277584094716305e-06,
+      "loss": 0.2621,
+      "step": 3898
+    },
+    {
+      "epoch": 1.3018363939899833,
+      "grad_norm": 0.46842907712324094,
+      "learning_rate": 3.27484899168493e-06,
+      "loss": 0.245,
+      "step": 3899
+    },
+    {
+      "epoch": 1.302170283806344,
+      "grad_norm": 0.48137964638933844,
+      "learning_rate": 3.2721144745059173e-06,
+      "loss": 0.2557,
+      "step": 3900
+    },
+    {
+      "epoch": 1.3025041736227045,
+      "grad_norm": 0.4768899417249122,
+      "learning_rate": 3.2693805441078956e-06,
+      "loss": 0.2604,
+      "step": 3901
+    },
+    {
+      "epoch": 1.302838063439065,
+      "grad_norm": 0.48126411643577116,
+      "learning_rate": 3.2666472014192942e-06,
+      "loss": 0.2515,
+      "step": 3902
+    },
+    {
+      "epoch": 1.3031719532554258,
+      "grad_norm": 0.4937814089254811,
+      "learning_rate": 3.2639144473683393e-06,
+      "loss": 0.2587,
+      "step": 3903
+    },
+    {
+      "epoch": 1.3035058430717863,
+      "grad_norm": 0.47392922714657565,
+      "learning_rate": 3.261182282883063e-06,
+      "loss": 0.2592,
+      "step": 3904
+    },
+    {
+      "epoch": 1.303839732888147,
+      "grad_norm": 0.46368404915817063,
+      "learning_rate": 3.258450708891292e-06,
+      "loss": 0.2466,
+      "step": 3905
+    },
+    {
+      "epoch": 1.3041736227045075,
+      "grad_norm": 0.46655518739916557,
+      "learning_rate": 3.255719726320655e-06,
+      "loss": 0.2569,
+      "step": 3906
+    },
+    {
+      "epoch": 1.304507512520868,
+      "grad_norm": 0.4760856078127407,
+      "learning_rate": 3.252989336098583e-06,
+      "loss": 0.2459,
+      "step": 3907
+    },
+    {
+      "epoch": 1.3048414023372288,
+      "grad_norm": 0.4771467584195235,
+      "learning_rate": 3.2502595391522985e-06,
+      "loss": 0.2461,
+      "step": 3908
+    },
+    {
+      "epoch": 1.3051752921535893,
+      "grad_norm": 0.4875653551816249,
+      "learning_rate": 3.2475303364088283e-06,
+      "loss": 0.2396,
+      "step": 3909
+    },
+    {
+      "epoch": 1.30550918196995,
+      "grad_norm": 0.47386098436952834,
+      "learning_rate": 3.244801728794994e-06,
+      "loss": 0.2558,
+      "step": 3910
+    },
+    {
+      "epoch": 1.3058430717863105,
+      "grad_norm": 0.4821849304587213,
+      "learning_rate": 3.2420737172374184e-06,
+      "loss": 0.2533,
+      "step": 3911
+    },
+    {
+      "epoch": 1.306176961602671,
+      "grad_norm": 0.485242521446545,
+      "learning_rate": 3.2393463026625183e-06,
+      "loss": 0.2606,
+      "step": 3912
+    },
+    {
+      "epoch": 1.3065108514190318,
+      "grad_norm": 0.49255355060071904,
+      "learning_rate": 3.236619485996512e-06,
+      "loss": 0.2488,
+      "step": 3913
+    },
+    {
+      "epoch": 1.3068447412353923,
+      "grad_norm": 0.502667853873332,
+      "learning_rate": 3.23389326816541e-06,
+      "loss": 0.2592,
+      "step": 3914
+    },
+    {
+      "epoch": 1.307178631051753,
+      "grad_norm": 0.4610323714202998,
+      "learning_rate": 3.2311676500950217e-06,
+      "loss": 0.2535,
+      "step": 3915
+    },
+    {
+      "epoch": 1.3075125208681135,
+      "grad_norm": 0.4801327652749297,
+      "learning_rate": 3.2284426327109553e-06,
+      "loss": 0.2598,
+      "step": 3916
+    },
+    {
+      "epoch": 1.307846410684474,
+      "grad_norm": 0.4998882266497486,
+      "learning_rate": 3.2257182169386102e-06,
+      "loss": 0.2461,
+      "step": 3917
+    },
+    {
+      "epoch": 1.3081803005008348,
+      "grad_norm": 0.4910478911981061,
+      "learning_rate": 3.222994403703187e-06,
+      "loss": 0.2543,
+      "step": 3918
+    },
+    {
+      "epoch": 1.3085141903171953,
+      "grad_norm": 0.4904675378653864,
+      "learning_rate": 3.220271193929675e-06,
+      "loss": 0.2559,
+      "step": 3919
+    },
+    {
+      "epoch": 1.308848080133556,
+      "grad_norm": 0.4595487567901312,
+      "learning_rate": 3.217548588542864e-06,
+      "loss": 0.251,
+      "step": 3920
+    },
+    {
+      "epoch": 1.3091819699499165,
+      "grad_norm": 0.4784661647484766,
+      "learning_rate": 3.2148265884673365e-06,
+      "loss": 0.2535,
+      "step": 3921
+    },
+    {
+      "epoch": 1.309515859766277,
+      "grad_norm": 0.46660802628643644,
+      "learning_rate": 3.2121051946274694e-06,
+      "loss": 0.2493,
+      "step": 3922
+    },
+    {
+      "epoch": 1.3098497495826378,
+      "grad_norm": 0.4594984871467606,
+      "learning_rate": 3.2093844079474335e-06,
+      "loss": 0.2508,
+      "step": 3923
+    },
+    {
+      "epoch": 1.3101836393989983,
+      "grad_norm": 0.4796784494285418,
+      "learning_rate": 3.206664229351194e-06,
+      "loss": 0.2588,
+      "step": 3924
+    },
+    {
+      "epoch": 1.310517529215359,
+      "grad_norm": 0.4855049178243787,
+      "learning_rate": 3.2039446597625124e-06,
+      "loss": 0.2552,
+      "step": 3925
+    },
+    {
+      "epoch": 1.3108514190317195,
+      "grad_norm": 0.47951444677778315,
+      "learning_rate": 3.2012257001049364e-06,
+      "loss": 0.2628,
+      "step": 3926
+    },
+    {
+      "epoch": 1.31118530884808,
+      "grad_norm": 0.4705805627791561,
+      "learning_rate": 3.198507351301813e-06,
+      "loss": 0.2498,
+      "step": 3927
+    },
+    {
+      "epoch": 1.3115191986644408,
+      "grad_norm": 0.4669914907248905,
+      "learning_rate": 3.1957896142762778e-06,
+      "loss": 0.2489,
+      "step": 3928
+    },
+    {
+      "epoch": 1.3118530884808013,
+      "grad_norm": 0.46917051826923256,
+      "learning_rate": 3.1930724899512633e-06,
+      "loss": 0.2552,
+      "step": 3929
+    },
+    {
+      "epoch": 1.312186978297162,
+      "grad_norm": 0.4719538021063392,
+      "learning_rate": 3.1903559792494864e-06,
+      "loss": 0.2534,
+      "step": 3930
+    },
+    {
+      "epoch": 1.3125208681135225,
+      "grad_norm": 0.4568707976314068,
+      "learning_rate": 3.1876400830934637e-06,
+      "loss": 0.2529,
+      "step": 3931
+    },
+    {
+      "epoch": 1.312854757929883,
+      "grad_norm": 0.4565844845413321,
+      "learning_rate": 3.184924802405497e-06,
+      "loss": 0.2451,
+      "step": 3932
+    },
+    {
+      "epoch": 1.3131886477462438,
+      "grad_norm": 0.4797971309105394,
+      "learning_rate": 3.1822101381076832e-06,
+      "loss": 0.249,
+      "step": 3933
+    },
+    {
+      "epoch": 1.3135225375626043,
+      "grad_norm": 0.4959103145717375,
+      "learning_rate": 3.1794960911219085e-06,
+      "loss": 0.2587,
+      "step": 3934
+    },
+    {
+      "epoch": 1.313856427378965,
+      "grad_norm": 0.4737664587549118,
+      "learning_rate": 3.1767826623698483e-06,
+      "loss": 0.258,
+      "step": 3935
+    },
+    {
+      "epoch": 1.3141903171953255,
+      "grad_norm": 0.4776797441649633,
+      "learning_rate": 3.17406985277297e-06,
+      "loss": 0.2529,
+      "step": 3936
+    },
+    {
+      "epoch": 1.314524207011686,
+      "grad_norm": 0.48138484829306716,
+      "learning_rate": 3.1713576632525273e-06,
+      "loss": 0.2598,
+      "step": 3937
+    },
+    {
+      "epoch": 1.3148580968280468,
+      "grad_norm": 0.4682326945528885,
+      "learning_rate": 3.1686460947295695e-06,
+      "loss": 0.2548,
+      "step": 3938
+    },
+    {
+      "epoch": 1.3151919866444073,
+      "grad_norm": 0.44200579577976407,
+      "learning_rate": 3.165935148124927e-06,
+      "loss": 0.2423,
+      "step": 3939
+    },
+    {
+      "epoch": 1.315525876460768,
+      "grad_norm": 0.4774437823263737,
+      "learning_rate": 3.163224824359225e-06,
+      "loss": 0.2539,
+      "step": 3940
+    },
+    {
+      "epoch": 1.3158597662771285,
+      "grad_norm": 0.4675385973800583,
+      "learning_rate": 3.160515124352878e-06,
+      "loss": 0.2543,
+      "step": 3941
+    },
+    {
+      "epoch": 1.316193656093489,
+      "grad_norm": 0.5093059401398684,
+      "learning_rate": 3.157806049026083e-06,
+      "loss": 0.2691,
+      "step": 3942
+    },
+    {
+      "epoch": 1.3165275459098498,
+      "grad_norm": 0.4715731636257213,
+      "learning_rate": 3.1550975992988298e-06,
+      "loss": 0.2544,
+      "step": 3943
+    },
+    {
+      "epoch": 1.3168614357262103,
+      "grad_norm": 0.4739167114296232,
+      "learning_rate": 3.1523897760908916e-06,
+      "loss": 0.2569,
+      "step": 3944
+    },
+    {
+      "epoch": 1.317195325542571,
+      "grad_norm": 0.4685267633691241,
+      "learning_rate": 3.1496825803218344e-06,
+      "loss": 0.2528,
+      "step": 3945
+    },
+    {
+      "epoch": 1.3175292153589315,
+      "grad_norm": 0.4632978628971496,
+      "learning_rate": 3.1469760129110046e-06,
+      "loss": 0.2639,
+      "step": 3946
+    },
+    {
+      "epoch": 1.317863105175292,
+      "grad_norm": 0.4778178649119029,
+      "learning_rate": 3.1442700747775413e-06,
+      "loss": 0.2505,
+      "step": 3947
+    },
+    {
+      "epoch": 1.3181969949916528,
+      "grad_norm": 0.4854097392173701,
+      "learning_rate": 3.1415647668403643e-06,
+      "loss": 0.2574,
+      "step": 3948
+    },
+    {
+      "epoch": 1.3185308848080133,
+      "grad_norm": 0.4593029363665182,
+      "learning_rate": 3.1388600900181832e-06,
+      "loss": 0.2537,
+      "step": 3949
+    },
+    {
+      "epoch": 1.318864774624374,
+      "grad_norm": 0.48233205411379054,
+      "learning_rate": 3.136156045229494e-06,
+      "loss": 0.2517,
+      "step": 3950
+    },
+    {
+      "epoch": 1.3191986644407345,
+      "grad_norm": 0.4757180359008764,
+      "learning_rate": 3.1334526333925754e-06,
+      "loss": 0.2544,
+      "step": 3951
+    },
+    {
+      "epoch": 1.319532554257095,
+      "grad_norm": 0.4907371139933925,
+      "learning_rate": 3.130749855425491e-06,
+      "loss": 0.2535,
+      "step": 3952
+    },
+    {
+      "epoch": 1.3198664440734558,
+      "grad_norm": 0.47261220236724294,
+      "learning_rate": 3.1280477122460917e-06,
+      "loss": 0.2568,
+      "step": 3953
+    },
+    {
+      "epoch": 1.3202003338898163,
+      "grad_norm": 0.48061589070745925,
+      "learning_rate": 3.1253462047720107e-06,
+      "loss": 0.2495,
+      "step": 3954
+    },
+    {
+      "epoch": 1.320534223706177,
+      "grad_norm": 0.46714772015185846,
+      "learning_rate": 3.122645333920665e-06,
+      "loss": 0.2467,
+      "step": 3955
+    },
+    {
+      "epoch": 1.3208681135225375,
+      "grad_norm": 0.4736238125604397,
+      "learning_rate": 3.1199451006092586e-06,
+      "loss": 0.2547,
+      "step": 3956
+    },
+    {
+      "epoch": 1.321202003338898,
+      "grad_norm": 0.4832082380481675,
+      "learning_rate": 3.117245505754773e-06,
+      "loss": 0.2511,
+      "step": 3957
+    },
+    {
+      "epoch": 1.3215358931552588,
+      "grad_norm": 0.5197690190477975,
+      "learning_rate": 3.11454655027398e-06,
+      "loss": 0.263,
+      "step": 3958
+    },
+    {
+      "epoch": 1.3218697829716195,
+      "grad_norm": 0.4716802621304285,
+      "learning_rate": 3.1118482350834316e-06,
+      "loss": 0.2564,
+      "step": 3959
+    },
+    {
+      "epoch": 1.32220367278798,
+      "grad_norm": 0.4933352391964484,
+      "learning_rate": 3.109150561099459e-06,
+      "loss": 0.2538,
+      "step": 3960
+    },
+    {
+      "epoch": 1.3225375626043405,
+      "grad_norm": 0.46878836255896217,
+      "learning_rate": 3.1064535292381803e-06,
+      "loss": 0.2456,
+      "step": 3961
+    },
+    {
+      "epoch": 1.322871452420701,
+      "grad_norm": 0.4780878760909182,
+      "learning_rate": 3.103757140415492e-06,
+      "loss": 0.2538,
+      "step": 3962
+    },
+    {
+      "epoch": 1.3232053422370618,
+      "grad_norm": 0.4890673528717047,
+      "learning_rate": 3.1010613955470754e-06,
+      "loss": 0.2477,
+      "step": 3963
+    },
+    {
+      "epoch": 1.3235392320534225,
+      "grad_norm": 0.47448458886404443,
+      "learning_rate": 3.0983662955483895e-06,
+      "loss": 0.2618,
+      "step": 3964
+    },
+    {
+      "epoch": 1.323873121869783,
+      "grad_norm": 0.48698491310085124,
+      "learning_rate": 3.0956718413346785e-06,
+      "loss": 0.2543,
+      "step": 3965
+    },
+    {
+      "epoch": 1.3242070116861435,
+      "grad_norm": 0.486401856333418,
+      "learning_rate": 3.0929780338209627e-06,
+      "loss": 0.2579,
+      "step": 3966
+    },
+    {
+      "epoch": 1.3245409015025043,
+      "grad_norm": 0.45882381609164324,
+      "learning_rate": 3.090284873922047e-06,
+      "loss": 0.2371,
+      "step": 3967
+    },
+    {
+      "epoch": 1.3248747913188648,
+      "grad_norm": 0.4929972973427026,
+      "learning_rate": 3.087592362552515e-06,
+      "loss": 0.263,
+      "step": 3968
+    },
+    {
+      "epoch": 1.3252086811352255,
+      "grad_norm": 0.4578192426915619,
+      "learning_rate": 3.084900500626728e-06,
+      "loss": 0.2466,
+      "step": 3969
+    },
+    {
+      "epoch": 1.325542570951586,
+      "grad_norm": 0.4785814871248793,
+      "learning_rate": 3.0822092890588307e-06,
+      "loss": 0.2484,
+      "step": 3970
+    },
+    {
+      "epoch": 1.3258764607679465,
+      "grad_norm": 0.4574137819995031,
+      "learning_rate": 3.0795187287627414e-06,
+      "loss": 0.2481,
+      "step": 3971
+    },
+    {
+      "epoch": 1.3262103505843073,
+      "grad_norm": 0.4660399729132793,
+      "learning_rate": 3.0768288206521636e-06,
+      "loss": 0.253,
+      "step": 3972
+    },
+    {
+      "epoch": 1.3265442404006678,
+      "grad_norm": 0.47362805876572567,
+      "learning_rate": 3.0741395656405735e-06,
+      "loss": 0.2574,
+      "step": 3973
+    },
+    {
+      "epoch": 1.3268781302170285,
+      "grad_norm": 0.46935979266237815,
+      "learning_rate": 3.0714509646412295e-06,
+      "loss": 0.2685,
+      "step": 3974
+    },
+    {
+      "epoch": 1.327212020033389,
+      "grad_norm": 0.4672521529061639,
+      "learning_rate": 3.0687630185671674e-06,
+      "loss": 0.2524,
+      "step": 3975
+    },
+    {
+      "epoch": 1.3275459098497495,
+      "grad_norm": 0.46670157475237667,
+      "learning_rate": 3.0660757283311986e-06,
+      "loss": 0.2589,
+      "step": 3976
+    },
+    {
+      "epoch": 1.3278797996661102,
+      "grad_norm": 0.4852085005420855,
+      "learning_rate": 3.0633890948459142e-06,
+      "loss": 0.2524,
+      "step": 3977
+    },
+    {
+      "epoch": 1.3282136894824708,
+      "grad_norm": 0.4682097698605569,
+      "learning_rate": 3.060703119023679e-06,
+      "loss": 0.2467,
+      "step": 3978
+    },
+    {
+      "epoch": 1.3285475792988315,
+      "grad_norm": 0.4642129933681047,
+      "learning_rate": 3.0580178017766393e-06,
+      "loss": 0.2415,
+      "step": 3979
+    },
+    {
+      "epoch": 1.328881469115192,
+      "grad_norm": 0.47231555724956636,
+      "learning_rate": 3.055333144016712e-06,
+      "loss": 0.2524,
+      "step": 3980
+    },
+    {
+      "epoch": 1.3292153589315525,
+      "grad_norm": 0.47705966577566133,
+      "learning_rate": 3.052649146655595e-06,
+      "loss": 0.2556,
+      "step": 3981
+    },
+    {
+      "epoch": 1.3295492487479132,
+      "grad_norm": 0.4783851480201294,
+      "learning_rate": 3.049965810604759e-06,
+      "loss": 0.2385,
+      "step": 3982
+    },
+    {
+      "epoch": 1.3298831385642738,
+      "grad_norm": 0.4914671257327012,
+      "learning_rate": 3.0472831367754496e-06,
+      "loss": 0.2492,
+      "step": 3983
+    },
+    {
+      "epoch": 1.3302170283806345,
+      "grad_norm": 0.4943343916815974,
+      "learning_rate": 3.0446011260786933e-06,
+      "loss": 0.2656,
+      "step": 3984
+    },
+    {
+      "epoch": 1.330550918196995,
+      "grad_norm": 0.45742520630173206,
+      "learning_rate": 3.041919779425283e-06,
+      "loss": 0.2429,
+      "step": 3985
+    },
+    {
+      "epoch": 1.3308848080133555,
+      "grad_norm": 0.48696370285710855,
+      "learning_rate": 3.0392390977257928e-06,
+      "loss": 0.2505,
+      "step": 3986
+    },
+    {
+      "epoch": 1.3312186978297162,
+      "grad_norm": 0.4693127519899506,
+      "learning_rate": 3.036559081890566e-06,
+      "loss": 0.2452,
+      "step": 3987
+    },
+    {
+      "epoch": 1.3315525876460768,
+      "grad_norm": 0.47092220564021725,
+      "learning_rate": 3.0338797328297243e-06,
+      "loss": 0.2461,
+      "step": 3988
+    },
+    {
+      "epoch": 1.3318864774624375,
+      "grad_norm": 0.4660028342158266,
+      "learning_rate": 3.0312010514531592e-06,
+      "loss": 0.2524,
+      "step": 3989
+    },
+    {
+      "epoch": 1.332220367278798,
+      "grad_norm": 0.46103931778243223,
+      "learning_rate": 3.028523038670538e-06,
+      "loss": 0.2395,
+      "step": 3990
+    },
+    {
+      "epoch": 1.3325542570951585,
+      "grad_norm": 0.4634871730334565,
+      "learning_rate": 3.0258456953912985e-06,
+      "loss": 0.2485,
+      "step": 3991
+    },
+    {
+      "epoch": 1.3328881469115192,
+      "grad_norm": 0.47988957350153166,
+      "learning_rate": 3.0231690225246537e-06,
+      "loss": 0.2471,
+      "step": 3992
+    },
+    {
+      "epoch": 1.3332220367278798,
+      "grad_norm": 0.47084274279360644,
+      "learning_rate": 3.020493020979589e-06,
+      "loss": 0.2461,
+      "step": 3993
+    },
+    {
+      "epoch": 1.3335559265442405,
+      "grad_norm": 0.49058874579190487,
+      "learning_rate": 3.0178176916648595e-06,
+      "loss": 0.2558,
+      "step": 3994
+    },
+    {
+      "epoch": 1.333889816360601,
+      "grad_norm": 0.4795795710066906,
+      "learning_rate": 3.015143035488993e-06,
+      "loss": 0.2574,
+      "step": 3995
+    },
+    {
+      "epoch": 1.3342237061769615,
+      "grad_norm": 0.46169000831509166,
+      "learning_rate": 3.0124690533602895e-06,
+      "loss": 0.2489,
+      "step": 3996
+    },
+    {
+      "epoch": 1.3345575959933222,
+      "grad_norm": 0.48867645747739813,
+      "learning_rate": 3.00979574618682e-06,
+      "loss": 0.2619,
+      "step": 3997
+    },
+    {
+      "epoch": 1.3348914858096828,
+      "grad_norm": 0.48574501941490944,
+      "learning_rate": 3.0071231148764235e-06,
+      "loss": 0.2527,
+      "step": 3998
+    },
+    {
+      "epoch": 1.3352253756260435,
+      "grad_norm": 0.47714079026496925,
+      "learning_rate": 3.0044511603367143e-06,
+      "loss": 0.2555,
+      "step": 3999
+    },
+    {
+      "epoch": 1.335559265442404,
+      "grad_norm": 0.45927404125258825,
+      "learning_rate": 3.001779883475071e-06,
+      "loss": 0.2565,
+      "step": 4000
+    },
+    {
+      "epoch": 1.3358931552587645,
+      "grad_norm": 0.46688760113325184,
+      "learning_rate": 2.999109285198649e-06,
+      "loss": 0.2522,
+      "step": 4001
+    },
+    {
+      "epoch": 1.3362270450751252,
+      "grad_norm": 0.49359750353168724,
+      "learning_rate": 2.99643936641437e-06,
+      "loss": 0.2531,
+      "step": 4002
+    },
+    {
+      "epoch": 1.3365609348914858,
+      "grad_norm": 0.4689253686357293,
+      "learning_rate": 2.9937701280289216e-06,
+      "loss": 0.2589,
+      "step": 4003
+    },
+    {
+      "epoch": 1.3368948247078465,
+      "grad_norm": 0.4805216381931073,
+      "learning_rate": 2.9911015709487654e-06,
+      "loss": 0.2528,
+      "step": 4004
+    },
+    {
+      "epoch": 1.337228714524207,
+      "grad_norm": 0.4804854100996269,
+      "learning_rate": 2.9884336960801284e-06,
+      "loss": 0.263,
+      "step": 4005
+    },
+    {
+      "epoch": 1.3375626043405675,
+      "grad_norm": 0.4534157732128834,
+      "learning_rate": 2.985766504329009e-06,
+      "loss": 0.2507,
+      "step": 4006
+    },
+    {
+      "epoch": 1.3378964941569282,
+      "grad_norm": 0.4705102147362671,
+      "learning_rate": 2.9830999966011688e-06,
+      "loss": 0.2568,
+      "step": 4007
+    },
+    {
+      "epoch": 1.3382303839732888,
+      "grad_norm": 0.46347925149167324,
+      "learning_rate": 2.9804341738021407e-06,
+      "loss": 0.2522,
+      "step": 4008
+    },
+    {
+      "epoch": 1.3385642737896495,
+      "grad_norm": 0.47806663180872244,
+      "learning_rate": 2.977769036837228e-06,
+      "loss": 0.2571,
+      "step": 4009
+    },
+    {
+      "epoch": 1.33889816360601,
+      "grad_norm": 0.46756504231185253,
+      "learning_rate": 2.9751045866114926e-06,
+      "loss": 0.2473,
+      "step": 4010
+    },
+    {
+      "epoch": 1.3392320534223705,
+      "grad_norm": 0.47500662289416967,
+      "learning_rate": 2.972440824029772e-06,
+      "loss": 0.2559,
+      "step": 4011
+    },
+    {
+      "epoch": 1.3395659432387312,
+      "grad_norm": 0.4630052328022044,
+      "learning_rate": 2.969777749996663e-06,
+      "loss": 0.2413,
+      "step": 4012
+    },
+    {
+      "epoch": 1.3398998330550917,
+      "grad_norm": 0.4679763992924414,
+      "learning_rate": 2.967115365416534e-06,
+      "loss": 0.2549,
+      "step": 4013
+    },
+    {
+      "epoch": 1.3402337228714525,
+      "grad_norm": 0.4852453576069965,
+      "learning_rate": 2.964453671193515e-06,
+      "loss": 0.2534,
+      "step": 4014
+    },
+    {
+      "epoch": 1.340567612687813,
+      "grad_norm": 0.47034542506863264,
+      "learning_rate": 2.9617926682315062e-06,
+      "loss": 0.2541,
+      "step": 4015
+    },
+    {
+      "epoch": 1.3409015025041735,
+      "grad_norm": 0.48770083166992795,
+      "learning_rate": 2.959132357434168e-06,
+      "loss": 0.2546,
+      "step": 4016
+    },
+    {
+      "epoch": 1.3412353923205342,
+      "grad_norm": 0.4778110036828003,
+      "learning_rate": 2.9564727397049274e-06,
+      "loss": 0.2609,
+      "step": 4017
+    },
+    {
+      "epoch": 1.3415692821368947,
+      "grad_norm": 0.4697322611776745,
+      "learning_rate": 2.9538138159469814e-06,
+      "loss": 0.2583,
+      "step": 4018
+    },
+    {
+      "epoch": 1.3419031719532555,
+      "grad_norm": 0.4830052278784066,
+      "learning_rate": 2.9511555870632824e-06,
+      "loss": 0.2498,
+      "step": 4019
+    },
+    {
+      "epoch": 1.342237061769616,
+      "grad_norm": 0.4707851257712897,
+      "learning_rate": 2.9484980539565546e-06,
+      "loss": 0.2545,
+      "step": 4020
+    },
+    {
+      "epoch": 1.3425709515859765,
+      "grad_norm": 0.48966496550052246,
+      "learning_rate": 2.945841217529279e-06,
+      "loss": 0.2602,
+      "step": 4021
+    },
+    {
+      "epoch": 1.3429048414023372,
+      "grad_norm": 0.4746824072243776,
+      "learning_rate": 2.9431850786837063e-06,
+      "loss": 0.2522,
+      "step": 4022
+    },
+    {
+      "epoch": 1.343238731218698,
+      "grad_norm": 0.4820889872719598,
+      "learning_rate": 2.940529638321845e-06,
+      "loss": 0.2568,
+      "step": 4023
+    },
+    {
+      "epoch": 1.3435726210350585,
+      "grad_norm": 0.48360632348503085,
+      "learning_rate": 2.937874897345472e-06,
+      "loss": 0.2551,
+      "step": 4024
+    },
+    {
+      "epoch": 1.343906510851419,
+      "grad_norm": 0.47789887879815846,
+      "learning_rate": 2.935220856656119e-06,
+      "loss": 0.2606,
+      "step": 4025
+    },
+    {
+      "epoch": 1.3442404006677795,
+      "grad_norm": 0.4706885280708992,
+      "learning_rate": 2.9325675171550895e-06,
+      "loss": 0.2598,
+      "step": 4026
+    },
+    {
+      "epoch": 1.3445742904841402,
+      "grad_norm": 0.48435169603019695,
+      "learning_rate": 2.9299148797434425e-06,
+      "loss": 0.2509,
+      "step": 4027
+    },
+    {
+      "epoch": 1.344908180300501,
+      "grad_norm": 0.47693220447400825,
+      "learning_rate": 2.927262945321998e-06,
+      "loss": 0.2589,
+      "step": 4028
+    },
+    {
+      "epoch": 1.3452420701168615,
+      "grad_norm": 0.48474038464616376,
+      "learning_rate": 2.9246117147913412e-06,
+      "loss": 0.251,
+      "step": 4029
+    },
+    {
+      "epoch": 1.345575959933222,
+      "grad_norm": 0.46773390848402957,
+      "learning_rate": 2.921961189051815e-06,
+      "loss": 0.2501,
+      "step": 4030
+    },
+    {
+      "epoch": 1.3459098497495827,
+      "grad_norm": 0.47341056072642995,
+      "learning_rate": 2.9193113690035256e-06,
+      "loss": 0.2459,
+      "step": 4031
+    },
+    {
+      "epoch": 1.3462437395659432,
+      "grad_norm": 0.46318934822192076,
+      "learning_rate": 2.9166622555463366e-06,
+      "loss": 0.2499,
+      "step": 4032
+    },
+    {
+      "epoch": 1.346577629382304,
+      "grad_norm": 0.47809226171828995,
+      "learning_rate": 2.9140138495798742e-06,
+      "loss": 0.2491,
+      "step": 4033
+    },
+    {
+      "epoch": 1.3469115191986645,
+      "grad_norm": 0.4891853249287526,
+      "learning_rate": 2.9113661520035218e-06,
+      "loss": 0.2652,
+      "step": 4034
+    },
+    {
+      "epoch": 1.347245409015025,
+      "grad_norm": 0.4996308404417567,
+      "learning_rate": 2.9087191637164268e-06,
+      "loss": 0.258,
+      "step": 4035
+    },
+    {
+      "epoch": 1.3475792988313857,
+      "grad_norm": 0.5027519965901721,
+      "learning_rate": 2.9060728856174898e-06,
+      "loss": 0.2656,
+      "step": 4036
+    },
+    {
+      "epoch": 1.3479131886477462,
+      "grad_norm": 0.48307477462689424,
+      "learning_rate": 2.903427318605376e-06,
+      "loss": 0.2567,
+      "step": 4037
+    },
+    {
+      "epoch": 1.348247078464107,
+      "grad_norm": 0.4572994458370103,
+      "learning_rate": 2.9007824635785047e-06,
+      "loss": 0.2413,
+      "step": 4038
+    },
+    {
+      "epoch": 1.3485809682804675,
+      "grad_norm": 0.47690123806503987,
+      "learning_rate": 2.898138321435055e-06,
+      "loss": 0.2528,
+      "step": 4039
+    },
+    {
+      "epoch": 1.348914858096828,
+      "grad_norm": 0.47882483274471577,
+      "learning_rate": 2.895494893072962e-06,
+      "loss": 0.2491,
+      "step": 4040
+    },
+    {
+      "epoch": 1.3492487479131887,
+      "grad_norm": 0.46564922911776263,
+      "learning_rate": 2.8928521793899246e-06,
+      "loss": 0.2538,
+      "step": 4041
+    },
+    {
+      "epoch": 1.3495826377295492,
+      "grad_norm": 0.48771955431302016,
+      "learning_rate": 2.89021018128339e-06,
+      "loss": 0.2602,
+      "step": 4042
+    },
+    {
+      "epoch": 1.34991652754591,
+      "grad_norm": 0.46618953693083093,
+      "learning_rate": 2.8875688996505713e-06,
+      "loss": 0.2473,
+      "step": 4043
+    },
+    {
+      "epoch": 1.3502504173622705,
+      "grad_norm": 0.4702243628077505,
+      "learning_rate": 2.884928335388433e-06,
+      "loss": 0.2443,
+      "step": 4044
+    },
+    {
+      "epoch": 1.350584307178631,
+      "grad_norm": 0.5077666894833464,
+      "learning_rate": 2.8822884893936933e-06,
+      "loss": 0.2507,
+      "step": 4045
+    },
+    {
+      "epoch": 1.3509181969949917,
+      "grad_norm": 0.4771034113618043,
+      "learning_rate": 2.8796493625628357e-06,
+      "loss": 0.2444,
+      "step": 4046
+    },
+    {
+      "epoch": 1.3512520868113522,
+      "grad_norm": 0.49412677394559745,
+      "learning_rate": 2.8770109557920916e-06,
+      "loss": 0.2623,
+      "step": 4047
+    },
+    {
+      "epoch": 1.351585976627713,
+      "grad_norm": 0.48682252415468524,
+      "learning_rate": 2.87437326997745e-06,
+      "loss": 0.2541,
+      "step": 4048
+    },
+    {
+      "epoch": 1.3519198664440735,
+      "grad_norm": 0.4701421149172766,
+      "learning_rate": 2.8717363060146543e-06,
+      "loss": 0.2564,
+      "step": 4049
+    },
+    {
+      "epoch": 1.352253756260434,
+      "grad_norm": 0.4547742948399459,
+      "learning_rate": 2.869100064799206e-06,
+      "loss": 0.2456,
+      "step": 4050
+    },
+    {
+      "epoch": 1.3525876460767947,
+      "grad_norm": 0.47687676895874714,
+      "learning_rate": 2.8664645472263575e-06,
+      "loss": 0.2419,
+      "step": 4051
+    },
+    {
+      "epoch": 1.3529215358931552,
+      "grad_norm": 0.4847379200499921,
+      "learning_rate": 2.8638297541911187e-06,
+      "loss": 0.2535,
+      "step": 4052
+    },
+    {
+      "epoch": 1.353255425709516,
+      "grad_norm": 0.4654109411460228,
+      "learning_rate": 2.8611956865882513e-06,
+      "loss": 0.252,
+      "step": 4053
+    },
+    {
+      "epoch": 1.3535893155258765,
+      "grad_norm": 0.45977495481094943,
+      "learning_rate": 2.858562345312268e-06,
+      "loss": 0.2424,
+      "step": 4054
+    },
+    {
+      "epoch": 1.353923205342237,
+      "grad_norm": 0.4860274092337715,
+      "learning_rate": 2.8559297312574417e-06,
+      "loss": 0.2575,
+      "step": 4055
+    },
+    {
+      "epoch": 1.3542570951585977,
+      "grad_norm": 0.4735092349699444,
+      "learning_rate": 2.853297845317794e-06,
+      "loss": 0.2522,
+      "step": 4056
+    },
+    {
+      "epoch": 1.3545909849749582,
+      "grad_norm": 0.4754836975273236,
+      "learning_rate": 2.8506666883870982e-06,
+      "loss": 0.2553,
+      "step": 4057
+    },
+    {
+      "epoch": 1.354924874791319,
+      "grad_norm": 0.479848015407775,
+      "learning_rate": 2.8480362613588817e-06,
+      "loss": 0.2522,
+      "step": 4058
+    },
+    {
+      "epoch": 1.3552587646076795,
+      "grad_norm": 0.460167679021279,
+      "learning_rate": 2.8454065651264212e-06,
+      "loss": 0.2536,
+      "step": 4059
+    },
+    {
+      "epoch": 1.35559265442404,
+      "grad_norm": 0.4909917289793325,
+      "learning_rate": 2.8427776005827546e-06,
+      "loss": 0.2567,
+      "step": 4060
+    },
+    {
+      "epoch": 1.3559265442404007,
+      "grad_norm": 0.47123876688163896,
+      "learning_rate": 2.8401493686206618e-06,
+      "loss": 0.2694,
+      "step": 4061
+    },
+    {
+      "epoch": 1.3562604340567612,
+      "grad_norm": 0.5005460040408946,
+      "learning_rate": 2.8375218701326757e-06,
+      "loss": 0.2406,
+      "step": 4062
+    },
+    {
+      "epoch": 1.356594323873122,
+      "grad_norm": 0.4679676094650976,
+      "learning_rate": 2.8348951060110796e-06,
+      "loss": 0.2543,
+      "step": 4063
+    },
+    {
+      "epoch": 1.3569282136894825,
+      "grad_norm": 0.46323832579671975,
+      "learning_rate": 2.8322690771479135e-06,
+      "loss": 0.2568,
+      "step": 4064
+    },
+    {
+      "epoch": 1.357262103505843,
+      "grad_norm": 0.4691816531199435,
+      "learning_rate": 2.829643784434961e-06,
+      "loss": 0.2416,
+      "step": 4065
+    },
+    {
+      "epoch": 1.3575959933222037,
+      "grad_norm": 0.4893424854541171,
+      "learning_rate": 2.827019228763758e-06,
+      "loss": 0.2451,
+      "step": 4066
+    },
+    {
+      "epoch": 1.3579298831385642,
+      "grad_norm": 0.49434426469320436,
+      "learning_rate": 2.824395411025589e-06,
+      "loss": 0.2501,
+      "step": 4067
+    },
+    {
+      "epoch": 1.358263772954925,
+      "grad_norm": 0.4673639440441596,
+      "learning_rate": 2.8217723321114855e-06,
+      "loss": 0.2472,
+      "step": 4068
+    },
+    {
+      "epoch": 1.3585976627712855,
+      "grad_norm": 0.4607987803138074,
+      "learning_rate": 2.81914999291224e-06,
+      "loss": 0.2494,
+      "step": 4069
+    },
+    {
+      "epoch": 1.358931552587646,
+      "grad_norm": 0.4839543012416478,
+      "learning_rate": 2.816528394318382e-06,
+      "loss": 0.2543,
+      "step": 4070
+    },
+    {
+      "epoch": 1.3592654424040067,
+      "grad_norm": 0.481162567601408,
+      "learning_rate": 2.813907537220191e-06,
+      "loss": 0.258,
+      "step": 4071
+    },
+    {
+      "epoch": 1.3595993322203672,
+      "grad_norm": 0.46060375323360786,
+      "learning_rate": 2.811287422507698e-06,
+      "loss": 0.2536,
+      "step": 4072
+    },
+    {
+      "epoch": 1.359933222036728,
+      "grad_norm": 0.477606111669602,
+      "learning_rate": 2.808668051070678e-06,
+      "loss": 0.247,
+      "step": 4073
+    },
+    {
+      "epoch": 1.3602671118530885,
+      "grad_norm": 0.4698486544512348,
+      "learning_rate": 2.806049423798659e-06,
+      "loss": 0.248,
+      "step": 4074
+    },
+    {
+      "epoch": 1.360601001669449,
+      "grad_norm": 0.46604079800787956,
+      "learning_rate": 2.8034315415809126e-06,
+      "loss": 0.254,
+      "step": 4075
+    },
+    {
+      "epoch": 1.3609348914858097,
+      "grad_norm": 0.47001376883453505,
+      "learning_rate": 2.8008144053064557e-06,
+      "loss": 0.258,
+      "step": 4076
+    },
+    {
+      "epoch": 1.3612687813021702,
+      "grad_norm": 0.45192672390080213,
+      "learning_rate": 2.798198015864058e-06,
+      "loss": 0.25,
+      "step": 4077
+    },
+    {
+      "epoch": 1.361602671118531,
+      "grad_norm": 0.4892426748832604,
+      "learning_rate": 2.7955823741422285e-06,
+      "loss": 0.2533,
+      "step": 4078
+    },
+    {
+      "epoch": 1.3619365609348915,
+      "grad_norm": 0.46673512559824276,
+      "learning_rate": 2.7929674810292284e-06,
+      "loss": 0.2505,
+      "step": 4079
+    },
+    {
+      "epoch": 1.362270450751252,
+      "grad_norm": 0.4681526983561642,
+      "learning_rate": 2.7903533374130607e-06,
+      "loss": 0.2492,
+      "step": 4080
+    },
+    {
+      "epoch": 1.3626043405676127,
+      "grad_norm": 0.453933427898063,
+      "learning_rate": 2.787739944181475e-06,
+      "loss": 0.252,
+      "step": 4081
+    },
+    {
+      "epoch": 1.3629382303839732,
+      "grad_norm": 0.47515126316574874,
+      "learning_rate": 2.7851273022219645e-06,
+      "loss": 0.2609,
+      "step": 4082
+    },
+    {
+      "epoch": 1.363272120200334,
+      "grad_norm": 0.4946207123258846,
+      "learning_rate": 2.7825154124217716e-06,
+      "loss": 0.2561,
+      "step": 4083
+    },
+    {
+      "epoch": 1.3636060100166945,
+      "grad_norm": 0.484106601739074,
+      "learning_rate": 2.77990427566788e-06,
+      "loss": 0.2465,
+      "step": 4084
+    },
+    {
+      "epoch": 1.363939899833055,
+      "grad_norm": 0.5125555165797138,
+      "learning_rate": 2.777293892847016e-06,
+      "loss": 0.2621,
+      "step": 4085
+    },
+    {
+      "epoch": 1.3642737896494157,
+      "grad_norm": 0.48051294524873656,
+      "learning_rate": 2.7746842648456552e-06,
+      "loss": 0.2432,
+      "step": 4086
+    },
+    {
+      "epoch": 1.3646076794657762,
+      "grad_norm": 0.4780092952194629,
+      "learning_rate": 2.7720753925500105e-06,
+      "loss": 0.2519,
+      "step": 4087
+    },
+    {
+      "epoch": 1.364941569282137,
+      "grad_norm": 0.46318581650509905,
+      "learning_rate": 2.769467276846045e-06,
+      "loss": 0.2577,
+      "step": 4088
+    },
+    {
+      "epoch": 1.3652754590984975,
+      "grad_norm": 0.45172215573410635,
+      "learning_rate": 2.76685991861946e-06,
+      "loss": 0.2489,
+      "step": 4089
+    },
+    {
+      "epoch": 1.365609348914858,
+      "grad_norm": 0.46461078343226575,
+      "learning_rate": 2.764253318755701e-06,
+      "loss": 0.2551,
+      "step": 4090
+    },
+    {
+      "epoch": 1.3659432387312187,
+      "grad_norm": 0.469050138710678,
+      "learning_rate": 2.7616474781399527e-06,
+      "loss": 0.252,
+      "step": 4091
+    },
+    {
+      "epoch": 1.3662771285475794,
+      "grad_norm": 0.46486857832964584,
+      "learning_rate": 2.759042397657149e-06,
+      "loss": 0.2424,
+      "step": 4092
+    },
+    {
+      "epoch": 1.36661101836394,
+      "grad_norm": 0.47833393506727173,
+      "learning_rate": 2.75643807819196e-06,
+      "loss": 0.2571,
+      "step": 4093
+    },
+    {
+      "epoch": 1.3669449081803005,
+      "grad_norm": 0.460821703733965,
+      "learning_rate": 2.7538345206288005e-06,
+      "loss": 0.2391,
+      "step": 4094
+    },
+    {
+      "epoch": 1.367278797996661,
+      "grad_norm": 0.455351091022819,
+      "learning_rate": 2.7512317258518244e-06,
+      "loss": 0.2485,
+      "step": 4095
+    },
+    {
+      "epoch": 1.3676126878130217,
+      "grad_norm": 0.47068609494766417,
+      "learning_rate": 2.7486296947449254e-06,
+      "loss": 0.255,
+      "step": 4096
+    },
+    {
+      "epoch": 1.3679465776293824,
+      "grad_norm": 0.4775608085959764,
+      "learning_rate": 2.746028428191744e-06,
+      "loss": 0.2588,
+      "step": 4097
+    },
+    {
+      "epoch": 1.368280467445743,
+      "grad_norm": 0.470281340222106,
+      "learning_rate": 2.7434279270756536e-06,
+      "loss": 0.2519,
+      "step": 4098
+    },
+    {
+      "epoch": 1.3686143572621035,
+      "grad_norm": 0.4694047247726791,
+      "learning_rate": 2.740828192279773e-06,
+      "loss": 0.2417,
+      "step": 4099
+    },
+    {
+      "epoch": 1.3689482470784642,
+      "grad_norm": 0.4858699449674361,
+      "learning_rate": 2.7382292246869548e-06,
+      "loss": 0.2548,
+      "step": 4100
+    },
+    {
+      "epoch": 1.3692821368948247,
+      "grad_norm": 0.4557096379997235,
+      "learning_rate": 2.7356310251798e-06,
+      "loss": 0.2397,
+      "step": 4101
+    },
+    {
+      "epoch": 1.3696160267111854,
+      "grad_norm": 0.4612561612934116,
+      "learning_rate": 2.7330335946406394e-06,
+      "loss": 0.2456,
+      "step": 4102
+    },
+    {
+      "epoch": 1.369949916527546,
+      "grad_norm": 0.4742343885236233,
+      "learning_rate": 2.7304369339515515e-06,
+      "loss": 0.255,
+      "step": 4103
+    },
+    {
+      "epoch": 1.3702838063439065,
+      "grad_norm": 0.4849383607417421,
+      "learning_rate": 2.7278410439943454e-06,
+      "loss": 0.2523,
+      "step": 4104
+    },
+    {
+      "epoch": 1.3706176961602672,
+      "grad_norm": 0.4854557577332508,
+      "learning_rate": 2.7252459256505715e-06,
+      "loss": 0.2438,
+      "step": 4105
+    },
+    {
+      "epoch": 1.3709515859766277,
+      "grad_norm": 0.48302444578409665,
+      "learning_rate": 2.7226515798015214e-06,
+      "loss": 0.2513,
+      "step": 4106
+    },
+    {
+      "epoch": 1.3712854757929884,
+      "grad_norm": 0.4567651615864983,
+      "learning_rate": 2.72005800732822e-06,
+      "loss": 0.2385,
+      "step": 4107
+    },
+    {
+      "epoch": 1.371619365609349,
+      "grad_norm": 0.4667758695960134,
+      "learning_rate": 2.7174652091114316e-06,
+      "loss": 0.2444,
+      "step": 4108
+    },
+    {
+      "epoch": 1.3719532554257095,
+      "grad_norm": 0.4912129983075742,
+      "learning_rate": 2.7148731860316544e-06,
+      "loss": 0.2519,
+      "step": 4109
+    },
+    {
+      "epoch": 1.3722871452420702,
+      "grad_norm": 0.4821646203132138,
+      "learning_rate": 2.712281938969129e-06,
+      "loss": 0.2565,
+      "step": 4110
+    },
+    {
+      "epoch": 1.3726210350584307,
+      "grad_norm": 0.5087310736106384,
+      "learning_rate": 2.70969146880383e-06,
+      "loss": 0.2532,
+      "step": 4111
+    },
+    {
+      "epoch": 1.3729549248747914,
+      "grad_norm": 0.4807123022391582,
+      "learning_rate": 2.707101776415467e-06,
+      "loss": 0.2498,
+      "step": 4112
+    },
+    {
+      "epoch": 1.373288814691152,
+      "grad_norm": 0.489752919870198,
+      "learning_rate": 2.704512862683485e-06,
+      "loss": 0.266,
+      "step": 4113
+    },
+    {
+      "epoch": 1.3736227045075124,
+      "grad_norm": 0.4578094770470805,
+      "learning_rate": 2.701924728487065e-06,
+      "loss": 0.2426,
+      "step": 4114
+    },
+    {
+      "epoch": 1.3739565943238732,
+      "grad_norm": 0.45078674349106523,
+      "learning_rate": 2.6993373747051265e-06,
+      "loss": 0.2453,
+      "step": 4115
+    },
+    {
+      "epoch": 1.3742904841402337,
+      "grad_norm": 0.48559815221440067,
+      "learning_rate": 2.696750802216321e-06,
+      "loss": 0.2468,
+      "step": 4116
+    },
+    {
+      "epoch": 1.3746243739565944,
+      "grad_norm": 0.4769171112816047,
+      "learning_rate": 2.6941650118990338e-06,
+      "loss": 0.2541,
+      "step": 4117
+    },
+    {
+      "epoch": 1.374958263772955,
+      "grad_norm": 0.5101041383185315,
+      "learning_rate": 2.6915800046313852e-06,
+      "loss": 0.2711,
+      "step": 4118
+    },
+    {
+      "epoch": 1.3752921535893154,
+      "grad_norm": 0.4648692102695922,
+      "learning_rate": 2.688995781291231e-06,
+      "loss": 0.2455,
+      "step": 4119
+    },
+    {
+      "epoch": 1.3756260434056762,
+      "grad_norm": 0.47931160022779,
+      "learning_rate": 2.6864123427561627e-06,
+      "loss": 0.2502,
+      "step": 4120
+    },
+    {
+      "epoch": 1.3759599332220367,
+      "grad_norm": 0.47619340625087464,
+      "learning_rate": 2.683829689903501e-06,
+      "loss": 0.2504,
+      "step": 4121
+    },
+    {
+      "epoch": 1.3762938230383974,
+      "grad_norm": 0.4737131696223613,
+      "learning_rate": 2.6812478236103014e-06,
+      "loss": 0.2549,
+      "step": 4122
+    },
+    {
+      "epoch": 1.376627712854758,
+      "grad_norm": 0.4804252361470575,
+      "learning_rate": 2.6786667447533498e-06,
+      "loss": 0.2519,
+      "step": 4123
+    },
+    {
+      "epoch": 1.3769616026711184,
+      "grad_norm": 0.4630859903986632,
+      "learning_rate": 2.6760864542091717e-06,
+      "loss": 0.2438,
+      "step": 4124
+    },
+    {
+      "epoch": 1.3772954924874792,
+      "grad_norm": 0.49163754421647227,
+      "learning_rate": 2.673506952854018e-06,
+      "loss": 0.2527,
+      "step": 4125
+    },
+    {
+      "epoch": 1.3776293823038397,
+      "grad_norm": 0.4847517453419395,
+      "learning_rate": 2.670928241563875e-06,
+      "loss": 0.2479,
+      "step": 4126
+    },
+    {
+      "epoch": 1.3779632721202004,
+      "grad_norm": 0.4667048932233535,
+      "learning_rate": 2.6683503212144563e-06,
+      "loss": 0.2583,
+      "step": 4127
+    },
+    {
+      "epoch": 1.378297161936561,
+      "grad_norm": 0.4743340727210411,
+      "learning_rate": 2.6657731926812132e-06,
+      "loss": 0.2546,
+      "step": 4128
+    },
+    {
+      "epoch": 1.3786310517529214,
+      "grad_norm": 0.47759141672375505,
+      "learning_rate": 2.6631968568393273e-06,
+      "loss": 0.2429,
+      "step": 4129
+    },
+    {
+      "epoch": 1.3789649415692822,
+      "grad_norm": 0.45831687794208104,
+      "learning_rate": 2.660621314563706e-06,
+      "loss": 0.239,
+      "step": 4130
+    },
+    {
+      "epoch": 1.3792988313856427,
+      "grad_norm": 0.4676189058880371,
+      "learning_rate": 2.6580465667289912e-06,
+      "loss": 0.2479,
+      "step": 4131
+    },
+    {
+      "epoch": 1.3796327212020034,
+      "grad_norm": 0.4693264421112457,
+      "learning_rate": 2.6554726142095524e-06,
+      "loss": 0.2514,
+      "step": 4132
+    },
+    {
+      "epoch": 1.379966611018364,
+      "grad_norm": 0.4716046086588294,
+      "learning_rate": 2.652899457879492e-06,
+      "loss": 0.24,
+      "step": 4133
+    },
+    {
+      "epoch": 1.3803005008347244,
+      "grad_norm": 0.4532857766141229,
+      "learning_rate": 2.6503270986126417e-06,
+      "loss": 0.2493,
+      "step": 4134
+    },
+    {
+      "epoch": 1.3806343906510852,
+      "grad_norm": 0.4759604335442302,
+      "learning_rate": 2.6477555372825573e-06,
+      "loss": 0.2491,
+      "step": 4135
+    },
+    {
+      "epoch": 1.3809682804674457,
+      "grad_norm": 0.4482204475650956,
+      "learning_rate": 2.645184774762533e-06,
+      "loss": 0.2424,
+      "step": 4136
+    },
+    {
+      "epoch": 1.3813021702838064,
+      "grad_norm": 0.4436758827143962,
+      "learning_rate": 2.6426148119255817e-06,
+      "loss": 0.2492,
+      "step": 4137
+    },
+    {
+      "epoch": 1.381636060100167,
+      "grad_norm": 0.500808541997587,
+      "learning_rate": 2.6400456496444536e-06,
+      "loss": 0.2651,
+      "step": 4138
+    },
+    {
+      "epoch": 1.3819699499165274,
+      "grad_norm": 0.5013840437206303,
+      "learning_rate": 2.6374772887916213e-06,
+      "loss": 0.2532,
+      "step": 4139
+    },
+    {
+      "epoch": 1.3823038397328882,
+      "grad_norm": 0.46326611133331724,
+      "learning_rate": 2.6349097302392867e-06,
+      "loss": 0.2529,
+      "step": 4140
+    },
+    {
+      "epoch": 1.3826377295492487,
+      "grad_norm": 0.49531196656662463,
+      "learning_rate": 2.632342974859377e-06,
+      "loss": 0.2627,
+      "step": 4141
+    },
+    {
+      "epoch": 1.3829716193656094,
+      "grad_norm": 0.47528255049490725,
+      "learning_rate": 2.629777023523553e-06,
+      "loss": 0.2571,
+      "step": 4142
+    },
+    {
+      "epoch": 1.38330550918197,
+      "grad_norm": 0.49360657475631636,
+      "learning_rate": 2.627211877103198e-06,
+      "loss": 0.2516,
+      "step": 4143
+    },
+    {
+      "epoch": 1.3836393989983304,
+      "grad_norm": 0.4705928734689301,
+      "learning_rate": 2.6246475364694183e-06,
+      "loss": 0.2459,
+      "step": 4144
+    },
+    {
+      "epoch": 1.3839732888146912,
+      "grad_norm": 0.4612282350310652,
+      "learning_rate": 2.6220840024930562e-06,
+      "loss": 0.2424,
+      "step": 4145
+    },
+    {
+      "epoch": 1.3843071786310517,
+      "grad_norm": 0.4845140287836429,
+      "learning_rate": 2.61952127604467e-06,
+      "loss": 0.2462,
+      "step": 4146
+    },
+    {
+      "epoch": 1.3846410684474124,
+      "grad_norm": 0.47300908954043847,
+      "learning_rate": 2.616959357994553e-06,
+      "loss": 0.2566,
+      "step": 4147
+    },
+    {
+      "epoch": 1.384974958263773,
+      "grad_norm": 0.47841345095586557,
+      "learning_rate": 2.614398249212716e-06,
+      "loss": 0.2528,
+      "step": 4148
+    },
+    {
+      "epoch": 1.3853088480801334,
+      "grad_norm": 0.48661830094210634,
+      "learning_rate": 2.6118379505689003e-06,
+      "loss": 0.2459,
+      "step": 4149
+    },
+    {
+      "epoch": 1.3856427378964942,
+      "grad_norm": 0.4893987325690486,
+      "learning_rate": 2.609278462932566e-06,
+      "loss": 0.2524,
+      "step": 4150
+    },
+    {
+      "epoch": 1.3859766277128547,
+      "grad_norm": 0.4729386589112648,
+      "learning_rate": 2.606719787172907e-06,
+      "loss": 0.2483,
+      "step": 4151
+    },
+    {
+      "epoch": 1.3863105175292154,
+      "grad_norm": 0.49673242295804054,
+      "learning_rate": 2.604161924158832e-06,
+      "loss": 0.254,
+      "step": 4152
+    },
+    {
+      "epoch": 1.386644407345576,
+      "grad_norm": 0.5014006423183521,
+      "learning_rate": 2.601604874758982e-06,
+      "loss": 0.2609,
+      "step": 4153
+    },
+    {
+      "epoch": 1.3869782971619364,
+      "grad_norm": 0.47390829260131395,
+      "learning_rate": 2.599048639841717e-06,
+      "loss": 0.248,
+      "step": 4154
+    },
+    {
+      "epoch": 1.3873121869782972,
+      "grad_norm": 0.5165318638465842,
+      "learning_rate": 2.596493220275117e-06,
+      "loss": 0.2571,
+      "step": 4155
+    },
+    {
+      "epoch": 1.387646076794658,
+      "grad_norm": 0.4679056618163404,
+      "learning_rate": 2.5939386169269947e-06,
+      "loss": 0.241,
+      "step": 4156
+    },
+    {
+      "epoch": 1.3879799666110184,
+      "grad_norm": 0.4737644443863347,
+      "learning_rate": 2.591384830664878e-06,
+      "loss": 0.254,
+      "step": 4157
+    },
+    {
+      "epoch": 1.388313856427379,
+      "grad_norm": 0.4893779417061027,
+      "learning_rate": 2.5888318623560192e-06,
+      "loss": 0.2532,
+      "step": 4158
+    },
+    {
+      "epoch": 1.3886477462437394,
+      "grad_norm": 0.47837653900540406,
+      "learning_rate": 2.586279712867391e-06,
+      "loss": 0.2481,
+      "step": 4159
+    },
+    {
+      "epoch": 1.3889816360601002,
+      "grad_norm": 0.4461674192390258,
+      "learning_rate": 2.5837283830656955e-06,
+      "loss": 0.2457,
+      "step": 4160
+    },
+    {
+      "epoch": 1.389315525876461,
+      "grad_norm": 0.4782958998588953,
+      "learning_rate": 2.5811778738173453e-06,
+      "loss": 0.2521,
+      "step": 4161
+    },
+    {
+      "epoch": 1.3896494156928214,
+      "grad_norm": 0.48442744679854993,
+      "learning_rate": 2.5786281859884856e-06,
+      "loss": 0.2523,
+      "step": 4162
+    },
+    {
+      "epoch": 1.389983305509182,
+      "grad_norm": 0.4828457000276984,
+      "learning_rate": 2.5760793204449737e-06,
+      "loss": 0.2608,
+      "step": 4163
+    },
+    {
+      "epoch": 1.3903171953255427,
+      "grad_norm": 0.47858854374566,
+      "learning_rate": 2.57353127805239e-06,
+      "loss": 0.2569,
+      "step": 4164
+    },
+    {
+      "epoch": 1.3906510851419032,
+      "grad_norm": 0.4569778255919143,
+      "learning_rate": 2.570984059676041e-06,
+      "loss": 0.2477,
+      "step": 4165
+    },
+    {
+      "epoch": 1.390984974958264,
+      "grad_norm": 0.47697526740412294,
+      "learning_rate": 2.5684376661809465e-06,
+      "loss": 0.2455,
+      "step": 4166
+    },
+    {
+      "epoch": 1.3913188647746244,
+      "grad_norm": 0.48039933384145295,
+      "learning_rate": 2.565892098431848e-06,
+      "loss": 0.2482,
+      "step": 4167
+    },
+    {
+      "epoch": 1.391652754590985,
+      "grad_norm": 0.4847859412650437,
+      "learning_rate": 2.5633473572932066e-06,
+      "loss": 0.2605,
+      "step": 4168
+    },
+    {
+      "epoch": 1.3919866444073457,
+      "grad_norm": 0.4842183844562091,
+      "learning_rate": 2.5608034436292036e-06,
+      "loss": 0.2442,
+      "step": 4169
+    },
+    {
+      "epoch": 1.3923205342237062,
+      "grad_norm": 0.45560863260773643,
+      "learning_rate": 2.5582603583037415e-06,
+      "loss": 0.2395,
+      "step": 4170
+    },
+    {
+      "epoch": 1.392654424040067,
+      "grad_norm": 0.46539648246489557,
+      "learning_rate": 2.555718102180437e-06,
+      "loss": 0.2479,
+      "step": 4171
+    },
+    {
+      "epoch": 1.3929883138564274,
+      "grad_norm": 0.4971452154576658,
+      "learning_rate": 2.553176676122627e-06,
+      "loss": 0.2633,
+      "step": 4172
+    },
+    {
+      "epoch": 1.393322203672788,
+      "grad_norm": 0.48661216249253303,
+      "learning_rate": 2.5506360809933657e-06,
+      "loss": 0.2607,
+      "step": 4173
+    },
+    {
+      "epoch": 1.3936560934891487,
+      "grad_norm": 0.46380072026548114,
+      "learning_rate": 2.5480963176554285e-06,
+      "loss": 0.2473,
+      "step": 4174
+    },
+    {
+      "epoch": 1.3939899833055092,
+      "grad_norm": 0.4818854997483026,
+      "learning_rate": 2.5455573869713046e-06,
+      "loss": 0.251,
+      "step": 4175
+    },
+    {
+      "epoch": 1.39432387312187,
+      "grad_norm": 0.49041099595749343,
+      "learning_rate": 2.543019289803201e-06,
+      "loss": 0.2574,
+      "step": 4176
+    },
+    {
+      "epoch": 1.3946577629382304,
+      "grad_norm": 0.4956258983995786,
+      "learning_rate": 2.5404820270130424e-06,
+      "loss": 0.2697,
+      "step": 4177
+    },
+    {
+      "epoch": 1.394991652754591,
+      "grad_norm": 0.45421697785293397,
+      "learning_rate": 2.537945599462471e-06,
+      "loss": 0.2479,
+      "step": 4178
+    },
+    {
+      "epoch": 1.3953255425709516,
+      "grad_norm": 0.46318585858703976,
+      "learning_rate": 2.535410008012845e-06,
+      "loss": 0.2531,
+      "step": 4179
+    },
+    {
+      "epoch": 1.3956594323873122,
+      "grad_norm": 0.48429530837391366,
+      "learning_rate": 2.532875253525238e-06,
+      "loss": 0.2467,
+      "step": 4180
+    },
+    {
+      "epoch": 1.395993322203673,
+      "grad_norm": 0.4520754385482453,
+      "learning_rate": 2.530341336860439e-06,
+      "loss": 0.2346,
+      "step": 4181
+    },
+    {
+      "epoch": 1.3963272120200334,
+      "grad_norm": 0.44588225422501027,
+      "learning_rate": 2.527808258878951e-06,
+      "loss": 0.248,
+      "step": 4182
+    },
+    {
+      "epoch": 1.396661101836394,
+      "grad_norm": 0.4637362434810362,
+      "learning_rate": 2.525276020440998e-06,
+      "loss": 0.2534,
+      "step": 4183
+    },
+    {
+      "epoch": 1.3969949916527546,
+      "grad_norm": 0.46212832520684705,
+      "learning_rate": 2.5227446224065134e-06,
+      "loss": 0.257,
+      "step": 4184
+    },
+    {
+      "epoch": 1.3973288814691152,
+      "grad_norm": 0.48819957833564415,
+      "learning_rate": 2.520214065635147e-06,
+      "loss": 0.2578,
+      "step": 4185
+    },
+    {
+      "epoch": 1.397662771285476,
+      "grad_norm": 0.47862797395104073,
+      "learning_rate": 2.5176843509862604e-06,
+      "loss": 0.2555,
+      "step": 4186
+    },
+    {
+      "epoch": 1.3979966611018364,
+      "grad_norm": 0.4757640508386217,
+      "learning_rate": 2.5151554793189325e-06,
+      "loss": 0.2499,
+      "step": 4187
+    },
+    {
+      "epoch": 1.398330550918197,
+      "grad_norm": 0.48696051493952563,
+      "learning_rate": 2.5126274514919597e-06,
+      "loss": 0.2556,
+      "step": 4188
+    },
+    {
+      "epoch": 1.3986644407345576,
+      "grad_norm": 0.4689279272302198,
+      "learning_rate": 2.510100268363843e-06,
+      "loss": 0.2507,
+      "step": 4189
+    },
+    {
+      "epoch": 1.3989983305509182,
+      "grad_norm": 0.4812932589349834,
+      "learning_rate": 2.5075739307928017e-06,
+      "loss": 0.2498,
+      "step": 4190
+    },
+    {
+      "epoch": 1.399332220367279,
+      "grad_norm": 0.4839701591939875,
+      "learning_rate": 2.505048439636765e-06,
+      "loss": 0.2522,
+      "step": 4191
+    },
+    {
+      "epoch": 1.3996661101836394,
+      "grad_norm": 0.48162872204825036,
+      "learning_rate": 2.5025237957533797e-06,
+      "loss": 0.2491,
+      "step": 4192
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.46754151107164343,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.2592,
+      "step": 4193
+    },
+    {
+      "epoch": 1.4003338898163606,
+      "grad_norm": 0.500406786083437,
+      "learning_rate": 2.4974770532336967e-06,
+      "loss": 0.2545,
+      "step": 4194
+    },
+    {
+      "epoch": 1.4006677796327212,
+      "grad_norm": 0.46790266985889406,
+      "learning_rate": 2.4949549563112434e-06,
+      "loss": 0.2467,
+      "step": 4195
+    },
+    {
+      "epoch": 1.4010016694490819,
+      "grad_norm": 0.4796019420029514,
+      "learning_rate": 2.4924337100891355e-06,
+      "loss": 0.2493,
+      "step": 4196
+    },
+    {
+      "epoch": 1.4013355592654424,
+      "grad_norm": 0.46408656846590335,
+      "learning_rate": 2.489913315423576e-06,
+      "loss": 0.2466,
+      "step": 4197
+    },
+    {
+      "epoch": 1.401669449081803,
+      "grad_norm": 0.48068506758537666,
+      "learning_rate": 2.487393773170477e-06,
+      "loss": 0.2511,
+      "step": 4198
+    },
+    {
+      "epoch": 1.4020033388981636,
+      "grad_norm": 0.4694911254882121,
+      "learning_rate": 2.484875084185462e-06,
+      "loss": 0.2547,
+      "step": 4199
+    },
+    {
+      "epoch": 1.4023372287145242,
+      "grad_norm": 0.46756894643869545,
+      "learning_rate": 2.4823572493238613e-06,
+      "loss": 0.2497,
+      "step": 4200
+    },
+    {
+      "epoch": 1.4026711185308849,
+      "grad_norm": 0.49635065628155034,
+      "learning_rate": 2.479840269440724e-06,
+      "loss": 0.2511,
+      "step": 4201
+    },
+    {
+      "epoch": 1.4030050083472454,
+      "grad_norm": 0.46499564124294895,
+      "learning_rate": 2.477324145390801e-06,
+      "loss": 0.2408,
+      "step": 4202
+    },
+    {
+      "epoch": 1.403338898163606,
+      "grad_norm": 0.49826985493957127,
+      "learning_rate": 2.4748088780285533e-06,
+      "loss": 0.2629,
+      "step": 4203
+    },
+    {
+      "epoch": 1.4036727879799666,
+      "grad_norm": 0.4636815361462638,
+      "learning_rate": 2.472294468208156e-06,
+      "loss": 0.2515,
+      "step": 4204
+    },
+    {
+      "epoch": 1.4040066777963272,
+      "grad_norm": 0.475031979539754,
+      "learning_rate": 2.4697809167834863e-06,
+      "loss": 0.2547,
+      "step": 4205
+    },
+    {
+      "epoch": 1.4043405676126879,
+      "grad_norm": 0.4752527513694065,
+      "learning_rate": 2.4672682246081375e-06,
+      "loss": 0.2504,
+      "step": 4206
+    },
+    {
+      "epoch": 1.4046744574290484,
+      "grad_norm": 0.48585439977993167,
+      "learning_rate": 2.464756392535404e-06,
+      "loss": 0.2489,
+      "step": 4207
+    },
+    {
+      "epoch": 1.405008347245409,
+      "grad_norm": 0.48086514083419923,
+      "learning_rate": 2.462245421418292e-06,
+      "loss": 0.2514,
+      "step": 4208
+    },
+    {
+      "epoch": 1.4053422370617696,
+      "grad_norm": 0.4588645682544067,
+      "learning_rate": 2.459735312109511e-06,
+      "loss": 0.2552,
+      "step": 4209
+    },
+    {
+      "epoch": 1.4056761268781301,
+      "grad_norm": 0.469057183894216,
+      "learning_rate": 2.457226065461486e-06,
+      "loss": 0.2508,
+      "step": 4210
+    },
+    {
+      "epoch": 1.4060100166944909,
+      "grad_norm": 0.4679786508792219,
+      "learning_rate": 2.4547176823263423e-06,
+      "loss": 0.2398,
+      "step": 4211
+    },
+    {
+      "epoch": 1.4063439065108514,
+      "grad_norm": 0.4940754188910155,
+      "learning_rate": 2.452210163555912e-06,
+      "loss": 0.2543,
+      "step": 4212
+    },
+    {
+      "epoch": 1.406677796327212,
+      "grad_norm": 0.47823510193938523,
+      "learning_rate": 2.4497035100017384e-06,
+      "loss": 0.2514,
+      "step": 4213
+    },
+    {
+      "epoch": 1.4070116861435726,
+      "grad_norm": 0.49457854253980194,
+      "learning_rate": 2.4471977225150655e-06,
+      "loss": 0.2586,
+      "step": 4214
+    },
+    {
+      "epoch": 1.4073455759599331,
+      "grad_norm": 0.48147601614381397,
+      "learning_rate": 2.444692801946849e-06,
+      "loss": 0.264,
+      "step": 4215
+    },
+    {
+      "epoch": 1.4076794657762939,
+      "grad_norm": 0.48162914170886434,
+      "learning_rate": 2.442188749147745e-06,
+      "loss": 0.2384,
+      "step": 4216
+    },
+    {
+      "epoch": 1.4080133555926544,
+      "grad_norm": 0.4623200256496862,
+      "learning_rate": 2.4396855649681166e-06,
+      "loss": 0.2544,
+      "step": 4217
+    },
+    {
+      "epoch": 1.408347245409015,
+      "grad_norm": 0.46940794043959017,
+      "learning_rate": 2.437183250258031e-06,
+      "loss": 0.2589,
+      "step": 4218
+    },
+    {
+      "epoch": 1.4086811352253756,
+      "grad_norm": 0.4736983766766649,
+      "learning_rate": 2.4346818058672645e-06,
+      "loss": 0.2565,
+      "step": 4219
+    },
+    {
+      "epoch": 1.4090150250417361,
+      "grad_norm": 0.4562442864306187,
+      "learning_rate": 2.4321812326452916e-06,
+      "loss": 0.25,
+      "step": 4220
+    },
+    {
+      "epoch": 1.4093489148580969,
+      "grad_norm": 0.4860411644491457,
+      "learning_rate": 2.4296815314412976e-06,
+      "loss": 0.249,
+      "step": 4221
+    },
+    {
+      "epoch": 1.4096828046744574,
+      "grad_norm": 0.47772927114794705,
+      "learning_rate": 2.4271827031041654e-06,
+      "loss": 0.2502,
+      "step": 4222
+    },
+    {
+      "epoch": 1.410016694490818,
+      "grad_norm": 0.4802401160819477,
+      "learning_rate": 2.424684748482483e-06,
+      "loss": 0.2622,
+      "step": 4223
+    },
+    {
+      "epoch": 1.4103505843071786,
+      "grad_norm": 0.4744569274239283,
+      "learning_rate": 2.422187668424547e-06,
+      "loss": 0.2415,
+      "step": 4224
+    },
+    {
+      "epoch": 1.4106844741235394,
+      "grad_norm": 0.4931346018283585,
+      "learning_rate": 2.4196914637783504e-06,
+      "loss": 0.2538,
+      "step": 4225
+    },
+    {
+      "epoch": 1.4110183639398999,
+      "grad_norm": 0.46405939018733877,
+      "learning_rate": 2.4171961353915914e-06,
+      "loss": 0.2576,
+      "step": 4226
+    },
+    {
+      "epoch": 1.4113522537562604,
+      "grad_norm": 0.4920100582061388,
+      "learning_rate": 2.4147016841116687e-06,
+      "loss": 0.2546,
+      "step": 4227
+    },
+    {
+      "epoch": 1.411686143572621,
+      "grad_norm": 0.4620944709260905,
+      "learning_rate": 2.412208110785689e-06,
+      "loss": 0.2535,
+      "step": 4228
+    },
+    {
+      "epoch": 1.4120200333889816,
+      "grad_norm": 0.467916160545023,
+      "learning_rate": 2.4097154162604537e-06,
+      "loss": 0.2449,
+      "step": 4229
+    },
+    {
+      "epoch": 1.4123539232053424,
+      "grad_norm": 0.47905133594140087,
+      "learning_rate": 2.4072236013824718e-06,
+      "loss": 0.2506,
+      "step": 4230
+    },
+    {
+      "epoch": 1.4126878130217029,
+      "grad_norm": 0.4622487848591893,
+      "learning_rate": 2.40473266699795e-06,
+      "loss": 0.2534,
+      "step": 4231
+    },
+    {
+      "epoch": 1.4130217028380634,
+      "grad_norm": 0.46569777618151853,
+      "learning_rate": 2.4022426139527934e-06,
+      "loss": 0.2516,
+      "step": 4232
+    },
+    {
+      "epoch": 1.4133555926544241,
+      "grad_norm": 0.4617690680805236,
+      "learning_rate": 2.3997534430926155e-06,
+      "loss": 0.2492,
+      "step": 4233
+    },
+    {
+      "epoch": 1.4136894824707846,
+      "grad_norm": 0.452145283563589,
+      "learning_rate": 2.397265155262724e-06,
+      "loss": 0.2514,
+      "step": 4234
+    },
+    {
+      "epoch": 1.4140233722871454,
+      "grad_norm": 0.4995005196778955,
+      "learning_rate": 2.394777751308129e-06,
+      "loss": 0.2573,
+      "step": 4235
+    },
+    {
+      "epoch": 1.4143572621035059,
+      "grad_norm": 0.48325727133757707,
+      "learning_rate": 2.392291232073538e-06,
+      "loss": 0.2507,
+      "step": 4236
+    },
+    {
+      "epoch": 1.4146911519198664,
+      "grad_norm": 0.4790481936836298,
+      "learning_rate": 2.3898055984033615e-06,
+      "loss": 0.2533,
+      "step": 4237
+    },
+    {
+      "epoch": 1.4150250417362271,
+      "grad_norm": 0.5023599006301053,
+      "learning_rate": 2.3873208511417102e-06,
+      "loss": 0.2617,
+      "step": 4238
+    },
+    {
+      "epoch": 1.4153589315525876,
+      "grad_norm": 0.4672814300293956,
+      "learning_rate": 2.384836991132389e-06,
+      "loss": 0.2497,
+      "step": 4239
+    },
+    {
+      "epoch": 1.4156928213689484,
+      "grad_norm": 0.45484846059195033,
+      "learning_rate": 2.3823540192189047e-06,
+      "loss": 0.2525,
+      "step": 4240
+    },
+    {
+      "epoch": 1.4160267111853089,
+      "grad_norm": 0.46952379368832364,
+      "learning_rate": 2.379871936244459e-06,
+      "loss": 0.2515,
+      "step": 4241
+    },
+    {
+      "epoch": 1.4163606010016694,
+      "grad_norm": 0.4832677485752608,
+      "learning_rate": 2.377390743051959e-06,
+      "loss": 0.2533,
+      "step": 4242
+    },
+    {
+      "epoch": 1.4166944908180301,
+      "grad_norm": 0.47309179581943844,
+      "learning_rate": 2.3749104404840023e-06,
+      "loss": 0.2418,
+      "step": 4243
+    },
+    {
+      "epoch": 1.4170283806343906,
+      "grad_norm": 0.4664673742992815,
+      "learning_rate": 2.372431029382888e-06,
+      "loss": 0.2474,
+      "step": 4244
+    },
+    {
+      "epoch": 1.4173622704507514,
+      "grad_norm": 0.47654883936099024,
+      "learning_rate": 2.3699525105906095e-06,
+      "loss": 0.2587,
+      "step": 4245
+    },
+    {
+      "epoch": 1.4176961602671119,
+      "grad_norm": 0.4789374949863525,
+      "learning_rate": 2.367474884948857e-06,
+      "loss": 0.245,
+      "step": 4246
+    },
+    {
+      "epoch": 1.4180300500834724,
+      "grad_norm": 0.49879300728084863,
+      "learning_rate": 2.364998153299027e-06,
+      "loss": 0.2511,
+      "step": 4247
+    },
+    {
+      "epoch": 1.4183639398998331,
+      "grad_norm": 0.46443603308234865,
+      "learning_rate": 2.3625223164821992e-06,
+      "loss": 0.25,
+      "step": 4248
+    },
+    {
+      "epoch": 1.4186978297161936,
+      "grad_norm": 0.5048939748015947,
+      "learning_rate": 2.360047375339156e-06,
+      "loss": 0.2602,
+      "step": 4249
+    },
+    {
+      "epoch": 1.4190317195325544,
+      "grad_norm": 0.4732310575403959,
+      "learning_rate": 2.357573330710373e-06,
+      "loss": 0.2555,
+      "step": 4250
+    },
+    {
+      "epoch": 1.4193656093489149,
+      "grad_norm": 0.4662739065262745,
+      "learning_rate": 2.3551001834360263e-06,
+      "loss": 0.2292,
+      "step": 4251
+    },
+    {
+      "epoch": 1.4196994991652754,
+      "grad_norm": 0.48137472448741164,
+      "learning_rate": 2.3526279343559826e-06,
+      "loss": 0.2636,
+      "step": 4252
+    },
+    {
+      "epoch": 1.4200333889816361,
+      "grad_norm": 0.4970615958171411,
+      "learning_rate": 2.350156584309804e-06,
+      "loss": 0.2539,
+      "step": 4253
+    },
+    {
+      "epoch": 1.4203672787979966,
+      "grad_norm": 0.4744830913941054,
+      "learning_rate": 2.3476861341367474e-06,
+      "loss": 0.2555,
+      "step": 4254
+    },
+    {
+      "epoch": 1.4207011686143574,
+      "grad_norm": 0.4812010248294851,
+      "learning_rate": 2.345216584675766e-06,
+      "loss": 0.249,
+      "step": 4255
+    },
+    {
+      "epoch": 1.4210350584307179,
+      "grad_norm": 0.4663465624523122,
+      "learning_rate": 2.3427479367655076e-06,
+      "loss": 0.2377,
+      "step": 4256
+    },
+    {
+      "epoch": 1.4213689482470784,
+      "grad_norm": 0.4844656594298752,
+      "learning_rate": 2.340280191244311e-06,
+      "loss": 0.2578,
+      "step": 4257
+    },
+    {
+      "epoch": 1.4217028380634391,
+      "grad_norm": 0.5055576821920061,
+      "learning_rate": 2.3378133489502105e-06,
+      "loss": 0.2495,
+      "step": 4258
+    },
+    {
+      "epoch": 1.4220367278797996,
+      "grad_norm": 0.4859658139451491,
+      "learning_rate": 2.335347410720931e-06,
+      "loss": 0.2619,
+      "step": 4259
+    },
+    {
+      "epoch": 1.4223706176961604,
+      "grad_norm": 0.4765495436115512,
+      "learning_rate": 2.332882377393893e-06,
+      "loss": 0.2651,
+      "step": 4260
+    },
+    {
+      "epoch": 1.4227045075125209,
+      "grad_norm": 0.4850044720059754,
+      "learning_rate": 2.3304182498062104e-06,
+      "loss": 0.2513,
+      "step": 4261
+    },
+    {
+      "epoch": 1.4230383973288814,
+      "grad_norm": 0.4623500934298817,
+      "learning_rate": 2.3279550287946883e-06,
+      "loss": 0.2543,
+      "step": 4262
+    },
+    {
+      "epoch": 1.423372287145242,
+      "grad_norm": 0.4779319869501573,
+      "learning_rate": 2.325492715195821e-06,
+      "loss": 0.2493,
+      "step": 4263
+    },
+    {
+      "epoch": 1.4237061769616026,
+      "grad_norm": 0.48007866686024253,
+      "learning_rate": 2.3230313098457992e-06,
+      "loss": 0.2531,
+      "step": 4264
+    },
+    {
+      "epoch": 1.4240400667779634,
+      "grad_norm": 0.48781536788362245,
+      "learning_rate": 2.320570813580505e-06,
+      "loss": 0.244,
+      "step": 4265
+    },
+    {
+      "epoch": 1.4243739565943239,
+      "grad_norm": 0.4771684673577719,
+      "learning_rate": 2.3181112272355095e-06,
+      "loss": 0.2532,
+      "step": 4266
+    },
+    {
+      "epoch": 1.4247078464106844,
+      "grad_norm": 0.49053462883093985,
+      "learning_rate": 2.315652551646075e-06,
+      "loss": 0.2523,
+      "step": 4267
+    },
+    {
+      "epoch": 1.425041736227045,
+      "grad_norm": 0.49308053932413176,
+      "learning_rate": 2.3131947876471543e-06,
+      "loss": 0.2548,
+      "step": 4268
+    },
+    {
+      "epoch": 1.4253756260434056,
+      "grad_norm": 0.4935027547973609,
+      "learning_rate": 2.3107379360733905e-06,
+      "loss": 0.259,
+      "step": 4269
+    },
+    {
+      "epoch": 1.4257095158597664,
+      "grad_norm": 0.4702552608010724,
+      "learning_rate": 2.308281997759121e-06,
+      "loss": 0.2455,
+      "step": 4270
+    },
+    {
+      "epoch": 1.4260434056761269,
+      "grad_norm": 0.4880188752299044,
+      "learning_rate": 2.305826973538366e-06,
+      "loss": 0.2523,
+      "step": 4271
+    },
+    {
+      "epoch": 1.4263772954924874,
+      "grad_norm": 0.47566074594779284,
+      "learning_rate": 2.3033728642448426e-06,
+      "loss": 0.2619,
+      "step": 4272
+    },
+    {
+      "epoch": 1.426711185308848,
+      "grad_norm": 0.4635722364121097,
+      "learning_rate": 2.3009196707119526e-06,
+      "loss": 0.2397,
+      "step": 4273
+    },
+    {
+      "epoch": 1.4270450751252086,
+      "grad_norm": 0.48303399275628167,
+      "learning_rate": 2.298467393772784e-06,
+      "loss": 0.2534,
+      "step": 4274
+    },
+    {
+      "epoch": 1.4273789649415694,
+      "grad_norm": 0.47978675685225497,
+      "learning_rate": 2.2960160342601234e-06,
+      "loss": 0.2561,
+      "step": 4275
+    },
+    {
+      "epoch": 1.4277128547579299,
+      "grad_norm": 0.4576912004552267,
+      "learning_rate": 2.293565593006437e-06,
+      "loss": 0.242,
+      "step": 4276
+    },
+    {
+      "epoch": 1.4280467445742904,
+      "grad_norm": 0.47657963467972647,
+      "learning_rate": 2.2911160708438816e-06,
+      "loss": 0.2611,
+      "step": 4277
+    },
+    {
+      "epoch": 1.428380634390651,
+      "grad_norm": 0.48845768305817133,
+      "learning_rate": 2.2886674686043008e-06,
+      "loss": 0.2626,
+      "step": 4278
+    },
+    {
+      "epoch": 1.4287145242070116,
+      "grad_norm": 0.4710980118952764,
+      "learning_rate": 2.2862197871192314e-06,
+      "loss": 0.2455,
+      "step": 4279
+    },
+    {
+      "epoch": 1.4290484140233723,
+      "grad_norm": 0.4740030442125757,
+      "learning_rate": 2.2837730272198886e-06,
+      "loss": 0.2528,
+      "step": 4280
+    },
+    {
+      "epoch": 1.4293823038397329,
+      "grad_norm": 0.4815923583392656,
+      "learning_rate": 2.281327189737184e-06,
+      "loss": 0.2454,
+      "step": 4281
+    },
+    {
+      "epoch": 1.4297161936560934,
+      "grad_norm": 0.4749707801892474,
+      "learning_rate": 2.278882275501709e-06,
+      "loss": 0.2514,
+      "step": 4282
+    },
+    {
+      "epoch": 1.430050083472454,
+      "grad_norm": 0.4582457559797428,
+      "learning_rate": 2.2764382853437427e-06,
+      "loss": 0.2339,
+      "step": 4283
+    },
+    {
+      "epoch": 1.4303839732888146,
+      "grad_norm": 0.4761179312197704,
+      "learning_rate": 2.2739952200932543e-06,
+      "loss": 0.2431,
+      "step": 4284
+    },
+    {
+      "epoch": 1.4307178631051753,
+      "grad_norm": 0.49537345926580834,
+      "learning_rate": 2.2715530805798945e-06,
+      "loss": 0.2551,
+      "step": 4285
+    },
+    {
+      "epoch": 1.4310517529215359,
+      "grad_norm": 0.504048125341069,
+      "learning_rate": 2.269111867633002e-06,
+      "loss": 0.2536,
+      "step": 4286
+    },
+    {
+      "epoch": 1.4313856427378964,
+      "grad_norm": 0.4698693098192748,
+      "learning_rate": 2.2666715820815975e-06,
+      "loss": 0.2471,
+      "step": 4287
+    },
+    {
+      "epoch": 1.431719532554257,
+      "grad_norm": 0.4792455966243841,
+      "learning_rate": 2.264232224754391e-06,
+      "loss": 0.2538,
+      "step": 4288
+    },
+    {
+      "epoch": 1.4320534223706178,
+      "grad_norm": 0.4750349894309066,
+      "learning_rate": 2.2617937964797786e-06,
+      "loss": 0.2521,
+      "step": 4289
+    },
+    {
+      "epoch": 1.4323873121869783,
+      "grad_norm": 0.4830441399312429,
+      "learning_rate": 2.2593562980858353e-06,
+      "loss": 0.2557,
+      "step": 4290
+    },
+    {
+      "epoch": 1.4327212020033389,
+      "grad_norm": 0.463170623624267,
+      "learning_rate": 2.2569197304003227e-06,
+      "loss": 0.251,
+      "step": 4291
+    },
+    {
+      "epoch": 1.4330550918196994,
+      "grad_norm": 0.47834739106232904,
+      "learning_rate": 2.254484094250685e-06,
+      "loss": 0.2416,
+      "step": 4292
+    },
+    {
+      "epoch": 1.43338898163606,
+      "grad_norm": 0.46554769119466016,
+      "learning_rate": 2.2520493904640555e-06,
+      "loss": 0.2446,
+      "step": 4293
+    },
+    {
+      "epoch": 1.4337228714524208,
+      "grad_norm": 0.47273668219912096,
+      "learning_rate": 2.2496156198672453e-06,
+      "loss": 0.2496,
+      "step": 4294
+    },
+    {
+      "epoch": 1.4340567612687813,
+      "grad_norm": 0.4681032238747119,
+      "learning_rate": 2.24718278328675e-06,
+      "loss": 0.2472,
+      "step": 4295
+    },
+    {
+      "epoch": 1.4343906510851419,
+      "grad_norm": 0.47532642379534196,
+      "learning_rate": 2.244750881548746e-06,
+      "loss": 0.2462,
+      "step": 4296
+    },
+    {
+      "epoch": 1.4347245409015026,
+      "grad_norm": 0.4922103168211831,
+      "learning_rate": 2.2423199154790975e-06,
+      "loss": 0.2516,
+      "step": 4297
+    },
+    {
+      "epoch": 1.435058430717863,
+      "grad_norm": 0.5009661451242415,
+      "learning_rate": 2.2398898859033496e-06,
+      "loss": 0.2522,
+      "step": 4298
+    },
+    {
+      "epoch": 1.4353923205342238,
+      "grad_norm": 0.4941378316422841,
+      "learning_rate": 2.237460793646725e-06,
+      "loss": 0.2503,
+      "step": 4299
+    },
+    {
+      "epoch": 1.4357262103505843,
+      "grad_norm": 0.48606384373178235,
+      "learning_rate": 2.235032639534132e-06,
+      "loss": 0.268,
+      "step": 4300
+    },
+    {
+      "epoch": 1.4360601001669449,
+      "grad_norm": 0.4789634901277727,
+      "learning_rate": 2.2326054243901567e-06,
+      "loss": 0.2589,
+      "step": 4301
+    },
+    {
+      "epoch": 1.4363939899833056,
+      "grad_norm": 0.4624580598487496,
+      "learning_rate": 2.2301791490390727e-06,
+      "loss": 0.2498,
+      "step": 4302
+    },
+    {
+      "epoch": 1.436727879799666,
+      "grad_norm": 0.4704095964428702,
+      "learning_rate": 2.227753814304829e-06,
+      "loss": 0.2506,
+      "step": 4303
+    },
+    {
+      "epoch": 1.4370617696160268,
+      "grad_norm": 0.4846989694572654,
+      "learning_rate": 2.225329421011056e-06,
+      "loss": 0.2482,
+      "step": 4304
+    },
+    {
+      "epoch": 1.4373956594323873,
+      "grad_norm": 0.47214767197882357,
+      "learning_rate": 2.222905969981064e-06,
+      "loss": 0.2518,
+      "step": 4305
+    },
+    {
+      "epoch": 1.4377295492487479,
+      "grad_norm": 0.4840503309457011,
+      "learning_rate": 2.2204834620378455e-06,
+      "loss": 0.2454,
+      "step": 4306
+    },
+    {
+      "epoch": 1.4380634390651086,
+      "grad_norm": 0.48041362640416957,
+      "learning_rate": 2.2180618980040747e-06,
+      "loss": 0.2587,
+      "step": 4307
+    },
+    {
+      "epoch": 1.438397328881469,
+      "grad_norm": 0.4854657233521175,
+      "learning_rate": 2.2156412787020993e-06,
+      "loss": 0.2543,
+      "step": 4308
+    },
+    {
+      "epoch": 1.4387312186978298,
+      "grad_norm": 0.4544265955427475,
+      "learning_rate": 2.2132216049539495e-06,
+      "loss": 0.2482,
+      "step": 4309
+    },
+    {
+      "epoch": 1.4390651085141903,
+      "grad_norm": 0.4651583186641174,
+      "learning_rate": 2.210802877581332e-06,
+      "loss": 0.2421,
+      "step": 4310
+    },
+    {
+      "epoch": 1.4393989983305508,
+      "grad_norm": 0.47978647363428384,
+      "learning_rate": 2.208385097405638e-06,
+      "loss": 0.2532,
+      "step": 4311
+    },
+    {
+      "epoch": 1.4397328881469116,
+      "grad_norm": 0.5079005600379547,
+      "learning_rate": 2.2059682652479313e-06,
+      "loss": 0.2586,
+      "step": 4312
+    },
+    {
+      "epoch": 1.440066777963272,
+      "grad_norm": 0.48105987249029053,
+      "learning_rate": 2.203552381928956e-06,
+      "loss": 0.2538,
+      "step": 4313
+    },
+    {
+      "epoch": 1.4404006677796328,
+      "grad_norm": 0.4790340978965464,
+      "learning_rate": 2.2011374482691306e-06,
+      "loss": 0.2522,
+      "step": 4314
+    },
+    {
+      "epoch": 1.4407345575959933,
+      "grad_norm": 0.4851346629868164,
+      "learning_rate": 2.198723465088557e-06,
+      "loss": 0.2522,
+      "step": 4315
+    },
+    {
+      "epoch": 1.4410684474123538,
+      "grad_norm": 0.46386142817173953,
+      "learning_rate": 2.196310433207013e-06,
+      "loss": 0.2434,
+      "step": 4316
+    },
+    {
+      "epoch": 1.4414023372287146,
+      "grad_norm": 0.4972945624078559,
+      "learning_rate": 2.1938983534439496e-06,
+      "loss": 0.2603,
+      "step": 4317
+    },
+    {
+      "epoch": 1.441736227045075,
+      "grad_norm": 0.4728063637700807,
+      "learning_rate": 2.191487226618497e-06,
+      "loss": 0.2469,
+      "step": 4318
+    },
+    {
+      "epoch": 1.4420701168614358,
+      "grad_norm": 0.49056850583170597,
+      "learning_rate": 2.189077053549459e-06,
+      "loss": 0.249,
+      "step": 4319
+    },
+    {
+      "epoch": 1.4424040066777963,
+      "grad_norm": 0.47670076451497795,
+      "learning_rate": 2.1866678350553215e-06,
+      "loss": 0.2517,
+      "step": 4320
+    },
+    {
+      "epoch": 1.4427378964941568,
+      "grad_norm": 0.4905156434055598,
+      "learning_rate": 2.184259571954242e-06,
+      "loss": 0.257,
+      "step": 4321
+    },
+    {
+      "epoch": 1.4430717863105176,
+      "grad_norm": 0.46683417935674976,
+      "learning_rate": 2.181852265064051e-06,
+      "loss": 0.2571,
+      "step": 4322
+    },
+    {
+      "epoch": 1.443405676126878,
+      "grad_norm": 0.49611520247322566,
+      "learning_rate": 2.179445915202262e-06,
+      "loss": 0.2518,
+      "step": 4323
+    },
+    {
+      "epoch": 1.4437395659432388,
+      "grad_norm": 0.47344274790615454,
+      "learning_rate": 2.1770405231860557e-06,
+      "loss": 0.2548,
+      "step": 4324
+    },
+    {
+      "epoch": 1.4440734557595993,
+      "grad_norm": 0.4769081646431022,
+      "learning_rate": 2.174636089832293e-06,
+      "loss": 0.24,
+      "step": 4325
+    },
+    {
+      "epoch": 1.4444073455759598,
+      "grad_norm": 0.4660359884662173,
+      "learning_rate": 2.1722326159575068e-06,
+      "loss": 0.2489,
+      "step": 4326
+    },
+    {
+      "epoch": 1.4447412353923206,
+      "grad_norm": 0.4785735737320632,
+      "learning_rate": 2.169830102377903e-06,
+      "loss": 0.2535,
+      "step": 4327
+    },
+    {
+      "epoch": 1.445075125208681,
+      "grad_norm": 0.4710838663113144,
+      "learning_rate": 2.1674285499093618e-06,
+      "loss": 0.2488,
+      "step": 4328
+    },
+    {
+      "epoch": 1.4454090150250418,
+      "grad_norm": 0.47496485985014997,
+      "learning_rate": 2.165027959367441e-06,
+      "loss": 0.2416,
+      "step": 4329
+    },
+    {
+      "epoch": 1.4457429048414023,
+      "grad_norm": 0.48207391415539413,
+      "learning_rate": 2.162628331567368e-06,
+      "loss": 0.2576,
+      "step": 4330
+    },
+    {
+      "epoch": 1.4460767946577628,
+      "grad_norm": 0.4735547587124692,
+      "learning_rate": 2.1602296673240416e-06,
+      "loss": 0.2602,
+      "step": 4331
+    },
+    {
+      "epoch": 1.4464106844741236,
+      "grad_norm": 0.4952788754891133,
+      "learning_rate": 2.1578319674520385e-06,
+      "loss": 0.2538,
+      "step": 4332
+    },
+    {
+      "epoch": 1.446744574290484,
+      "grad_norm": 0.48047217517416324,
+      "learning_rate": 2.155435232765603e-06,
+      "loss": 0.258,
+      "step": 4333
+    },
+    {
+      "epoch": 1.4470784641068448,
+      "grad_norm": 0.4708846646992267,
+      "learning_rate": 2.1530394640786567e-06,
+      "loss": 0.2627,
+      "step": 4334
+    },
+    {
+      "epoch": 1.4474123539232053,
+      "grad_norm": 0.47248383821973927,
+      "learning_rate": 2.1506446622047887e-06,
+      "loss": 0.2455,
+      "step": 4335
+    },
+    {
+      "epoch": 1.4477462437395658,
+      "grad_norm": 0.466822233848356,
+      "learning_rate": 2.148250827957261e-06,
+      "loss": 0.251,
+      "step": 4336
+    },
+    {
+      "epoch": 1.4480801335559266,
+      "grad_norm": 0.4594629197881194,
+      "learning_rate": 2.1458579621490056e-06,
+      "loss": 0.2519,
+      "step": 4337
+    },
+    {
+      "epoch": 1.448414023372287,
+      "grad_norm": 0.4665700656805018,
+      "learning_rate": 2.1434660655926316e-06,
+      "loss": 0.2503,
+      "step": 4338
+    },
+    {
+      "epoch": 1.4487479131886478,
+      "grad_norm": 0.4674848707107587,
+      "learning_rate": 2.1410751391004105e-06,
+      "loss": 0.2617,
+      "step": 4339
+    },
+    {
+      "epoch": 1.4490818030050083,
+      "grad_norm": 0.4824049001635254,
+      "learning_rate": 2.1386851834842916e-06,
+      "loss": 0.251,
+      "step": 4340
+    },
+    {
+      "epoch": 1.4494156928213688,
+      "grad_norm": 0.4541581042375257,
+      "learning_rate": 2.1362961995558904e-06,
+      "loss": 0.245,
+      "step": 4341
+    },
+    {
+      "epoch": 1.4497495826377296,
+      "grad_norm": 0.4700140584306345,
+      "learning_rate": 2.133908188126492e-06,
+      "loss": 0.2505,
+      "step": 4342
+    },
+    {
+      "epoch": 1.45008347245409,
+      "grad_norm": 0.4721719468201512,
+      "learning_rate": 2.131521150007056e-06,
+      "loss": 0.2478,
+      "step": 4343
+    },
+    {
+      "epoch": 1.4504173622704508,
+      "grad_norm": 0.4788069691934053,
+      "learning_rate": 2.1291350860082066e-06,
+      "loss": 0.2591,
+      "step": 4344
+    },
+    {
+      "epoch": 1.4507512520868113,
+      "grad_norm": 0.4647938396284026,
+      "learning_rate": 2.1267499969402383e-06,
+      "loss": 0.2426,
+      "step": 4345
+    },
+    {
+      "epoch": 1.4510851419031718,
+      "grad_norm": 0.4660740664880144,
+      "learning_rate": 2.1243658836131133e-06,
+      "loss": 0.2488,
+      "step": 4346
+    },
+    {
+      "epoch": 1.4514190317195326,
+      "grad_norm": 0.4781330658737238,
+      "learning_rate": 2.121982746836468e-06,
+      "loss": 0.2499,
+      "step": 4347
+    },
+    {
+      "epoch": 1.451752921535893,
+      "grad_norm": 0.4725935014940002,
+      "learning_rate": 2.1196005874195995e-06,
+      "loss": 0.2514,
+      "step": 4348
+    },
+    {
+      "epoch": 1.4520868113522538,
+      "grad_norm": 0.4800302544563655,
+      "learning_rate": 2.117219406171481e-06,
+      "loss": 0.2494,
+      "step": 4349
+    },
+    {
+      "epoch": 1.4524207011686143,
+      "grad_norm": 0.45635906823665784,
+      "learning_rate": 2.114839203900747e-06,
+      "loss": 0.2383,
+      "step": 4350
+    },
+    {
+      "epoch": 1.4527545909849748,
+      "grad_norm": 0.4864444156057404,
+      "learning_rate": 2.1124599814157006e-06,
+      "loss": 0.2619,
+      "step": 4351
+    },
+    {
+      "epoch": 1.4530884808013356,
+      "grad_norm": 0.46689193314248006,
+      "learning_rate": 2.110081739524316e-06,
+      "loss": 0.2483,
+      "step": 4352
+    },
+    {
+      "epoch": 1.453422370617696,
+      "grad_norm": 0.47161094218743405,
+      "learning_rate": 2.107704479034231e-06,
+      "loss": 0.245,
+      "step": 4353
+    },
+    {
+      "epoch": 1.4537562604340568,
+      "grad_norm": 0.45280380246226004,
+      "learning_rate": 2.1053282007527503e-06,
+      "loss": 0.247,
+      "step": 4354
+    },
+    {
+      "epoch": 1.4540901502504173,
+      "grad_norm": 0.504710469621831,
+      "learning_rate": 2.1029529054868447e-06,
+      "loss": 0.2695,
+      "step": 4355
+    },
+    {
+      "epoch": 1.4544240400667778,
+      "grad_norm": 0.4698896470486617,
+      "learning_rate": 2.1005785940431533e-06,
+      "loss": 0.2535,
+      "step": 4356
+    },
+    {
+      "epoch": 1.4547579298831386,
+      "grad_norm": 0.487017593669449,
+      "learning_rate": 2.0982052672279825e-06,
+      "loss": 0.256,
+      "step": 4357
+    },
+    {
+      "epoch": 1.4550918196994993,
+      "grad_norm": 0.46652547943086525,
+      "learning_rate": 2.095832925847299e-06,
+      "loss": 0.256,
+      "step": 4358
+    },
+    {
+      "epoch": 1.4554257095158598,
+      "grad_norm": 0.4790245184567173,
+      "learning_rate": 2.093461570706739e-06,
+      "loss": 0.2483,
+      "step": 4359
+    },
+    {
+      "epoch": 1.4557595993322203,
+      "grad_norm": 0.4788301360845562,
+      "learning_rate": 2.0910912026116e-06,
+      "loss": 0.258,
+      "step": 4360
+    },
+    {
+      "epoch": 1.4560934891485808,
+      "grad_norm": 0.45614715792345645,
+      "learning_rate": 2.0887218223668493e-06,
+      "loss": 0.2367,
+      "step": 4361
+    },
+    {
+      "epoch": 1.4564273789649416,
+      "grad_norm": 0.47780038054412055,
+      "learning_rate": 2.086353430777116e-06,
+      "loss": 0.261,
+      "step": 4362
+    },
+    {
+      "epoch": 1.4567612687813023,
+      "grad_norm": 0.4714050897053654,
+      "learning_rate": 2.0839860286466927e-06,
+      "loss": 0.2419,
+      "step": 4363
+    },
+    {
+      "epoch": 1.4570951585976628,
+      "grad_norm": 0.4705959259431321,
+      "learning_rate": 2.081619616779536e-06,
+      "loss": 0.2461,
+      "step": 4364
+    },
+    {
+      "epoch": 1.4574290484140233,
+      "grad_norm": 0.4714757843816254,
+      "learning_rate": 2.0792541959792676e-06,
+      "loss": 0.2501,
+      "step": 4365
+    },
+    {
+      "epoch": 1.457762938230384,
+      "grad_norm": 0.47310778355891425,
+      "learning_rate": 2.076889767049175e-06,
+      "loss": 0.2579,
+      "step": 4366
+    },
+    {
+      "epoch": 1.4580968280467446,
+      "grad_norm": 0.45410892888473653,
+      "learning_rate": 2.0745263307922043e-06,
+      "loss": 0.2419,
+      "step": 4367
+    },
+    {
+      "epoch": 1.4584307178631053,
+      "grad_norm": 0.45990088119572264,
+      "learning_rate": 2.072163888010966e-06,
+      "loss": 0.2491,
+      "step": 4368
+    },
+    {
+      "epoch": 1.4587646076794658,
+      "grad_norm": 0.48155862602452293,
+      "learning_rate": 2.0698024395077303e-06,
+      "loss": 0.2459,
+      "step": 4369
+    },
+    {
+      "epoch": 1.4590984974958263,
+      "grad_norm": 0.47562521897731563,
+      "learning_rate": 2.0674419860844385e-06,
+      "loss": 0.2373,
+      "step": 4370
+    },
+    {
+      "epoch": 1.459432387312187,
+      "grad_norm": 0.46093733969862405,
+      "learning_rate": 2.065082528542686e-06,
+      "loss": 0.2506,
+      "step": 4371
+    },
+    {
+      "epoch": 1.4597662771285476,
+      "grad_norm": 0.4977748483925918,
+      "learning_rate": 2.0627240676837325e-06,
+      "loss": 0.252,
+      "step": 4372
+    },
+    {
+      "epoch": 1.4601001669449083,
+      "grad_norm": 0.47234584707453403,
+      "learning_rate": 2.0603666043084964e-06,
+      "loss": 0.2595,
+      "step": 4373
+    },
+    {
+      "epoch": 1.4604340567612688,
+      "grad_norm": 0.4816484282207652,
+      "learning_rate": 2.0580101392175626e-06,
+      "loss": 0.2513,
+      "step": 4374
+    },
+    {
+      "epoch": 1.4607679465776293,
+      "grad_norm": 0.4851363740573362,
+      "learning_rate": 2.0556546732111758e-06,
+      "loss": 0.2553,
+      "step": 4375
+    },
+    {
+      "epoch": 1.46110183639399,
+      "grad_norm": 0.44908371085198473,
+      "learning_rate": 2.0533002070892395e-06,
+      "loss": 0.2481,
+      "step": 4376
+    },
+    {
+      "epoch": 1.4614357262103506,
+      "grad_norm": 0.4906374666964387,
+      "learning_rate": 2.050946741651317e-06,
+      "loss": 0.2575,
+      "step": 4377
+    },
+    {
+      "epoch": 1.4617696160267113,
+      "grad_norm": 0.48031174816968214,
+      "learning_rate": 2.0485942776966313e-06,
+      "loss": 0.2506,
+      "step": 4378
+    },
+    {
+      "epoch": 1.4621035058430718,
+      "grad_norm": 0.46985594226127125,
+      "learning_rate": 2.046242816024071e-06,
+      "loss": 0.2382,
+      "step": 4379
+    },
+    {
+      "epoch": 1.4624373956594323,
+      "grad_norm": 0.46671094819189113,
+      "learning_rate": 2.043892357432178e-06,
+      "loss": 0.2507,
+      "step": 4380
+    },
+    {
+      "epoch": 1.462771285475793,
+      "grad_norm": 0.48127889417101405,
+      "learning_rate": 2.041542902719157e-06,
+      "loss": 0.2556,
+      "step": 4381
+    },
+    {
+      "epoch": 1.4631051752921536,
+      "grad_norm": 0.4673466639614098,
+      "learning_rate": 2.039194452682867e-06,
+      "loss": 0.2548,
+      "step": 4382
+    },
+    {
+      "epoch": 1.4634390651085143,
+      "grad_norm": 0.4920561020232792,
+      "learning_rate": 2.0368470081208335e-06,
+      "loss": 0.2632,
+      "step": 4383
+    },
+    {
+      "epoch": 1.4637729549248748,
+      "grad_norm": 0.4907738394197905,
+      "learning_rate": 2.0345005698302367e-06,
+      "loss": 0.2567,
+      "step": 4384
+    },
+    {
+      "epoch": 1.4641068447412353,
+      "grad_norm": 0.46929735136382245,
+      "learning_rate": 2.032155138607914e-06,
+      "loss": 0.2598,
+      "step": 4385
+    },
+    {
+      "epoch": 1.464440734557596,
+      "grad_norm": 0.49328421161968933,
+      "learning_rate": 2.029810715250361e-06,
+      "loss": 0.2552,
+      "step": 4386
+    },
+    {
+      "epoch": 1.4647746243739566,
+      "grad_norm": 0.4691680161236366,
+      "learning_rate": 2.0274673005537305e-06,
+      "loss": 0.2459,
+      "step": 4387
+    },
+    {
+      "epoch": 1.4651085141903173,
+      "grad_norm": 0.4770583641746202,
+      "learning_rate": 2.0251248953138377e-06,
+      "loss": 0.2547,
+      "step": 4388
+    },
+    {
+      "epoch": 1.4654424040066778,
+      "grad_norm": 0.4623230437921756,
+      "learning_rate": 2.0227835003261486e-06,
+      "loss": 0.2487,
+      "step": 4389
+    },
+    {
+      "epoch": 1.4657762938230383,
+      "grad_norm": 0.4685642881709904,
+      "learning_rate": 2.0204431163857887e-06,
+      "loss": 0.2434,
+      "step": 4390
+    },
+    {
+      "epoch": 1.466110183639399,
+      "grad_norm": 0.46530817858371387,
+      "learning_rate": 2.0181037442875417e-06,
+      "loss": 0.2555,
+      "step": 4391
+    },
+    {
+      "epoch": 1.4664440734557596,
+      "grad_norm": 0.4713631480037795,
+      "learning_rate": 2.0157653848258445e-06,
+      "loss": 0.2521,
+      "step": 4392
+    },
+    {
+      "epoch": 1.4667779632721203,
+      "grad_norm": 0.4732353094528508,
+      "learning_rate": 2.0134280387947947e-06,
+      "loss": 0.2516,
+      "step": 4393
+    },
+    {
+      "epoch": 1.4671118530884808,
+      "grad_norm": 0.4729942179019217,
+      "learning_rate": 2.0110917069881413e-06,
+      "loss": 0.2445,
+      "step": 4394
+    },
+    {
+      "epoch": 1.4674457429048413,
+      "grad_norm": 0.47408391819401086,
+      "learning_rate": 2.008756390199291e-06,
+      "loss": 0.2466,
+      "step": 4395
+    },
+    {
+      "epoch": 1.467779632721202,
+      "grad_norm": 0.5118418984197962,
+      "learning_rate": 2.006422089221302e-06,
+      "loss": 0.2597,
+      "step": 4396
+    },
+    {
+      "epoch": 1.4681135225375626,
+      "grad_norm": 0.44673086034936643,
+      "learning_rate": 2.0040888048468954e-06,
+      "loss": 0.2545,
+      "step": 4397
+    },
+    {
+      "epoch": 1.4684474123539233,
+      "grad_norm": 0.46985908773719715,
+      "learning_rate": 2.001756537868441e-06,
+      "loss": 0.2478,
+      "step": 4398
+    },
+    {
+      "epoch": 1.4687813021702838,
+      "grad_norm": 0.4617772107859915,
+      "learning_rate": 1.999425289077963e-06,
+      "loss": 0.2463,
+      "step": 4399
+    },
+    {
+      "epoch": 1.4691151919866443,
+      "grad_norm": 0.4678891947679621,
+      "learning_rate": 1.997095059267144e-06,
+      "loss": 0.2553,
+      "step": 4400
+    },
+    {
+      "epoch": 1.469449081803005,
+      "grad_norm": 0.46594601839157423,
+      "learning_rate": 1.9947658492273153e-06,
+      "loss": 0.2509,
+      "step": 4401
+    },
+    {
+      "epoch": 1.4697829716193656,
+      "grad_norm": 0.4953728715922545,
+      "learning_rate": 1.9924376597494676e-06,
+      "loss": 0.2467,
+      "step": 4402
+    },
+    {
+      "epoch": 1.4701168614357263,
+      "grad_norm": 0.47723200650527814,
+      "learning_rate": 1.9901104916242403e-06,
+      "loss": 0.2515,
+      "step": 4403
+    },
+    {
+      "epoch": 1.4704507512520868,
+      "grad_norm": 0.4833657173873253,
+      "learning_rate": 1.987784345641928e-06,
+      "loss": 0.2455,
+      "step": 4404
+    },
+    {
+      "epoch": 1.4707846410684473,
+      "grad_norm": 0.4847017444278244,
+      "learning_rate": 1.985459222592476e-06,
+      "loss": 0.2625,
+      "step": 4405
+    },
+    {
+      "epoch": 1.471118530884808,
+      "grad_norm": 0.4498470571006663,
+      "learning_rate": 1.9831351232654874e-06,
+      "loss": 0.2498,
+      "step": 4406
+    },
+    {
+      "epoch": 1.4714524207011686,
+      "grad_norm": 0.4735918269874608,
+      "learning_rate": 1.98081204845021e-06,
+      "loss": 0.2449,
+      "step": 4407
+    },
+    {
+      "epoch": 1.4717863105175293,
+      "grad_norm": 0.48652999846086503,
+      "learning_rate": 1.978489998935554e-06,
+      "loss": 0.2572,
+      "step": 4408
+    },
+    {
+      "epoch": 1.4721202003338898,
+      "grad_norm": 0.4832385177958067,
+      "learning_rate": 1.976168975510071e-06,
+      "loss": 0.2521,
+      "step": 4409
+    },
+    {
+      "epoch": 1.4724540901502503,
+      "grad_norm": 0.5002485912092974,
+      "learning_rate": 1.973848978961968e-06,
+      "loss": 0.2512,
+      "step": 4410
+    },
+    {
+      "epoch": 1.472787979966611,
+      "grad_norm": 0.4592989424527345,
+      "learning_rate": 1.9715300100791073e-06,
+      "loss": 0.2447,
+      "step": 4411
+    },
+    {
+      "epoch": 1.4731218697829715,
+      "grad_norm": 0.45743338997627486,
+      "learning_rate": 1.969212069648997e-06,
+      "loss": 0.2567,
+      "step": 4412
+    },
+    {
+      "epoch": 1.4734557595993323,
+      "grad_norm": 0.47592299871096827,
+      "learning_rate": 1.966895158458797e-06,
+      "loss": 0.2457,
+      "step": 4413
+    },
+    {
+      "epoch": 1.4737896494156928,
+      "grad_norm": 0.462649307462366,
+      "learning_rate": 1.9645792772953175e-06,
+      "loss": 0.2428,
+      "step": 4414
+    },
+    {
+      "epoch": 1.4741235392320533,
+      "grad_norm": 0.4850723992958,
+      "learning_rate": 1.962264426945023e-06,
+      "loss": 0.2514,
+      "step": 4415
+    },
+    {
+      "epoch": 1.474457429048414,
+      "grad_norm": 0.48851125445296495,
+      "learning_rate": 1.9599506081940212e-06,
+      "loss": 0.2507,
+      "step": 4416
+    },
+    {
+      "epoch": 1.4747913188647745,
+      "grad_norm": 0.47502227089892635,
+      "learning_rate": 1.9576378218280763e-06,
+      "loss": 0.2488,
+      "step": 4417
+    },
+    {
+      "epoch": 1.4751252086811353,
+      "grad_norm": 0.47623076016169824,
+      "learning_rate": 1.955326068632597e-06,
+      "loss": 0.26,
+      "step": 4418
+    },
+    {
+      "epoch": 1.4754590984974958,
+      "grad_norm": 0.48309550978530835,
+      "learning_rate": 1.953015349392641e-06,
+      "loss": 0.2541,
+      "step": 4419
+    },
+    {
+      "epoch": 1.4757929883138563,
+      "grad_norm": 0.4873406104515571,
+      "learning_rate": 1.95070566489292e-06,
+      "loss": 0.2568,
+      "step": 4420
+    },
+    {
+      "epoch": 1.476126878130217,
+      "grad_norm": 0.48001011905762325,
+      "learning_rate": 1.9483970159177896e-06,
+      "loss": 0.2477,
+      "step": 4421
+    },
+    {
+      "epoch": 1.4764607679465778,
+      "grad_norm": 0.5120426280151493,
+      "learning_rate": 1.9460894032512538e-06,
+      "loss": 0.2455,
+      "step": 4422
+    },
+    {
+      "epoch": 1.4767946577629383,
+      "grad_norm": 0.471987912707281,
+      "learning_rate": 1.943782827676966e-06,
+      "loss": 0.2496,
+      "step": 4423
+    },
+    {
+      "epoch": 1.4771285475792988,
+      "grad_norm": 0.4763604795514734,
+      "learning_rate": 1.9414772899782274e-06,
+      "loss": 0.2555,
+      "step": 4424
+    },
+    {
+      "epoch": 1.4774624373956593,
+      "grad_norm": 0.4869813778450272,
+      "learning_rate": 1.9391727909379897e-06,
+      "loss": 0.2528,
+      "step": 4425
+    },
+    {
+      "epoch": 1.47779632721202,
+      "grad_norm": 0.48985441941283836,
+      "learning_rate": 1.9368693313388465e-06,
+      "loss": 0.256,
+      "step": 4426
+    },
+    {
+      "epoch": 1.4781302170283808,
+      "grad_norm": 0.46968315868271354,
+      "learning_rate": 1.9345669119630413e-06,
+      "loss": 0.2437,
+      "step": 4427
+    },
+    {
+      "epoch": 1.4784641068447413,
+      "grad_norm": 0.4884692863919192,
+      "learning_rate": 1.9322655335924605e-06,
+      "loss": 0.2578,
+      "step": 4428
+    },
+    {
+      "epoch": 1.4787979966611018,
+      "grad_norm": 0.48628623729409326,
+      "learning_rate": 1.9299651970086457e-06,
+      "loss": 0.2514,
+      "step": 4429
+    },
+    {
+      "epoch": 1.4791318864774623,
+      "grad_norm": 0.5027504568722786,
+      "learning_rate": 1.9276659029927768e-06,
+      "loss": 0.2594,
+      "step": 4430
+    },
+    {
+      "epoch": 1.479465776293823,
+      "grad_norm": 0.47291524163937815,
+      "learning_rate": 1.925367652325682e-06,
+      "loss": 0.2374,
+      "step": 4431
+    },
+    {
+      "epoch": 1.4797996661101838,
+      "grad_norm": 0.4714945473246823,
+      "learning_rate": 1.9230704457878335e-06,
+      "loss": 0.2416,
+      "step": 4432
+    },
+    {
+      "epoch": 1.4801335559265443,
+      "grad_norm": 0.479429782101291,
+      "learning_rate": 1.920774284159353e-06,
+      "loss": 0.2483,
+      "step": 4433
+    },
+    {
+      "epoch": 1.4804674457429048,
+      "grad_norm": 0.47695033112439517,
+      "learning_rate": 1.9184791682200064e-06,
+      "loss": 0.2353,
+      "step": 4434
+    },
+    {
+      "epoch": 1.4808013355592655,
+      "grad_norm": 0.4723945882620745,
+      "learning_rate": 1.916185098749202e-06,
+      "loss": 0.2615,
+      "step": 4435
+    },
+    {
+      "epoch": 1.481135225375626,
+      "grad_norm": 0.48435265686090057,
+      "learning_rate": 1.913892076525993e-06,
+      "loss": 0.2507,
+      "step": 4436
+    },
+    {
+      "epoch": 1.4814691151919868,
+      "grad_norm": 0.4797767623705524,
+      "learning_rate": 1.9116001023290776e-06,
+      "loss": 0.2429,
+      "step": 4437
+    },
+    {
+      "epoch": 1.4818030050083473,
+      "grad_norm": 0.4696082570692082,
+      "learning_rate": 1.9093091769368006e-06,
+      "loss": 0.2467,
+      "step": 4438
+    },
+    {
+      "epoch": 1.4821368948247078,
+      "grad_norm": 0.4625500759814401,
+      "learning_rate": 1.907019301127147e-06,
+      "loss": 0.2468,
+      "step": 4439
+    },
+    {
+      "epoch": 1.4824707846410685,
+      "grad_norm": 0.4970579927461097,
+      "learning_rate": 1.9047304756777473e-06,
+      "loss": 0.2592,
+      "step": 4440
+    },
+    {
+      "epoch": 1.482804674457429,
+      "grad_norm": 0.4639499018365728,
+      "learning_rate": 1.9024427013658725e-06,
+      "loss": 0.2438,
+      "step": 4441
+    },
+    {
+      "epoch": 1.4831385642737898,
+      "grad_norm": 0.4568543355596004,
+      "learning_rate": 1.9001559789684403e-06,
+      "loss": 0.255,
+      "step": 4442
+    },
+    {
+      "epoch": 1.4834724540901503,
+      "grad_norm": 0.48106740559999833,
+      "learning_rate": 1.8978703092620128e-06,
+      "loss": 0.2527,
+      "step": 4443
+    },
+    {
+      "epoch": 1.4838063439065108,
+      "grad_norm": 0.4763997376339374,
+      "learning_rate": 1.8955856930227884e-06,
+      "loss": 0.2501,
+      "step": 4444
+    },
+    {
+      "epoch": 1.4841402337228715,
+      "grad_norm": 0.4598168501760532,
+      "learning_rate": 1.893302131026612e-06,
+      "loss": 0.242,
+      "step": 4445
+    },
+    {
+      "epoch": 1.484474123539232,
+      "grad_norm": 0.4709963654152814,
+      "learning_rate": 1.8910196240489676e-06,
+      "loss": 0.2434,
+      "step": 4446
+    },
+    {
+      "epoch": 1.4848080133555928,
+      "grad_norm": 0.47734394710335465,
+      "learning_rate": 1.8887381728649856e-06,
+      "loss": 0.2572,
+      "step": 4447
+    },
+    {
+      "epoch": 1.4851419031719533,
+      "grad_norm": 0.4748401085669183,
+      "learning_rate": 1.8864577782494343e-06,
+      "loss": 0.2586,
+      "step": 4448
+    },
+    {
+      "epoch": 1.4854757929883138,
+      "grad_norm": 0.46300033396334506,
+      "learning_rate": 1.8841784409767228e-06,
+      "loss": 0.2467,
+      "step": 4449
+    },
+    {
+      "epoch": 1.4858096828046745,
+      "grad_norm": 0.48443294071703324,
+      "learning_rate": 1.881900161820901e-06,
+      "loss": 0.2495,
+      "step": 4450
+    },
+    {
+      "epoch": 1.486143572621035,
+      "grad_norm": 0.4706559044606471,
+      "learning_rate": 1.8796229415556628e-06,
+      "loss": 0.2512,
+      "step": 4451
+    },
+    {
+      "epoch": 1.4864774624373958,
+      "grad_norm": 0.47103875187809024,
+      "learning_rate": 1.8773467809543422e-06,
+      "loss": 0.2567,
+      "step": 4452
+    },
+    {
+      "epoch": 1.4868113522537563,
+      "grad_norm": 0.46756675968019984,
+      "learning_rate": 1.8750716807899089e-06,
+      "loss": 0.2544,
+      "step": 4453
+    },
+    {
+      "epoch": 1.4871452420701168,
+      "grad_norm": 0.45437928621924384,
+      "learning_rate": 1.8727976418349758e-06,
+      "loss": 0.2552,
+      "step": 4454
+    },
+    {
+      "epoch": 1.4874791318864775,
+      "grad_norm": 0.48098426696739927,
+      "learning_rate": 1.8705246648617953e-06,
+      "loss": 0.2467,
+      "step": 4455
+    },
+    {
+      "epoch": 1.487813021702838,
+      "grad_norm": 0.4568894144836691,
+      "learning_rate": 1.868252750642256e-06,
+      "loss": 0.2492,
+      "step": 4456
+    },
+    {
+      "epoch": 1.4881469115191988,
+      "grad_norm": 0.49475797300614444,
+      "learning_rate": 1.8659818999478918e-06,
+      "loss": 0.2554,
+      "step": 4457
+    },
+    {
+      "epoch": 1.4884808013355593,
+      "grad_norm": 0.5193110418773145,
+      "learning_rate": 1.8637121135498693e-06,
+      "loss": 0.2556,
+      "step": 4458
+    },
+    {
+      "epoch": 1.4888146911519198,
+      "grad_norm": 0.46129011846874973,
+      "learning_rate": 1.8614433922189984e-06,
+      "loss": 0.2463,
+      "step": 4459
+    },
+    {
+      "epoch": 1.4891485809682805,
+      "grad_norm": 0.47356313350641627,
+      "learning_rate": 1.859175736725724e-06,
+      "loss": 0.2369,
+      "step": 4460
+    },
+    {
+      "epoch": 1.489482470784641,
+      "grad_norm": 0.4659052358443554,
+      "learning_rate": 1.8569091478401286e-06,
+      "loss": 0.239,
+      "step": 4461
+    },
+    {
+      "epoch": 1.4898163606010018,
+      "grad_norm": 0.5017350289581284,
+      "learning_rate": 1.8546436263319379e-06,
+      "loss": 0.2672,
+      "step": 4462
+    },
+    {
+      "epoch": 1.4901502504173623,
+      "grad_norm": 0.4530900857367731,
+      "learning_rate": 1.8523791729705082e-06,
+      "loss": 0.2508,
+      "step": 4463
+    },
+    {
+      "epoch": 1.4904841402337228,
+      "grad_norm": 0.4623411167739793,
+      "learning_rate": 1.8501157885248373e-06,
+      "loss": 0.2448,
+      "step": 4464
+    },
+    {
+      "epoch": 1.4908180300500835,
+      "grad_norm": 0.4683118178297707,
+      "learning_rate": 1.8478534737635568e-06,
+      "loss": 0.2473,
+      "step": 4465
+    },
+    {
+      "epoch": 1.491151919866444,
+      "grad_norm": 0.4721136403932729,
+      "learning_rate": 1.845592229454941e-06,
+      "loss": 0.2532,
+      "step": 4466
+    },
+    {
+      "epoch": 1.4914858096828048,
+      "grad_norm": 0.47993560748760433,
+      "learning_rate": 1.8433320563668921e-06,
+      "loss": 0.2563,
+      "step": 4467
+    },
+    {
+      "epoch": 1.4918196994991653,
+      "grad_norm": 0.4910885281669451,
+      "learning_rate": 1.8410729552669582e-06,
+      "loss": 0.2504,
+      "step": 4468
+    },
+    {
+      "epoch": 1.4921535893155258,
+      "grad_norm": 0.4877683593554069,
+      "learning_rate": 1.8388149269223153e-06,
+      "loss": 0.2602,
+      "step": 4469
+    },
+    {
+      "epoch": 1.4924874791318865,
+      "grad_norm": 0.4699531374864247,
+      "learning_rate": 1.8365579720997779e-06,
+      "loss": 0.2528,
+      "step": 4470
+    },
+    {
+      "epoch": 1.492821368948247,
+      "grad_norm": 0.47710216290290125,
+      "learning_rate": 1.8343020915657977e-06,
+      "loss": 0.2643,
+      "step": 4471
+    },
+    {
+      "epoch": 1.4931552587646078,
+      "grad_norm": 0.4624210137604584,
+      "learning_rate": 1.8320472860864596e-06,
+      "loss": 0.2484,
+      "step": 4472
+    },
+    {
+      "epoch": 1.4934891485809683,
+      "grad_norm": 0.48337385812471684,
+      "learning_rate": 1.8297935564274844e-06,
+      "loss": 0.2521,
+      "step": 4473
+    },
+    {
+      "epoch": 1.4938230383973288,
+      "grad_norm": 0.45791498877021297,
+      "learning_rate": 1.827540903354224e-06,
+      "loss": 0.2499,
+      "step": 4474
+    },
+    {
+      "epoch": 1.4941569282136895,
+      "grad_norm": 0.4698985871238666,
+      "learning_rate": 1.8252893276316696e-06,
+      "loss": 0.2607,
+      "step": 4475
+    },
+    {
+      "epoch": 1.49449081803005,
+      "grad_norm": 0.4721463014940266,
+      "learning_rate": 1.823038830024448e-06,
+      "loss": 0.2577,
+      "step": 4476
+    },
+    {
+      "epoch": 1.4948247078464107,
+      "grad_norm": 0.47215234094260977,
+      "learning_rate": 1.8207894112968126e-06,
+      "loss": 0.2526,
+      "step": 4477
+    },
+    {
+      "epoch": 1.4951585976627713,
+      "grad_norm": 0.46327425627685315,
+      "learning_rate": 1.8185410722126556e-06,
+      "loss": 0.2622,
+      "step": 4478
+    },
+    {
+      "epoch": 1.4954924874791318,
+      "grad_norm": 0.47648503269232145,
+      "learning_rate": 1.8162938135354995e-06,
+      "loss": 0.2567,
+      "step": 4479
+    },
+    {
+      "epoch": 1.4958263772954925,
+      "grad_norm": 0.47047199514332855,
+      "learning_rate": 1.8140476360285053e-06,
+      "loss": 0.2463,
+      "step": 4480
+    },
+    {
+      "epoch": 1.496160267111853,
+      "grad_norm": 0.4628555391133361,
+      "learning_rate": 1.8118025404544614e-06,
+      "loss": 0.2478,
+      "step": 4481
+    },
+    {
+      "epoch": 1.4964941569282137,
+      "grad_norm": 0.4539291489464441,
+      "learning_rate": 1.8095585275757904e-06,
+      "loss": 0.2434,
+      "step": 4482
+    },
+    {
+      "epoch": 1.4968280467445743,
+      "grad_norm": 0.46345200178874035,
+      "learning_rate": 1.8073155981545465e-06,
+      "loss": 0.2446,
+      "step": 4483
+    },
+    {
+      "epoch": 1.4971619365609348,
+      "grad_norm": 0.45236943544990355,
+      "learning_rate": 1.8050737529524176e-06,
+      "loss": 0.2355,
+      "step": 4484
+    },
+    {
+      "epoch": 1.4974958263772955,
+      "grad_norm": 0.4554984289684917,
+      "learning_rate": 1.8028329927307247e-06,
+      "loss": 0.2583,
+      "step": 4485
+    },
+    {
+      "epoch": 1.497829716193656,
+      "grad_norm": 0.5024841788608543,
+      "learning_rate": 1.8005933182504175e-06,
+      "loss": 0.2618,
+      "step": 4486
+    },
+    {
+      "epoch": 1.4981636060100167,
+      "grad_norm": 0.47917334901535424,
+      "learning_rate": 1.7983547302720773e-06,
+      "loss": 0.2507,
+      "step": 4487
+    },
+    {
+      "epoch": 1.4984974958263773,
+      "grad_norm": 0.46874995084068355,
+      "learning_rate": 1.796117229555915e-06,
+      "loss": 0.2408,
+      "step": 4488
+    },
+    {
+      "epoch": 1.4988313856427378,
+      "grad_norm": 0.4698731173226217,
+      "learning_rate": 1.7938808168617776e-06,
+      "loss": 0.2423,
+      "step": 4489
+    },
+    {
+      "epoch": 1.4991652754590985,
+      "grad_norm": 0.4707640583843599,
+      "learning_rate": 1.791645492949139e-06,
+      "loss": 0.2588,
+      "step": 4490
+    },
+    {
+      "epoch": 1.4994991652754592,
+      "grad_norm": 0.5021723155713317,
+      "learning_rate": 1.7894112585771017e-06,
+      "loss": 0.2599,
+      "step": 4491
+    },
+    {
+      "epoch": 1.4998330550918197,
+      "grad_norm": 0.47357857532160363,
+      "learning_rate": 1.7871781145044e-06,
+      "loss": 0.2439,
+      "step": 4492
+    },
+    {
+      "epoch": 1.5001669449081803,
+      "grad_norm": 0.4647956183688045,
+      "learning_rate": 1.7849460614893987e-06,
+      "loss": 0.2484,
+      "step": 4493
+    },
+    {
+      "epoch": 1.5005008347245408,
+      "grad_norm": 0.48299194693085323,
+      "learning_rate": 1.7827151002900938e-06,
+      "loss": 0.2561,
+      "step": 4494
+    },
+    {
+      "epoch": 1.5008347245409015,
+      "grad_norm": 0.480504283478015,
+      "learning_rate": 1.7804852316641063e-06,
+      "loss": 0.2503,
+      "step": 4495
+    },
+    {
+      "epoch": 1.5011686143572622,
+      "grad_norm": 0.4751374577203341,
+      "learning_rate": 1.7782564563686882e-06,
+      "loss": 0.255,
+      "step": 4496
+    },
+    {
+      "epoch": 1.5015025041736227,
+      "grad_norm": 0.4666105522245239,
+      "learning_rate": 1.776028775160718e-06,
+      "loss": 0.2602,
+      "step": 4497
+    },
+    {
+      "epoch": 1.5018363939899833,
+      "grad_norm": 0.4782064425966174,
+      "learning_rate": 1.7738021887967083e-06,
+      "loss": 0.256,
+      "step": 4498
+    },
+    {
+      "epoch": 1.5021702838063438,
+      "grad_norm": 0.4782567157727596,
+      "learning_rate": 1.7715766980327947e-06,
+      "loss": 0.2621,
+      "step": 4499
+    },
+    {
+      "epoch": 1.5025041736227045,
+      "grad_norm": 0.4848948640936508,
+      "learning_rate": 1.7693523036247423e-06,
+      "loss": 0.2553,
+      "step": 4500
+    },
+    {
+      "epoch": 1.5028380634390652,
+      "grad_norm": 0.46822922777036863,
+      "learning_rate": 1.7671290063279423e-06,
+      "loss": 0.2501,
+      "step": 4501
+    },
+    {
+      "epoch": 1.5031719532554257,
+      "grad_norm": 0.4757759232152548,
+      "learning_rate": 1.7649068068974167e-06,
+      "loss": 0.2381,
+      "step": 4502
+    },
+    {
+      "epoch": 1.5035058430717863,
+      "grad_norm": 0.4876939015883604,
+      "learning_rate": 1.7626857060878143e-06,
+      "loss": 0.2482,
+      "step": 4503
+    },
+    {
+      "epoch": 1.5038397328881468,
+      "grad_norm": 0.4795944933491858,
+      "learning_rate": 1.760465704653408e-06,
+      "loss": 0.2565,
+      "step": 4504
+    },
+    {
+      "epoch": 1.5041736227045075,
+      "grad_norm": 0.46358622931014537,
+      "learning_rate": 1.7582468033480992e-06,
+      "loss": 0.2469,
+      "step": 4505
+    },
+    {
+      "epoch": 1.5045075125208682,
+      "grad_norm": 0.49042582246433536,
+      "learning_rate": 1.7560290029254129e-06,
+      "loss": 0.2514,
+      "step": 4506
+    },
+    {
+      "epoch": 1.5048414023372287,
+      "grad_norm": 0.4679557067026633,
+      "learning_rate": 1.7538123041385064e-06,
+      "loss": 0.2549,
+      "step": 4507
+    },
+    {
+      "epoch": 1.5051752921535893,
+      "grad_norm": 0.47687873974371503,
+      "learning_rate": 1.7515967077401574e-06,
+      "loss": 0.2568,
+      "step": 4508
+    },
+    {
+      "epoch": 1.5055091819699498,
+      "grad_norm": 0.47568092812138735,
+      "learning_rate": 1.7493822144827694e-06,
+      "loss": 0.2539,
+      "step": 4509
+    },
+    {
+      "epoch": 1.5058430717863105,
+      "grad_norm": 0.46364350833547924,
+      "learning_rate": 1.747168825118376e-06,
+      "loss": 0.2427,
+      "step": 4510
+    },
+    {
+      "epoch": 1.5061769616026712,
+      "grad_norm": 0.47423874330159926,
+      "learning_rate": 1.7449565403986301e-06,
+      "loss": 0.251,
+      "step": 4511
+    },
+    {
+      "epoch": 1.5065108514190317,
+      "grad_norm": 0.4634818560421081,
+      "learning_rate": 1.7427453610748146e-06,
+      "loss": 0.2467,
+      "step": 4512
+    },
+    {
+      "epoch": 1.5068447412353922,
+      "grad_norm": 0.4515530057758644,
+      "learning_rate": 1.7405352878978332e-06,
+      "loss": 0.2393,
+      "step": 4513
+    },
+    {
+      "epoch": 1.5071786310517528,
+      "grad_norm": 0.46467679117107713,
+      "learning_rate": 1.7383263216182155e-06,
+      "loss": 0.2421,
+      "step": 4514
+    },
+    {
+      "epoch": 1.5075125208681135,
+      "grad_norm": 0.4873317771655171,
+      "learning_rate": 1.7361184629861133e-06,
+      "loss": 0.2518,
+      "step": 4515
+    },
+    {
+      "epoch": 1.5078464106844742,
+      "grad_norm": 0.4530811702213615,
+      "learning_rate": 1.7339117127513072e-06,
+      "loss": 0.231,
+      "step": 4516
+    },
+    {
+      "epoch": 1.5081803005008347,
+      "grad_norm": 0.5002749128284556,
+      "learning_rate": 1.7317060716631967e-06,
+      "loss": 0.2627,
+      "step": 4517
+    },
+    {
+      "epoch": 1.5085141903171952,
+      "grad_norm": 0.4875094043584194,
+      "learning_rate": 1.7295015404708038e-06,
+      "loss": 0.2569,
+      "step": 4518
+    },
+    {
+      "epoch": 1.508848080133556,
+      "grad_norm": 0.4779680650698429,
+      "learning_rate": 1.7272981199227802e-06,
+      "loss": 0.244,
+      "step": 4519
+    },
+    {
+      "epoch": 1.5091819699499165,
+      "grad_norm": 0.4722453291627795,
+      "learning_rate": 1.725095810767392e-06,
+      "loss": 0.2474,
+      "step": 4520
+    },
+    {
+      "epoch": 1.5095158597662772,
+      "grad_norm": 0.4733220096595431,
+      "learning_rate": 1.7228946137525366e-06,
+      "loss": 0.2501,
+      "step": 4521
+    },
+    {
+      "epoch": 1.5098497495826377,
+      "grad_norm": 0.47265529540613593,
+      "learning_rate": 1.7206945296257255e-06,
+      "loss": 0.2548,
+      "step": 4522
+    },
+    {
+      "epoch": 1.5101836393989982,
+      "grad_norm": 0.4687953179837605,
+      "learning_rate": 1.7184955591340974e-06,
+      "loss": 0.2585,
+      "step": 4523
+    },
+    {
+      "epoch": 1.510517529215359,
+      "grad_norm": 0.47876212083530484,
+      "learning_rate": 1.7162977030244088e-06,
+      "loss": 0.2488,
+      "step": 4524
+    },
+    {
+      "epoch": 1.5108514190317195,
+      "grad_norm": 0.46486724324912226,
+      "learning_rate": 1.7141009620430442e-06,
+      "loss": 0.2491,
+      "step": 4525
+    },
+    {
+      "epoch": 1.5111853088480802,
+      "grad_norm": 0.46490553745849766,
+      "learning_rate": 1.7119053369360023e-06,
+      "loss": 0.2506,
+      "step": 4526
+    },
+    {
+      "epoch": 1.5115191986644407,
+      "grad_norm": 0.46482237732118425,
+      "learning_rate": 1.7097108284489084e-06,
+      "loss": 0.2543,
+      "step": 4527
+    },
+    {
+      "epoch": 1.5118530884808012,
+      "grad_norm": 0.46039917193536783,
+      "learning_rate": 1.7075174373270053e-06,
+      "loss": 0.2464,
+      "step": 4528
+    },
+    {
+      "epoch": 1.512186978297162,
+      "grad_norm": 0.47553306700946624,
+      "learning_rate": 1.7053251643151554e-06,
+      "loss": 0.2544,
+      "step": 4529
+    },
+    {
+      "epoch": 1.5125208681135225,
+      "grad_norm": 0.4703929691476318,
+      "learning_rate": 1.7031340101578469e-06,
+      "loss": 0.2481,
+      "step": 4530
+    },
+    {
+      "epoch": 1.5128547579298832,
+      "grad_norm": 0.4724975184880526,
+      "learning_rate": 1.7009439755991825e-06,
+      "loss": 0.2574,
+      "step": 4531
+    },
+    {
+      "epoch": 1.5131886477462437,
+      "grad_norm": 0.4906391253124488,
+      "learning_rate": 1.6987550613828863e-06,
+      "loss": 0.2638,
+      "step": 4532
+    },
+    {
+      "epoch": 1.5135225375626042,
+      "grad_norm": 0.46187323663441415,
+      "learning_rate": 1.6965672682523015e-06,
+      "loss": 0.2442,
+      "step": 4533
+    },
+    {
+      "epoch": 1.513856427378965,
+      "grad_norm": 0.46730744236975474,
+      "learning_rate": 1.6943805969503936e-06,
+      "loss": 0.25,
+      "step": 4534
+    },
+    {
+      "epoch": 1.5141903171953257,
+      "grad_norm": 0.48993853952918176,
+      "learning_rate": 1.6921950482197425e-06,
+      "loss": 0.2563,
+      "step": 4535
+    },
+    {
+      "epoch": 1.5145242070116862,
+      "grad_norm": 0.466545920148219,
+      "learning_rate": 1.6900106228025516e-06,
+      "loss": 0.251,
+      "step": 4536
+    },
+    {
+      "epoch": 1.5148580968280467,
+      "grad_norm": 0.4622693535329713,
+      "learning_rate": 1.6878273214406393e-06,
+      "loss": 0.2532,
+      "step": 4537
+    },
+    {
+      "epoch": 1.5151919866444072,
+      "grad_norm": 0.5022964418003112,
+      "learning_rate": 1.6856451448754423e-06,
+      "loss": 0.2525,
+      "step": 4538
+    },
+    {
+      "epoch": 1.515525876460768,
+      "grad_norm": 0.4922260525102104,
+      "learning_rate": 1.683464093848019e-06,
+      "loss": 0.2513,
+      "step": 4539
+    },
+    {
+      "epoch": 1.5158597662771287,
+      "grad_norm": 0.4810451057754166,
+      "learning_rate": 1.6812841690990423e-06,
+      "loss": 0.2491,
+      "step": 4540
+    },
+    {
+      "epoch": 1.5161936560934892,
+      "grad_norm": 0.4604073015168515,
+      "learning_rate": 1.6791053713688022e-06,
+      "loss": 0.2423,
+      "step": 4541
+    },
+    {
+      "epoch": 1.5165275459098497,
+      "grad_norm": 0.4651383090447779,
+      "learning_rate": 1.6769277013972064e-06,
+      "loss": 0.2504,
+      "step": 4542
+    },
+    {
+      "epoch": 1.5168614357262102,
+      "grad_norm": 0.46675105651932464,
+      "learning_rate": 1.6747511599237816e-06,
+      "loss": 0.2375,
+      "step": 4543
+    },
+    {
+      "epoch": 1.517195325542571,
+      "grad_norm": 0.47199397079244043,
+      "learning_rate": 1.6725757476876714e-06,
+      "loss": 0.2394,
+      "step": 4544
+    },
+    {
+      "epoch": 1.5175292153589317,
+      "grad_norm": 0.4787677014861461,
+      "learning_rate": 1.6704014654276335e-06,
+      "loss": 0.2552,
+      "step": 4545
+    },
+    {
+      "epoch": 1.5178631051752922,
+      "grad_norm": 0.48591784741933575,
+      "learning_rate": 1.6682283138820426e-06,
+      "loss": 0.2438,
+      "step": 4546
+    },
+    {
+      "epoch": 1.5181969949916527,
+      "grad_norm": 0.4625640738133976,
+      "learning_rate": 1.6660562937888886e-06,
+      "loss": 0.2581,
+      "step": 4547
+    },
+    {
+      "epoch": 1.5185308848080132,
+      "grad_norm": 0.459092372784831,
+      "learning_rate": 1.6638854058857806e-06,
+      "loss": 0.238,
+      "step": 4548
+    },
+    {
+      "epoch": 1.518864774624374,
+      "grad_norm": 0.46135162510649447,
+      "learning_rate": 1.6617156509099403e-06,
+      "loss": 0.2573,
+      "step": 4549
+    },
+    {
+      "epoch": 1.5191986644407347,
+      "grad_norm": 0.4680928254966776,
+      "learning_rate": 1.6595470295982047e-06,
+      "loss": 0.2492,
+      "step": 4550
+    },
+    {
+      "epoch": 1.5195325542570952,
+      "grad_norm": 0.47532693846286267,
+      "learning_rate": 1.6573795426870247e-06,
+      "loss": 0.2488,
+      "step": 4551
+    },
+    {
+      "epoch": 1.5198664440734557,
+      "grad_norm": 0.4498531261568033,
+      "learning_rate": 1.655213190912469e-06,
+      "loss": 0.2419,
+      "step": 4552
+    },
+    {
+      "epoch": 1.5202003338898162,
+      "grad_norm": 0.4539160995508871,
+      "learning_rate": 1.6530479750102224e-06,
+      "loss": 0.2479,
+      "step": 4553
+    },
+    {
+      "epoch": 1.520534223706177,
+      "grad_norm": 0.4719337608098083,
+      "learning_rate": 1.6508838957155782e-06,
+      "loss": 0.2514,
+      "step": 4554
+    },
+    {
+      "epoch": 1.5208681135225377,
+      "grad_norm": 0.47442997999852915,
+      "learning_rate": 1.6487209537634463e-06,
+      "loss": 0.26,
+      "step": 4555
+    },
+    {
+      "epoch": 1.5212020033388982,
+      "grad_norm": 0.47628795581300204,
+      "learning_rate": 1.6465591498883503e-06,
+      "loss": 0.2505,
+      "step": 4556
+    },
+    {
+      "epoch": 1.5215358931552587,
+      "grad_norm": 0.47957035302856826,
+      "learning_rate": 1.64439848482443e-06,
+      "loss": 0.2473,
+      "step": 4557
+    },
+    {
+      "epoch": 1.5218697829716192,
+      "grad_norm": 0.5032024668738625,
+      "learning_rate": 1.6422389593054338e-06,
+      "loss": 0.2531,
+      "step": 4558
+    },
+    {
+      "epoch": 1.52220367278798,
+      "grad_norm": 0.47360166035536383,
+      "learning_rate": 1.6400805740647269e-06,
+      "loss": 0.2523,
+      "step": 4559
+    },
+    {
+      "epoch": 1.5225375626043407,
+      "grad_norm": 0.46351516657791225,
+      "learning_rate": 1.6379233298352832e-06,
+      "loss": 0.2501,
+      "step": 4560
+    },
+    {
+      "epoch": 1.5228714524207012,
+      "grad_norm": 0.4659034808188622,
+      "learning_rate": 1.6357672273496932e-06,
+      "loss": 0.2554,
+      "step": 4561
+    },
+    {
+      "epoch": 1.5232053422370617,
+      "grad_norm": 0.48055467140014,
+      "learning_rate": 1.6336122673401606e-06,
+      "loss": 0.2516,
+      "step": 4562
+    },
+    {
+      "epoch": 1.5235392320534222,
+      "grad_norm": 0.4777854932173947,
+      "learning_rate": 1.6314584505384961e-06,
+      "loss": 0.2706,
+      "step": 4563
+    },
+    {
+      "epoch": 1.523873121869783,
+      "grad_norm": 0.4637277488362835,
+      "learning_rate": 1.6293057776761257e-06,
+      "loss": 0.2574,
+      "step": 4564
+    },
+    {
+      "epoch": 1.5242070116861437,
+      "grad_norm": 0.47139509310482514,
+      "learning_rate": 1.6271542494840832e-06,
+      "loss": 0.2477,
+      "step": 4565
+    },
+    {
+      "epoch": 1.5245409015025042,
+      "grad_norm": 0.4580859797257306,
+      "learning_rate": 1.6250038666930207e-06,
+      "loss": 0.242,
+      "step": 4566
+    },
+    {
+      "epoch": 1.5248747913188647,
+      "grad_norm": 0.4611570619871044,
+      "learning_rate": 1.6228546300331953e-06,
+      "loss": 0.2537,
+      "step": 4567
+    },
+    {
+      "epoch": 1.5252086811352252,
+      "grad_norm": 0.47107349120141234,
+      "learning_rate": 1.6207065402344746e-06,
+      "loss": 0.2434,
+      "step": 4568
+    },
+    {
+      "epoch": 1.525542570951586,
+      "grad_norm": 0.4897488261922242,
+      "learning_rate": 1.6185595980263423e-06,
+      "loss": 0.2429,
+      "step": 4569
+    },
+    {
+      "epoch": 1.5258764607679467,
+      "grad_norm": 0.48926019720568775,
+      "learning_rate": 1.6164138041378857e-06,
+      "loss": 0.2614,
+      "step": 4570
+    },
+    {
+      "epoch": 1.5262103505843072,
+      "grad_norm": 0.48230053067808937,
+      "learning_rate": 1.6142691592978083e-06,
+      "loss": 0.2466,
+      "step": 4571
+    },
+    {
+      "epoch": 1.5265442404006677,
+      "grad_norm": 0.4566272673669822,
+      "learning_rate": 1.6121256642344186e-06,
+      "loss": 0.2384,
+      "step": 4572
+    },
+    {
+      "epoch": 1.5268781302170282,
+      "grad_norm": 0.4857859140890101,
+      "learning_rate": 1.609983319675637e-06,
+      "loss": 0.2644,
+      "step": 4573
+    },
+    {
+      "epoch": 1.527212020033389,
+      "grad_norm": 0.4631749072788222,
+      "learning_rate": 1.6078421263489914e-06,
+      "loss": 0.2472,
+      "step": 4574
+    },
+    {
+      "epoch": 1.5275459098497497,
+      "grad_norm": 0.4817191468203982,
+      "learning_rate": 1.605702084981622e-06,
+      "loss": 0.2508,
+      "step": 4575
+    },
+    {
+      "epoch": 1.5278797996661102,
+      "grad_norm": 0.45570465389945425,
+      "learning_rate": 1.6035631963002745e-06,
+      "loss": 0.2403,
+      "step": 4576
+    },
+    {
+      "epoch": 1.5282136894824707,
+      "grad_norm": 0.4698285820526085,
+      "learning_rate": 1.6014254610313035e-06,
+      "loss": 0.2506,
+      "step": 4577
+    },
+    {
+      "epoch": 1.5285475792988312,
+      "grad_norm": 0.45719471555856195,
+      "learning_rate": 1.5992888799006756e-06,
+      "loss": 0.2473,
+      "step": 4578
+    },
+    {
+      "epoch": 1.528881469115192,
+      "grad_norm": 0.46778771979297173,
+      "learning_rate": 1.59715345363396e-06,
+      "loss": 0.2482,
+      "step": 4579
+    },
+    {
+      "epoch": 1.5292153589315527,
+      "grad_norm": 0.47309099573067775,
+      "learning_rate": 1.5950191829563382e-06,
+      "loss": 0.2448,
+      "step": 4580
+    },
+    {
+      "epoch": 1.5295492487479132,
+      "grad_norm": 0.4651437758675542,
+      "learning_rate": 1.592886068592598e-06,
+      "loss": 0.2487,
+      "step": 4581
+    },
+    {
+      "epoch": 1.5298831385642737,
+      "grad_norm": 0.4986445252400927,
+      "learning_rate": 1.5907541112671321e-06,
+      "loss": 0.2418,
+      "step": 4582
+    },
+    {
+      "epoch": 1.5302170283806342,
+      "grad_norm": 0.4700102931221495,
+      "learning_rate": 1.5886233117039412e-06,
+      "loss": 0.249,
+      "step": 4583
+    },
+    {
+      "epoch": 1.530550918196995,
+      "grad_norm": 0.46398988382397777,
+      "learning_rate": 1.5864936706266376e-06,
+      "loss": 0.2514,
+      "step": 4584
+    },
+    {
+      "epoch": 1.5308848080133557,
+      "grad_norm": 0.4685126819211024,
+      "learning_rate": 1.5843651887584326e-06,
+      "loss": 0.2452,
+      "step": 4585
+    },
+    {
+      "epoch": 1.5312186978297162,
+      "grad_norm": 0.4829869740124558,
+      "learning_rate": 1.5822378668221511e-06,
+      "loss": 0.2517,
+      "step": 4586
+    },
+    {
+      "epoch": 1.5315525876460767,
+      "grad_norm": 0.4722773638457389,
+      "learning_rate": 1.5801117055402193e-06,
+      "loss": 0.2414,
+      "step": 4587
+    },
+    {
+      "epoch": 1.5318864774624374,
+      "grad_norm": 0.45651668399744616,
+      "learning_rate": 1.5779867056346682e-06,
+      "loss": 0.2417,
+      "step": 4588
+    },
+    {
+      "epoch": 1.532220367278798,
+      "grad_norm": 0.46779828614221747,
+      "learning_rate": 1.5758628678271403e-06,
+      "loss": 0.2484,
+      "step": 4589
+    },
+    {
+      "epoch": 1.5325542570951587,
+      "grad_norm": 0.464226874700506,
+      "learning_rate": 1.5737401928388785e-06,
+      "loss": 0.26,
+      "step": 4590
+    },
+    {
+      "epoch": 1.5328881469115192,
+      "grad_norm": 0.47084705522004905,
+      "learning_rate": 1.5716186813907325e-06,
+      "loss": 0.2528,
+      "step": 4591
+    },
+    {
+      "epoch": 1.5332220367278797,
+      "grad_norm": 0.47171908349381647,
+      "learning_rate": 1.569498334203154e-06,
+      "loss": 0.2408,
+      "step": 4592
+    },
+    {
+      "epoch": 1.5335559265442404,
+      "grad_norm": 0.4436044202479731,
+      "learning_rate": 1.5673791519962061e-06,
+      "loss": 0.2434,
+      "step": 4593
+    },
+    {
+      "epoch": 1.533889816360601,
+      "grad_norm": 0.4903380616925631,
+      "learning_rate": 1.565261135489549e-06,
+      "loss": 0.2558,
+      "step": 4594
+    },
+    {
+      "epoch": 1.5342237061769617,
+      "grad_norm": 0.46759896455631217,
+      "learning_rate": 1.563144285402453e-06,
+      "loss": 0.248,
+      "step": 4595
+    },
+    {
+      "epoch": 1.5345575959933222,
+      "grad_norm": 0.4648059450277117,
+      "learning_rate": 1.561028602453788e-06,
+      "loss": 0.2521,
+      "step": 4596
+    },
+    {
+      "epoch": 1.5348914858096827,
+      "grad_norm": 0.46879976511111554,
+      "learning_rate": 1.5589140873620267e-06,
+      "loss": 0.2428,
+      "step": 4597
+    },
+    {
+      "epoch": 1.5352253756260434,
+      "grad_norm": 0.4737355855554533,
+      "learning_rate": 1.5568007408452519e-06,
+      "loss": 0.2538,
+      "step": 4598
+    },
+    {
+      "epoch": 1.535559265442404,
+      "grad_norm": 0.44833429520608314,
+      "learning_rate": 1.5546885636211422e-06,
+      "loss": 0.2391,
+      "step": 4599
+    },
+    {
+      "epoch": 1.5358931552587647,
+      "grad_norm": 0.4818136961483913,
+      "learning_rate": 1.5525775564069833e-06,
+      "loss": 0.2446,
+      "step": 4600
+    },
+    {
+      "epoch": 1.5362270450751252,
+      "grad_norm": 0.4674806686183336,
+      "learning_rate": 1.55046771991966e-06,
+      "loss": 0.2479,
+      "step": 4601
+    },
+    {
+      "epoch": 1.5365609348914857,
+      "grad_norm": 0.46131085027064117,
+      "learning_rate": 1.5483590548756633e-06,
+      "loss": 0.2376,
+      "step": 4602
+    },
+    {
+      "epoch": 1.5368948247078464,
+      "grad_norm": 0.47748459875524385,
+      "learning_rate": 1.5462515619910873e-06,
+      "loss": 0.2483,
+      "step": 4603
+    },
+    {
+      "epoch": 1.5372287145242072,
+      "grad_norm": 0.4615221691019839,
+      "learning_rate": 1.5441452419816238e-06,
+      "loss": 0.2369,
+      "step": 4604
+    },
+    {
+      "epoch": 1.5375626043405677,
+      "grad_norm": 0.4843776159512481,
+      "learning_rate": 1.542040095562568e-06,
+      "loss": 0.2669,
+      "step": 4605
+    },
+    {
+      "epoch": 1.5378964941569282,
+      "grad_norm": 0.45852715047358605,
+      "learning_rate": 1.5399361234488154e-06,
+      "loss": 0.2463,
+      "step": 4606
+    },
+    {
+      "epoch": 1.5382303839732887,
+      "grad_norm": 0.4965861622021677,
+      "learning_rate": 1.5378333263548667e-06,
+      "loss": 0.2486,
+      "step": 4607
+    },
+    {
+      "epoch": 1.5385642737896494,
+      "grad_norm": 0.45521960099797165,
+      "learning_rate": 1.5357317049948194e-06,
+      "loss": 0.2427,
+      "step": 4608
+    },
+    {
+      "epoch": 1.5388981636060102,
+      "grad_norm": 0.4697900570323403,
+      "learning_rate": 1.5336312600823743e-06,
+      "loss": 0.2521,
+      "step": 4609
+    },
+    {
+      "epoch": 1.5392320534223707,
+      "grad_norm": 0.49229246278277233,
+      "learning_rate": 1.5315319923308291e-06,
+      "loss": 0.2457,
+      "step": 4610
+    },
+    {
+      "epoch": 1.5395659432387312,
+      "grad_norm": 0.47603001983885235,
+      "learning_rate": 1.5294339024530858e-06,
+      "loss": 0.2471,
+      "step": 4611
+    },
+    {
+      "epoch": 1.5398998330550917,
+      "grad_norm": 0.4686060298314569,
+      "learning_rate": 1.5273369911616465e-06,
+      "loss": 0.2479,
+      "step": 4612
+    },
+    {
+      "epoch": 1.5402337228714524,
+      "grad_norm": 0.4748179216256042,
+      "learning_rate": 1.5252412591686105e-06,
+      "loss": 0.2593,
+      "step": 4613
+    },
+    {
+      "epoch": 1.5405676126878132,
+      "grad_norm": 0.4831150906451978,
+      "learning_rate": 1.523146707185677e-06,
+      "loss": 0.2637,
+      "step": 4614
+    },
+    {
+      "epoch": 1.5409015025041737,
+      "grad_norm": 0.4769026921310551,
+      "learning_rate": 1.5210533359241436e-06,
+      "loss": 0.2532,
+      "step": 4615
+    },
+    {
+      "epoch": 1.5412353923205342,
+      "grad_norm": 0.48217876345260036,
+      "learning_rate": 1.5189611460949112e-06,
+      "loss": 0.2497,
+      "step": 4616
+    },
+    {
+      "epoch": 1.5415692821368947,
+      "grad_norm": 0.49599163622279163,
+      "learning_rate": 1.5168701384084754e-06,
+      "loss": 0.2515,
+      "step": 4617
+    },
+    {
+      "epoch": 1.5419031719532554,
+      "grad_norm": 0.46016756075907345,
+      "learning_rate": 1.5147803135749318e-06,
+      "loss": 0.2378,
+      "step": 4618
+    },
+    {
+      "epoch": 1.5422370617696162,
+      "grad_norm": 0.4717927157978281,
+      "learning_rate": 1.512691672303972e-06,
+      "loss": 0.2572,
+      "step": 4619
+    },
+    {
+      "epoch": 1.5425709515859767,
+      "grad_norm": 0.4790831620565414,
+      "learning_rate": 1.5106042153048894e-06,
+      "loss": 0.2528,
+      "step": 4620
+    },
+    {
+      "epoch": 1.5429048414023372,
+      "grad_norm": 0.480933654484693,
+      "learning_rate": 1.508517943286576e-06,
+      "loss": 0.2477,
+      "step": 4621
+    },
+    {
+      "epoch": 1.5432387312186977,
+      "grad_norm": 0.49114227486932427,
+      "learning_rate": 1.5064328569575166e-06,
+      "loss": 0.2548,
+      "step": 4622
+    },
+    {
+      "epoch": 1.5435726210350584,
+      "grad_norm": 0.4654549305857276,
+      "learning_rate": 1.5043489570257957e-06,
+      "loss": 0.245,
+      "step": 4623
+    },
+    {
+      "epoch": 1.5439065108514192,
+      "grad_norm": 0.47216837983841203,
+      "learning_rate": 1.5022662441990938e-06,
+      "loss": 0.2551,
+      "step": 4624
+    },
+    {
+      "epoch": 1.5442404006677797,
+      "grad_norm": 0.4681937442165451,
+      "learning_rate": 1.5001847191846924e-06,
+      "loss": 0.2475,
+      "step": 4625
+    },
+    {
+      "epoch": 1.5445742904841402,
+      "grad_norm": 0.4625875450306863,
+      "learning_rate": 1.4981043826894648e-06,
+      "loss": 0.2389,
+      "step": 4626
+    },
+    {
+      "epoch": 1.5449081803005007,
+      "grad_norm": 0.478241738330973,
+      "learning_rate": 1.4960252354198828e-06,
+      "loss": 0.2485,
+      "step": 4627
+    },
+    {
+      "epoch": 1.5452420701168614,
+      "grad_norm": 0.4804363443349148,
+      "learning_rate": 1.4939472780820125e-06,
+      "loss": 0.2572,
+      "step": 4628
+    },
+    {
+      "epoch": 1.5455759599332222,
+      "grad_norm": 0.45679806577462323,
+      "learning_rate": 1.491870511381519e-06,
+      "loss": 0.2442,
+      "step": 4629
+    },
+    {
+      "epoch": 1.5459098497495827,
+      "grad_norm": 0.48048186665392884,
+      "learning_rate": 1.4897949360236635e-06,
+      "loss": 0.247,
+      "step": 4630
+    },
+    {
+      "epoch": 1.5462437395659432,
+      "grad_norm": 0.4863643413865315,
+      "learning_rate": 1.4877205527132983e-06,
+      "loss": 0.2401,
+      "step": 4631
+    },
+    {
+      "epoch": 1.5465776293823037,
+      "grad_norm": 0.48310680923317695,
+      "learning_rate": 1.4856473621548733e-06,
+      "loss": 0.2495,
+      "step": 4632
+    },
+    {
+      "epoch": 1.5469115191986644,
+      "grad_norm": 0.4718380849321278,
+      "learning_rate": 1.4835753650524326e-06,
+      "loss": 0.2492,
+      "step": 4633
+    },
+    {
+      "epoch": 1.5472454090150252,
+      "grad_norm": 0.4991598665193056,
+      "learning_rate": 1.481504562109618e-06,
+      "loss": 0.256,
+      "step": 4634
+    },
+    {
+      "epoch": 1.5475792988313857,
+      "grad_norm": 0.47037707052788724,
+      "learning_rate": 1.4794349540296622e-06,
+      "loss": 0.2438,
+      "step": 4635
+    },
+    {
+      "epoch": 1.5479131886477462,
+      "grad_norm": 0.4511256455285778,
+      "learning_rate": 1.4773665415153915e-06,
+      "loss": 0.2408,
+      "step": 4636
+    },
+    {
+      "epoch": 1.5482470784641067,
+      "grad_norm": 0.479635708775768,
+      "learning_rate": 1.475299325269232e-06,
+      "loss": 0.2525,
+      "step": 4637
+    },
+    {
+      "epoch": 1.5485809682804674,
+      "grad_norm": 0.452573121120766,
+      "learning_rate": 1.4732333059931958e-06,
+      "loss": 0.2445,
+      "step": 4638
+    },
+    {
+      "epoch": 1.5489148580968282,
+      "grad_norm": 0.45525702840186166,
+      "learning_rate": 1.4711684843888963e-06,
+      "loss": 0.2395,
+      "step": 4639
+    },
+    {
+      "epoch": 1.5492487479131887,
+      "grad_norm": 0.4458718263713089,
+      "learning_rate": 1.4691048611575337e-06,
+      "loss": 0.2427,
+      "step": 4640
+    },
+    {
+      "epoch": 1.5495826377295492,
+      "grad_norm": 0.47782423900299037,
+      "learning_rate": 1.4670424369999043e-06,
+      "loss": 0.2504,
+      "step": 4641
+    },
+    {
+      "epoch": 1.5499165275459097,
+      "grad_norm": 0.49208940372816307,
+      "learning_rate": 1.4649812126163964e-06,
+      "loss": 0.2481,
+      "step": 4642
+    },
+    {
+      "epoch": 1.5502504173622704,
+      "grad_norm": 0.46913321859388674,
+      "learning_rate": 1.46292118870699e-06,
+      "loss": 0.2459,
+      "step": 4643
+    },
+    {
+      "epoch": 1.5505843071786312,
+      "grad_norm": 0.4855653287242945,
+      "learning_rate": 1.4608623659712612e-06,
+      "loss": 0.2509,
+      "step": 4644
+    },
+    {
+      "epoch": 1.5509181969949917,
+      "grad_norm": 0.48586557205219444,
+      "learning_rate": 1.458804745108373e-06,
+      "loss": 0.2446,
+      "step": 4645
+    },
+    {
+      "epoch": 1.5512520868113522,
+      "grad_norm": 0.4724785755642919,
+      "learning_rate": 1.4567483268170856e-06,
+      "loss": 0.2462,
+      "step": 4646
+    },
+    {
+      "epoch": 1.5515859766277127,
+      "grad_norm": 0.4904578871448311,
+      "learning_rate": 1.4546931117957452e-06,
+      "loss": 0.2566,
+      "step": 4647
+    },
+    {
+      "epoch": 1.5519198664440734,
+      "grad_norm": 0.48141940309536313,
+      "learning_rate": 1.4526391007422946e-06,
+      "loss": 0.2483,
+      "step": 4648
+    },
+    {
+      "epoch": 1.5522537562604342,
+      "grad_norm": 0.4696117419606062,
+      "learning_rate": 1.4505862943542643e-06,
+      "loss": 0.2559,
+      "step": 4649
+    },
+    {
+      "epoch": 1.5525876460767947,
+      "grad_norm": 0.4724851479154182,
+      "learning_rate": 1.4485346933287768e-06,
+      "loss": 0.2499,
+      "step": 4650
+    },
+    {
+      "epoch": 1.5529215358931552,
+      "grad_norm": 0.46401169334851694,
+      "learning_rate": 1.446484298362545e-06,
+      "loss": 0.2484,
+      "step": 4651
+    },
+    {
+      "epoch": 1.553255425709516,
+      "grad_norm": 0.4828950591497319,
+      "learning_rate": 1.444435110151871e-06,
+      "loss": 0.2598,
+      "step": 4652
+    },
+    {
+      "epoch": 1.5535893155258764,
+      "grad_norm": 0.48355092505081154,
+      "learning_rate": 1.4423871293926501e-06,
+      "loss": 0.2411,
+      "step": 4653
+    },
+    {
+      "epoch": 1.5539232053422372,
+      "grad_norm": 0.47452189690874313,
+      "learning_rate": 1.440340356780367e-06,
+      "loss": 0.2566,
+      "step": 4654
+    },
+    {
+      "epoch": 1.5542570951585977,
+      "grad_norm": 0.47032083616066256,
+      "learning_rate": 1.4382947930100944e-06,
+      "loss": 0.2522,
+      "step": 4655
+    },
+    {
+      "epoch": 1.5545909849749582,
+      "grad_norm": 0.48161552930751034,
+      "learning_rate": 1.4362504387764947e-06,
+      "loss": 0.248,
+      "step": 4656
+    },
+    {
+      "epoch": 1.554924874791319,
+      "grad_norm": 0.4596439013861462,
+      "learning_rate": 1.4342072947738183e-06,
+      "loss": 0.2468,
+      "step": 4657
+    },
+    {
+      "epoch": 1.5552587646076794,
+      "grad_norm": 0.4572202402436325,
+      "learning_rate": 1.4321653616959096e-06,
+      "loss": 0.2421,
+      "step": 4658
+    },
+    {
+      "epoch": 1.5555926544240402,
+      "grad_norm": 0.46607886154222544,
+      "learning_rate": 1.430124640236198e-06,
+      "loss": 0.2576,
+      "step": 4659
+    },
+    {
+      "epoch": 1.5559265442404007,
+      "grad_norm": 0.4742950796232806,
+      "learning_rate": 1.4280851310876998e-06,
+      "loss": 0.2553,
+      "step": 4660
+    },
+    {
+      "epoch": 1.5562604340567612,
+      "grad_norm": 0.4845284543011468,
+      "learning_rate": 1.4260468349430224e-06,
+      "loss": 0.2519,
+      "step": 4661
+    },
+    {
+      "epoch": 1.556594323873122,
+      "grad_norm": 0.45870103589656336,
+      "learning_rate": 1.4240097524943602e-06,
+      "loss": 0.2355,
+      "step": 4662
+    },
+    {
+      "epoch": 1.5569282136894824,
+      "grad_norm": 0.47276788769117223,
+      "learning_rate": 1.421973884433499e-06,
+      "loss": 0.2488,
+      "step": 4663
+    },
+    {
+      "epoch": 1.5572621035058432,
+      "grad_norm": 0.4733025982058025,
+      "learning_rate": 1.4199392314518068e-06,
+      "loss": 0.2513,
+      "step": 4664
+    },
+    {
+      "epoch": 1.5575959933222037,
+      "grad_norm": 0.4644281225539227,
+      "learning_rate": 1.417905794240242e-06,
+      "loss": 0.2423,
+      "step": 4665
+    },
+    {
+      "epoch": 1.5579298831385642,
+      "grad_norm": 0.4486871454797364,
+      "learning_rate": 1.4158735734893464e-06,
+      "loss": 0.239,
+      "step": 4666
+    },
+    {
+      "epoch": 1.558263772954925,
+      "grad_norm": 0.4557365529712523,
+      "learning_rate": 1.4138425698892555e-06,
+      "loss": 0.2486,
+      "step": 4667
+    },
+    {
+      "epoch": 1.5585976627712856,
+      "grad_norm": 0.4795888420312492,
+      "learning_rate": 1.4118127841296864e-06,
+      "loss": 0.2518,
+      "step": 4668
+    },
+    {
+      "epoch": 1.5589315525876462,
+      "grad_norm": 0.4527185428976977,
+      "learning_rate": 1.4097842168999437e-06,
+      "loss": 0.2461,
+      "step": 4669
+    },
+    {
+      "epoch": 1.5592654424040067,
+      "grad_norm": 0.47619091288879095,
+      "learning_rate": 1.4077568688889165e-06,
+      "loss": 0.2527,
+      "step": 4670
+    },
+    {
+      "epoch": 1.5595993322203672,
+      "grad_norm": 0.4794784468572937,
+      "learning_rate": 1.4057307407850834e-06,
+      "loss": 0.2541,
+      "step": 4671
+    },
+    {
+      "epoch": 1.559933222036728,
+      "grad_norm": 0.46639508201589297,
+      "learning_rate": 1.4037058332765085e-06,
+      "loss": 0.2454,
+      "step": 4672
+    },
+    {
+      "epoch": 1.5602671118530886,
+      "grad_norm": 0.44873205797932947,
+      "learning_rate": 1.4016821470508385e-06,
+      "loss": 0.2349,
+      "step": 4673
+    },
+    {
+      "epoch": 1.5606010016694492,
+      "grad_norm": 0.4659895140760718,
+      "learning_rate": 1.3996596827953063e-06,
+      "loss": 0.2395,
+      "step": 4674
+    },
+    {
+      "epoch": 1.5609348914858097,
+      "grad_norm": 0.478105003726061,
+      "learning_rate": 1.3976384411967297e-06,
+      "loss": 0.2505,
+      "step": 4675
+    },
+    {
+      "epoch": 1.5612687813021702,
+      "grad_norm": 0.47191785243528966,
+      "learning_rate": 1.3956184229415148e-06,
+      "loss": 0.2485,
+      "step": 4676
+    },
+    {
+      "epoch": 1.561602671118531,
+      "grad_norm": 0.47233139371708144,
+      "learning_rate": 1.3935996287156467e-06,
+      "loss": 0.2436,
+      "step": 4677
+    },
+    {
+      "epoch": 1.5619365609348916,
+      "grad_norm": 0.4731141044305972,
+      "learning_rate": 1.3915820592046986e-06,
+      "loss": 0.2397,
+      "step": 4678
+    },
+    {
+      "epoch": 1.5622704507512521,
+      "grad_norm": 0.46789580898386285,
+      "learning_rate": 1.3895657150938252e-06,
+      "loss": 0.2547,
+      "step": 4679
+    },
+    {
+      "epoch": 1.5626043405676127,
+      "grad_norm": 0.46980531749499327,
+      "learning_rate": 1.3875505970677672e-06,
+      "loss": 0.2495,
+      "step": 4680
+    },
+    {
+      "epoch": 1.5629382303839732,
+      "grad_norm": 0.48380709575091446,
+      "learning_rate": 1.3855367058108504e-06,
+      "loss": 0.2603,
+      "step": 4681
+    },
+    {
+      "epoch": 1.563272120200334,
+      "grad_norm": 0.49250749599132204,
+      "learning_rate": 1.3835240420069801e-06,
+      "loss": 0.2625,
+      "step": 4682
+    },
+    {
+      "epoch": 1.5636060100166946,
+      "grad_norm": 0.4713604141033455,
+      "learning_rate": 1.3815126063396467e-06,
+      "loss": 0.2468,
+      "step": 4683
+    },
+    {
+      "epoch": 1.5639398998330551,
+      "grad_norm": 0.4715102035347803,
+      "learning_rate": 1.379502399491922e-06,
+      "loss": 0.2417,
+      "step": 4684
+    },
+    {
+      "epoch": 1.5642737896494157,
+      "grad_norm": 0.4671046529057528,
+      "learning_rate": 1.3774934221464643e-06,
+      "loss": 0.2457,
+      "step": 4685
+    },
+    {
+      "epoch": 1.5646076794657762,
+      "grad_norm": 0.4635112010412601,
+      "learning_rate": 1.3754856749855105e-06,
+      "loss": 0.2532,
+      "step": 4686
+    },
+    {
+      "epoch": 1.564941569282137,
+      "grad_norm": 0.4639404671426744,
+      "learning_rate": 1.3734791586908808e-06,
+      "loss": 0.2395,
+      "step": 4687
+    },
+    {
+      "epoch": 1.5652754590984976,
+      "grad_norm": 0.46220236526148195,
+      "learning_rate": 1.3714738739439792e-06,
+      "loss": 0.2461,
+      "step": 4688
+    },
+    {
+      "epoch": 1.5656093489148581,
+      "grad_norm": 0.4560437061947862,
+      "learning_rate": 1.3694698214257879e-06,
+      "loss": 0.2483,
+      "step": 4689
+    },
+    {
+      "epoch": 1.5659432387312187,
+      "grad_norm": 0.5003675601065328,
+      "learning_rate": 1.367467001816875e-06,
+      "loss": 0.2497,
+      "step": 4690
+    },
+    {
+      "epoch": 1.5662771285475792,
+      "grad_norm": 0.4742627262647329,
+      "learning_rate": 1.3654654157973874e-06,
+      "loss": 0.2492,
+      "step": 4691
+    },
+    {
+      "epoch": 1.56661101836394,
+      "grad_norm": 0.479052123339044,
+      "learning_rate": 1.3634650640470525e-06,
+      "loss": 0.2477,
+      "step": 4692
+    },
+    {
+      "epoch": 1.5669449081803006,
+      "grad_norm": 0.48441335986174766,
+      "learning_rate": 1.361465947245178e-06,
+      "loss": 0.2523,
+      "step": 4693
+    },
+    {
+      "epoch": 1.5672787979966611,
+      "grad_norm": 0.4816841460309258,
+      "learning_rate": 1.3594680660706572e-06,
+      "loss": 0.2454,
+      "step": 4694
+    },
+    {
+      "epoch": 1.5676126878130217,
+      "grad_norm": 0.4843452213925038,
+      "learning_rate": 1.357471421201958e-06,
+      "loss": 0.2454,
+      "step": 4695
+    },
+    {
+      "epoch": 1.5679465776293822,
+      "grad_norm": 0.473177161535402,
+      "learning_rate": 1.355476013317129e-06,
+      "loss": 0.2616,
+      "step": 4696
+    },
+    {
+      "epoch": 1.568280467445743,
+      "grad_norm": 0.46898159185967003,
+      "learning_rate": 1.3534818430938046e-06,
+      "loss": 0.254,
+      "step": 4697
+    },
+    {
+      "epoch": 1.5686143572621036,
+      "grad_norm": 0.47181116474818313,
+      "learning_rate": 1.3514889112091905e-06,
+      "loss": 0.2408,
+      "step": 4698
+    },
+    {
+      "epoch": 1.5689482470784641,
+      "grad_norm": 0.4597013250300328,
+      "learning_rate": 1.3494972183400795e-06,
+      "loss": 0.2503,
+      "step": 4699
+    },
+    {
+      "epoch": 1.5692821368948247,
+      "grad_norm": 0.4631714632287056,
+      "learning_rate": 1.3475067651628388e-06,
+      "loss": 0.2509,
+      "step": 4700
+    },
+    {
+      "epoch": 1.5696160267111852,
+      "grad_norm": 0.4752918340537046,
+      "learning_rate": 1.3455175523534158e-06,
+      "loss": 0.2505,
+      "step": 4701
+    },
+    {
+      "epoch": 1.569949916527546,
+      "grad_norm": 0.4688679926456441,
+      "learning_rate": 1.3435295805873361e-06,
+      "loss": 0.2538,
+      "step": 4702
+    },
+    {
+      "epoch": 1.5702838063439066,
+      "grad_norm": 0.4481238378396957,
+      "learning_rate": 1.341542850539706e-06,
+      "loss": 0.2467,
+      "step": 4703
+    },
+    {
+      "epoch": 1.5706176961602671,
+      "grad_norm": 0.45628677270890766,
+      "learning_rate": 1.3395573628852076e-06,
+      "loss": 0.2398,
+      "step": 4704
+    },
+    {
+      "epoch": 1.5709515859766277,
+      "grad_norm": 0.47973249694193043,
+      "learning_rate": 1.337573118298104e-06,
+      "loss": 0.2494,
+      "step": 4705
+    },
+    {
+      "epoch": 1.5712854757929882,
+      "grad_norm": 0.4675715818068297,
+      "learning_rate": 1.3355901174522334e-06,
+      "loss": 0.2466,
+      "step": 4706
+    },
+    {
+      "epoch": 1.571619365609349,
+      "grad_norm": 0.47316946222759826,
+      "learning_rate": 1.3336083610210115e-06,
+      "loss": 0.2503,
+      "step": 4707
+    },
+    {
+      "epoch": 1.5719532554257096,
+      "grad_norm": 0.4800933975233366,
+      "learning_rate": 1.331627849677435e-06,
+      "loss": 0.248,
+      "step": 4708
+    },
+    {
+      "epoch": 1.5722871452420701,
+      "grad_norm": 0.4512049415152314,
+      "learning_rate": 1.3296485840940732e-06,
+      "loss": 0.2414,
+      "step": 4709
+    },
+    {
+      "epoch": 1.5726210350584306,
+      "grad_norm": 0.4602193275373224,
+      "learning_rate": 1.3276705649430754e-06,
+      "loss": 0.2508,
+      "step": 4710
+    },
+    {
+      "epoch": 1.5729549248747912,
+      "grad_norm": 0.4694476487329797,
+      "learning_rate": 1.3256937928961655e-06,
+      "loss": 0.2562,
+      "step": 4711
+    },
+    {
+      "epoch": 1.573288814691152,
+      "grad_norm": 0.4579912133447704,
+      "learning_rate": 1.323718268624647e-06,
+      "loss": 0.2535,
+      "step": 4712
+    },
+    {
+      "epoch": 1.5736227045075126,
+      "grad_norm": 0.4788537424961001,
+      "learning_rate": 1.3217439927993957e-06,
+      "loss": 0.2517,
+      "step": 4713
+    },
+    {
+      "epoch": 1.5739565943238731,
+      "grad_norm": 0.4727181010890767,
+      "learning_rate": 1.3197709660908676e-06,
+      "loss": 0.2441,
+      "step": 4714
+    },
+    {
+      "epoch": 1.5742904841402336,
+      "grad_norm": 0.4681402195710115,
+      "learning_rate": 1.3177991891690917e-06,
+      "loss": 0.2497,
+      "step": 4715
+    },
+    {
+      "epoch": 1.5746243739565942,
+      "grad_norm": 0.4562515386570161,
+      "learning_rate": 1.3158286627036709e-06,
+      "loss": 0.2447,
+      "step": 4716
+    },
+    {
+      "epoch": 1.574958263772955,
+      "grad_norm": 0.4655474986559904,
+      "learning_rate": 1.3138593873637894e-06,
+      "loss": 0.2401,
+      "step": 4717
+    },
+    {
+      "epoch": 1.5752921535893156,
+      "grad_norm": 0.472567241751802,
+      "learning_rate": 1.3118913638182012e-06,
+      "loss": 0.2496,
+      "step": 4718
+    },
+    {
+      "epoch": 1.5756260434056761,
+      "grad_norm": 0.4601480452795277,
+      "learning_rate": 1.3099245927352367e-06,
+      "loss": 0.2377,
+      "step": 4719
+    },
+    {
+      "epoch": 1.5759599332220366,
+      "grad_norm": 0.45979832756842337,
+      "learning_rate": 1.3079590747828003e-06,
+      "loss": 0.2456,
+      "step": 4720
+    },
+    {
+      "epoch": 1.5762938230383974,
+      "grad_norm": 0.45367809649838997,
+      "learning_rate": 1.3059948106283725e-06,
+      "loss": 0.2528,
+      "step": 4721
+    },
+    {
+      "epoch": 1.576627712854758,
+      "grad_norm": 0.48184150762713757,
+      "learning_rate": 1.3040318009390096e-06,
+      "loss": 0.2537,
+      "step": 4722
+    },
+    {
+      "epoch": 1.5769616026711186,
+      "grad_norm": 0.4614569044909795,
+      "learning_rate": 1.3020700463813374e-06,
+      "loss": 0.2544,
+      "step": 4723
+    },
+    {
+      "epoch": 1.5772954924874791,
+      "grad_norm": 0.4864422163736675,
+      "learning_rate": 1.3001095476215581e-06,
+      "loss": 0.252,
+      "step": 4724
+    },
+    {
+      "epoch": 1.5776293823038396,
+      "grad_norm": 0.48543608979036124,
+      "learning_rate": 1.2981503053254446e-06,
+      "loss": 0.2514,
+      "step": 4725
+    },
+    {
+      "epoch": 1.5779632721202004,
+      "grad_norm": 0.4748224717962334,
+      "learning_rate": 1.2961923201583493e-06,
+      "loss": 0.2443,
+      "step": 4726
+    },
+    {
+      "epoch": 1.5782971619365609,
+      "grad_norm": 0.4515181318023877,
+      "learning_rate": 1.2942355927851919e-06,
+      "loss": 0.2348,
+      "step": 4727
+    },
+    {
+      "epoch": 1.5786310517529216,
+      "grad_norm": 0.46206479276833834,
+      "learning_rate": 1.2922801238704675e-06,
+      "loss": 0.2497,
+      "step": 4728
+    },
+    {
+      "epoch": 1.5789649415692821,
+      "grad_norm": 0.478093121562446,
+      "learning_rate": 1.2903259140782404e-06,
+      "loss": 0.2406,
+      "step": 4729
+    },
+    {
+      "epoch": 1.5792988313856426,
+      "grad_norm": 0.4724976315270765,
+      "learning_rate": 1.2883729640721532e-06,
+      "loss": 0.248,
+      "step": 4730
+    },
+    {
+      "epoch": 1.5796327212020034,
+      "grad_norm": 0.4658997260768487,
+      "learning_rate": 1.286421274515418e-06,
+      "loss": 0.247,
+      "step": 4731
+    },
+    {
+      "epoch": 1.5799666110183639,
+      "grad_norm": 0.46082236574727786,
+      "learning_rate": 1.2844708460708177e-06,
+      "loss": 0.2467,
+      "step": 4732
+    },
+    {
+      "epoch": 1.5803005008347246,
+      "grad_norm": 0.48118028654450706,
+      "learning_rate": 1.2825216794007073e-06,
+      "loss": 0.2581,
+      "step": 4733
+    },
+    {
+      "epoch": 1.5806343906510851,
+      "grad_norm": 0.47036895037116006,
+      "learning_rate": 1.2805737751670122e-06,
+      "loss": 0.2578,
+      "step": 4734
+    },
+    {
+      "epoch": 1.5809682804674456,
+      "grad_norm": 0.47243185664774556,
+      "learning_rate": 1.278627134031234e-06,
+      "loss": 0.2466,
+      "step": 4735
+    },
+    {
+      "epoch": 1.5813021702838064,
+      "grad_norm": 0.4823450977372792,
+      "learning_rate": 1.2766817566544398e-06,
+      "loss": 0.2511,
+      "step": 4736
+    },
+    {
+      "epoch": 1.581636060100167,
+      "grad_norm": 0.44840671733449655,
+      "learning_rate": 1.2747376436972702e-06,
+      "loss": 0.2408,
+      "step": 4737
+    },
+    {
+      "epoch": 1.5819699499165276,
+      "grad_norm": 0.46942568997093514,
+      "learning_rate": 1.2727947958199349e-06,
+      "loss": 0.2463,
+      "step": 4738
+    },
+    {
+      "epoch": 1.5823038397328881,
+      "grad_norm": 0.4786182723187881,
+      "learning_rate": 1.2708532136822156e-06,
+      "loss": 0.2553,
+      "step": 4739
+    },
+    {
+      "epoch": 1.5826377295492486,
+      "grad_norm": 0.4450020393471056,
+      "learning_rate": 1.268912897943465e-06,
+      "loss": 0.2303,
+      "step": 4740
+    },
+    {
+      "epoch": 1.5829716193656094,
+      "grad_norm": 0.4700408266651774,
+      "learning_rate": 1.2669738492626033e-06,
+      "loss": 0.2543,
+      "step": 4741
+    },
+    {
+      "epoch": 1.58330550918197,
+      "grad_norm": 0.47327144109070557,
+      "learning_rate": 1.2650360682981216e-06,
+      "loss": 0.2478,
+      "step": 4742
+    },
+    {
+      "epoch": 1.5836393989983306,
+      "grad_norm": 0.4832961325114361,
+      "learning_rate": 1.2630995557080783e-06,
+      "loss": 0.254,
+      "step": 4743
+    },
+    {
+      "epoch": 1.5839732888146911,
+      "grad_norm": 0.4888465324313995,
+      "learning_rate": 1.2611643121501061e-06,
+      "loss": 0.2538,
+      "step": 4744
+    },
+    {
+      "epoch": 1.5843071786310516,
+      "grad_norm": 0.4788066814435775,
+      "learning_rate": 1.2592303382814024e-06,
+      "loss": 0.2501,
+      "step": 4745
+    },
+    {
+      "epoch": 1.5846410684474124,
+      "grad_norm": 0.48581577165344464,
+      "learning_rate": 1.2572976347587345e-06,
+      "loss": 0.2563,
+      "step": 4746
+    },
+    {
+      "epoch": 1.584974958263773,
+      "grad_norm": 0.4716785535564782,
+      "learning_rate": 1.2553662022384377e-06,
+      "loss": 0.2505,
+      "step": 4747
+    },
+    {
+      "epoch": 1.5853088480801336,
+      "grad_norm": 0.4817622888357497,
+      "learning_rate": 1.253436041376417e-06,
+      "loss": 0.2463,
+      "step": 4748
+    },
+    {
+      "epoch": 1.5856427378964941,
+      "grad_norm": 0.472940970079289,
+      "learning_rate": 1.2515071528281474e-06,
+      "loss": 0.2449,
+      "step": 4749
+    },
+    {
+      "epoch": 1.5859766277128546,
+      "grad_norm": 0.45292118855456276,
+      "learning_rate": 1.2495795372486673e-06,
+      "loss": 0.2508,
+      "step": 4750
+    },
+    {
+      "epoch": 1.5863105175292154,
+      "grad_norm": 0.4757766606791781,
+      "learning_rate": 1.2476531952925852e-06,
+      "loss": 0.2561,
+      "step": 4751
+    },
+    {
+      "epoch": 1.586644407345576,
+      "grad_norm": 0.460518100413153,
+      "learning_rate": 1.2457281276140759e-06,
+      "loss": 0.2459,
+      "step": 4752
+    },
+    {
+      "epoch": 1.5869782971619366,
+      "grad_norm": 0.47734012005070137,
+      "learning_rate": 1.243804334866885e-06,
+      "loss": 0.2445,
+      "step": 4753
+    },
+    {
+      "epoch": 1.5873121869782971,
+      "grad_norm": 0.4677268026926354,
+      "learning_rate": 1.2418818177043207e-06,
+      "loss": 0.2544,
+      "step": 4754
+    },
+    {
+      "epoch": 1.5876460767946576,
+      "grad_norm": 0.4605845669899324,
+      "learning_rate": 1.2399605767792588e-06,
+      "loss": 0.2432,
+      "step": 4755
+    },
+    {
+      "epoch": 1.5879799666110184,
+      "grad_norm": 0.5036158204788531,
+      "learning_rate": 1.2380406127441457e-06,
+      "loss": 0.2655,
+      "step": 4756
+    },
+    {
+      "epoch": 1.588313856427379,
+      "grad_norm": 0.4686324078702875,
+      "learning_rate": 1.2361219262509882e-06,
+      "loss": 0.2399,
+      "step": 4757
+    },
+    {
+      "epoch": 1.5886477462437396,
+      "grad_norm": 0.4596837553085632,
+      "learning_rate": 1.2342045179513655e-06,
+      "loss": 0.2542,
+      "step": 4758
+    },
+    {
+      "epoch": 1.5889816360601001,
+      "grad_norm": 0.49890305511387867,
+      "learning_rate": 1.2322883884964181e-06,
+      "loss": 0.2531,
+      "step": 4759
+    },
+    {
+      "epoch": 1.5893155258764606,
+      "grad_norm": 0.46775117620872,
+      "learning_rate": 1.2303735385368538e-06,
+      "loss": 0.2433,
+      "step": 4760
+    },
+    {
+      "epoch": 1.5896494156928214,
+      "grad_norm": 0.47041775012314846,
+      "learning_rate": 1.228459968722943e-06,
+      "loss": 0.2536,
+      "step": 4761
+    },
+    {
+      "epoch": 1.589983305509182,
+      "grad_norm": 0.4926800703974411,
+      "learning_rate": 1.2265476797045288e-06,
+      "loss": 0.2537,
+      "step": 4762
+    },
+    {
+      "epoch": 1.5903171953255426,
+      "grad_norm": 0.47625927038631766,
+      "learning_rate": 1.2246366721310126e-06,
+      "loss": 0.2447,
+      "step": 4763
+    },
+    {
+      "epoch": 1.5906510851419031,
+      "grad_norm": 0.4809953429828662,
+      "learning_rate": 1.2227269466513608e-06,
+      "loss": 0.2532,
+      "step": 4764
+    },
+    {
+      "epoch": 1.5909849749582636,
+      "grad_norm": 0.48689479141553155,
+      "learning_rate": 1.2208185039141096e-06,
+      "loss": 0.247,
+      "step": 4765
+    },
+    {
+      "epoch": 1.5913188647746244,
+      "grad_norm": 0.479666948659014,
+      "learning_rate": 1.2189113445673528e-06,
+      "loss": 0.2503,
+      "step": 4766
+    },
+    {
+      "epoch": 1.591652754590985,
+      "grad_norm": 0.4509046493667952,
+      "learning_rate": 1.2170054692587556e-06,
+      "loss": 0.2341,
+      "step": 4767
+    },
+    {
+      "epoch": 1.5919866444073456,
+      "grad_norm": 0.47972299735595814,
+      "learning_rate": 1.2151008786355411e-06,
+      "loss": 0.258,
+      "step": 4768
+    },
+    {
+      "epoch": 1.5923205342237061,
+      "grad_norm": 0.468723117894756,
+      "learning_rate": 1.2131975733444996e-06,
+      "loss": 0.2415,
+      "step": 4769
+    },
+    {
+      "epoch": 1.5926544240400666,
+      "grad_norm": 0.4812693363279319,
+      "learning_rate": 1.211295554031981e-06,
+      "loss": 0.2489,
+      "step": 4770
+    },
+    {
+      "epoch": 1.5929883138564274,
+      "grad_norm": 0.46256588232614143,
+      "learning_rate": 1.2093948213439049e-06,
+      "loss": 0.2533,
+      "step": 4771
+    },
+    {
+      "epoch": 1.593322203672788,
+      "grad_norm": 0.4604374920888097,
+      "learning_rate": 1.2074953759257474e-06,
+      "loss": 0.2408,
+      "step": 4772
+    },
+    {
+      "epoch": 1.5936560934891486,
+      "grad_norm": 0.45863182917045914,
+      "learning_rate": 1.2055972184225528e-06,
+      "loss": 0.2496,
+      "step": 4773
+    },
+    {
+      "epoch": 1.5939899833055091,
+      "grad_norm": 0.469363612104249,
+      "learning_rate": 1.2037003494789246e-06,
+      "loss": 0.2442,
+      "step": 4774
+    },
+    {
+      "epoch": 1.5943238731218696,
+      "grad_norm": 0.4814967452052739,
+      "learning_rate": 1.201804769739028e-06,
+      "loss": 0.2536,
+      "step": 4775
+    },
+    {
+      "epoch": 1.5946577629382304,
+      "grad_norm": 0.49437214785803396,
+      "learning_rate": 1.1999104798465943e-06,
+      "loss": 0.2471,
+      "step": 4776
+    },
+    {
+      "epoch": 1.594991652754591,
+      "grad_norm": 0.47196875911436925,
+      "learning_rate": 1.1980174804449145e-06,
+      "loss": 0.2517,
+      "step": 4777
+    },
+    {
+      "epoch": 1.5953255425709516,
+      "grad_norm": 0.4695740545271054,
+      "learning_rate": 1.1961257721768405e-06,
+      "loss": 0.2429,
+      "step": 4778
+    },
+    {
+      "epoch": 1.5956594323873121,
+      "grad_norm": 0.47167436411157904,
+      "learning_rate": 1.1942353556847852e-06,
+      "loss": 0.2562,
+      "step": 4779
+    },
+    {
+      "epoch": 1.5959933222036726,
+      "grad_norm": 0.49551127966533964,
+      "learning_rate": 1.1923462316107276e-06,
+      "loss": 0.2473,
+      "step": 4780
+    },
+    {
+      "epoch": 1.5963272120200334,
+      "grad_norm": 0.48369934287237415,
+      "learning_rate": 1.190458400596201e-06,
+      "loss": 0.2524,
+      "step": 4781
+    },
+    {
+      "epoch": 1.596661101836394,
+      "grad_norm": 0.4645281480601688,
+      "learning_rate": 1.1885718632823068e-06,
+      "loss": 0.2521,
+      "step": 4782
+    },
+    {
+      "epoch": 1.5969949916527546,
+      "grad_norm": 0.46622505501356226,
+      "learning_rate": 1.186686620309701e-06,
+      "loss": 0.2527,
+      "step": 4783
+    },
+    {
+      "epoch": 1.5973288814691151,
+      "grad_norm": 0.4598780614227544,
+      "learning_rate": 1.1848026723186013e-06,
+      "loss": 0.2358,
+      "step": 4784
+    },
+    {
+      "epoch": 1.5976627712854758,
+      "grad_norm": 0.4591289738204158,
+      "learning_rate": 1.1829200199487894e-06,
+      "loss": 0.2474,
+      "step": 4785
+    },
+    {
+      "epoch": 1.5979966611018364,
+      "grad_norm": 0.47723317240763746,
+      "learning_rate": 1.1810386638396032e-06,
+      "loss": 0.2475,
+      "step": 4786
+    },
+    {
+      "epoch": 1.598330550918197,
+      "grad_norm": 0.46170584039443174,
+      "learning_rate": 1.1791586046299414e-06,
+      "loss": 0.2495,
+      "step": 4787
+    },
+    {
+      "epoch": 1.5986644407345576,
+      "grad_norm": 0.4682458614945937,
+      "learning_rate": 1.177279842958261e-06,
+      "loss": 0.2472,
+      "step": 4788
+    },
+    {
+      "epoch": 1.5989983305509181,
+      "grad_norm": 0.4812354321540043,
+      "learning_rate": 1.1754023794625806e-06,
+      "loss": 0.2464,
+      "step": 4789
+    },
+    {
+      "epoch": 1.5993322203672788,
+      "grad_norm": 0.47107537437268443,
+      "learning_rate": 1.1735262147804793e-06,
+      "loss": 0.2511,
+      "step": 4790
+    },
+    {
+      "epoch": 1.5996661101836394,
+      "grad_norm": 0.4666719791954086,
+      "learning_rate": 1.1716513495490912e-06,
+      "loss": 0.259,
+      "step": 4791
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.47402448962886046,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.2606,
+      "step": 4792
+    },
+    {
+      "epoch": 1.6003338898163606,
+      "grad_norm": 0.4729935204667167,
+      "learning_rate": 1.1679055199847894e-06,
+      "loss": 0.2473,
+      "step": 4793
+    },
+    {
+      "epoch": 1.600667779632721,
+      "grad_norm": 0.49066027772209836,
+      "learning_rate": 1.1660345569239418e-06,
+      "loss": 0.2487,
+      "step": 4794
+    },
+    {
+      "epoch": 1.6010016694490818,
+      "grad_norm": 0.4714925686024002,
+      "learning_rate": 1.1641648958579354e-06,
+      "loss": 0.2445,
+      "step": 4795
+    },
+    {
+      "epoch": 1.6013355592654424,
+      "grad_norm": 0.47424682037219845,
+      "learning_rate": 1.1622965374216976e-06,
+      "loss": 0.2499,
+      "step": 4796
+    },
+    {
+      "epoch": 1.601669449081803,
+      "grad_norm": 0.4815135281558572,
+      "learning_rate": 1.1604294822497125e-06,
+      "loss": 0.2588,
+      "step": 4797
+    },
+    {
+      "epoch": 1.6020033388981636,
+      "grad_norm": 0.4882167133364399,
+      "learning_rate": 1.1585637309760229e-06,
+      "loss": 0.2564,
+      "step": 4798
+    },
+    {
+      "epoch": 1.602337228714524,
+      "grad_norm": 0.46231959401872735,
+      "learning_rate": 1.15669928423423e-06,
+      "loss": 0.2467,
+      "step": 4799
+    },
+    {
+      "epoch": 1.6026711185308848,
+      "grad_norm": 0.4677233835985707,
+      "learning_rate": 1.154836142657489e-06,
+      "loss": 0.2421,
+      "step": 4800
+    },
+    {
+      "epoch": 1.6030050083472456,
+      "grad_norm": 0.5017341441361839,
+      "learning_rate": 1.1529743068785131e-06,
+      "loss": 0.2452,
+      "step": 4801
+    },
+    {
+      "epoch": 1.603338898163606,
+      "grad_norm": 0.4546493400321069,
+      "learning_rate": 1.1511137775295705e-06,
+      "loss": 0.2377,
+      "step": 4802
+    },
+    {
+      "epoch": 1.6036727879799666,
+      "grad_norm": 0.47544052330410014,
+      "learning_rate": 1.14925455524249e-06,
+      "loss": 0.2443,
+      "step": 4803
+    },
+    {
+      "epoch": 1.604006677796327,
+      "grad_norm": 0.4693267864572999,
+      "learning_rate": 1.147396640648653e-06,
+      "loss": 0.2468,
+      "step": 4804
+    },
+    {
+      "epoch": 1.6043405676126878,
+      "grad_norm": 0.4859955961708985,
+      "learning_rate": 1.145540034378998e-06,
+      "loss": 0.2488,
+      "step": 4805
+    },
+    {
+      "epoch": 1.6046744574290486,
+      "grad_norm": 0.4631655606892396,
+      "learning_rate": 1.1436847370640164e-06,
+      "loss": 0.2441,
+      "step": 4806
+    },
+    {
+      "epoch": 1.605008347245409,
+      "grad_norm": 0.4666245447842212,
+      "learning_rate": 1.1418307493337594e-06,
+      "loss": 0.2539,
+      "step": 4807
+    },
+    {
+      "epoch": 1.6053422370617696,
+      "grad_norm": 0.4717101212139891,
+      "learning_rate": 1.1399780718178332e-06,
+      "loss": 0.2582,
+      "step": 4808
+    },
+    {
+      "epoch": 1.60567612687813,
+      "grad_norm": 0.4655047665254601,
+      "learning_rate": 1.1381267051453965e-06,
+      "loss": 0.2476,
+      "step": 4809
+    },
+    {
+      "epoch": 1.6060100166944908,
+      "grad_norm": 0.45535071599351706,
+      "learning_rate": 1.1362766499451633e-06,
+      "loss": 0.24,
+      "step": 4810
+    },
+    {
+      "epoch": 1.6063439065108516,
+      "grad_norm": 0.47710526147923055,
+      "learning_rate": 1.1344279068454012e-06,
+      "loss": 0.2608,
+      "step": 4811
+    },
+    {
+      "epoch": 1.606677796327212,
+      "grad_norm": 0.45065089182456636,
+      "learning_rate": 1.1325804764739372e-06,
+      "loss": 0.2522,
+      "step": 4812
+    },
+    {
+      "epoch": 1.6070116861435726,
+      "grad_norm": 0.4662461209542175,
+      "learning_rate": 1.130734359458147e-06,
+      "loss": 0.2508,
+      "step": 4813
+    },
+    {
+      "epoch": 1.607345575959933,
+      "grad_norm": 0.4523875306842065,
+      "learning_rate": 1.128889556424963e-06,
+      "loss": 0.2421,
+      "step": 4814
+    },
+    {
+      "epoch": 1.6076794657762938,
+      "grad_norm": 0.47955967957909534,
+      "learning_rate": 1.1270460680008693e-06,
+      "loss": 0.258,
+      "step": 4815
+    },
+    {
+      "epoch": 1.6080133555926546,
+      "grad_norm": 0.4797578134466955,
+      "learning_rate": 1.1252038948119065e-06,
+      "loss": 0.2498,
+      "step": 4816
+    },
+    {
+      "epoch": 1.608347245409015,
+      "grad_norm": 0.46110405774007074,
+      "learning_rate": 1.1233630374836679e-06,
+      "loss": 0.2489,
+      "step": 4817
+    },
+    {
+      "epoch": 1.6086811352253756,
+      "grad_norm": 0.44794884888982217,
+      "learning_rate": 1.121523496641298e-06,
+      "loss": 0.2386,
+      "step": 4818
+    },
+    {
+      "epoch": 1.609015025041736,
+      "grad_norm": 0.4868358044755078,
+      "learning_rate": 1.119685272909496e-06,
+      "loss": 0.2485,
+      "step": 4819
+    },
+    {
+      "epoch": 1.6093489148580968,
+      "grad_norm": 0.4887041669401457,
+      "learning_rate": 1.1178483669125112e-06,
+      "loss": 0.2528,
+      "step": 4820
+    },
+    {
+      "epoch": 1.6096828046744576,
+      "grad_norm": 0.4731014457829856,
+      "learning_rate": 1.11601277927415e-06,
+      "loss": 0.2562,
+      "step": 4821
+    },
+    {
+      "epoch": 1.610016694490818,
+      "grad_norm": 0.4821658560522468,
+      "learning_rate": 1.1141785106177672e-06,
+      "loss": 0.2533,
+      "step": 4822
+    },
+    {
+      "epoch": 1.6103505843071786,
+      "grad_norm": 0.4751094374554469,
+      "learning_rate": 1.1123455615662698e-06,
+      "loss": 0.2419,
+      "step": 4823
+    },
+    {
+      "epoch": 1.610684474123539,
+      "grad_norm": 0.47961037870027545,
+      "learning_rate": 1.1105139327421204e-06,
+      "loss": 0.2494,
+      "step": 4824
+    },
+    {
+      "epoch": 1.6110183639398998,
+      "grad_norm": 0.46967468837746207,
+      "learning_rate": 1.1086836247673282e-06,
+      "loss": 0.2517,
+      "step": 4825
+    },
+    {
+      "epoch": 1.6113522537562606,
+      "grad_norm": 0.4650794965559954,
+      "learning_rate": 1.1068546382634587e-06,
+      "loss": 0.2345,
+      "step": 4826
+    },
+    {
+      "epoch": 1.611686143572621,
+      "grad_norm": 0.4794245368990495,
+      "learning_rate": 1.105026973851625e-06,
+      "loss": 0.2477,
+      "step": 4827
+    },
+    {
+      "epoch": 1.6120200333889816,
+      "grad_norm": 0.4653696044081265,
+      "learning_rate": 1.1032006321524929e-06,
+      "loss": 0.2341,
+      "step": 4828
+    },
+    {
+      "epoch": 1.612353923205342,
+      "grad_norm": 0.4544230914880064,
+      "learning_rate": 1.101375613786278e-06,
+      "loss": 0.2412,
+      "step": 4829
+    },
+    {
+      "epoch": 1.6126878130217028,
+      "grad_norm": 0.4760565529038112,
+      "learning_rate": 1.0995519193727462e-06,
+      "loss": 0.247,
+      "step": 4830
+    },
+    {
+      "epoch": 1.6130217028380636,
+      "grad_norm": 0.4679480087745479,
+      "learning_rate": 1.0977295495312167e-06,
+      "loss": 0.2423,
+      "step": 4831
+    },
+    {
+      "epoch": 1.613355592654424,
+      "grad_norm": 0.45589841218702304,
+      "learning_rate": 1.0959085048805546e-06,
+      "loss": 0.2463,
+      "step": 4832
+    },
+    {
+      "epoch": 1.6136894824707846,
+      "grad_norm": 0.4848952169869333,
+      "learning_rate": 1.0940887860391802e-06,
+      "loss": 0.2464,
+      "step": 4833
+    },
+    {
+      "epoch": 1.614023372287145,
+      "grad_norm": 0.4874193064521673,
+      "learning_rate": 1.0922703936250578e-06,
+      "loss": 0.2529,
+      "step": 4834
+    },
+    {
+      "epoch": 1.6143572621035058,
+      "grad_norm": 0.4941212748305379,
+      "learning_rate": 1.0904533282557066e-06,
+      "loss": 0.2605,
+      "step": 4835
+    },
+    {
+      "epoch": 1.6146911519198666,
+      "grad_norm": 0.4882609938046427,
+      "learning_rate": 1.0886375905481917e-06,
+      "loss": 0.2444,
+      "step": 4836
+    },
+    {
+      "epoch": 1.615025041736227,
+      "grad_norm": 0.463985970022649,
+      "learning_rate": 1.0868231811191276e-06,
+      "loss": 0.2508,
+      "step": 4837
+    },
+    {
+      "epoch": 1.6153589315525876,
+      "grad_norm": 0.465283928664218,
+      "learning_rate": 1.0850101005846787e-06,
+      "loss": 0.2345,
+      "step": 4838
+    },
+    {
+      "epoch": 1.615692821368948,
+      "grad_norm": 0.4733621767992804,
+      "learning_rate": 1.0831983495605557e-06,
+      "loss": 0.2457,
+      "step": 4839
+    },
+    {
+      "epoch": 1.6160267111853088,
+      "grad_norm": 0.4847517879589195,
+      "learning_rate": 1.0813879286620227e-06,
+      "loss": 0.2509,
+      "step": 4840
+    },
+    {
+      "epoch": 1.6163606010016696,
+      "grad_norm": 0.46482711553590367,
+      "learning_rate": 1.0795788385038892e-06,
+      "loss": 0.2415,
+      "step": 4841
+    },
+    {
+      "epoch": 1.61669449081803,
+      "grad_norm": 0.4770810248876076,
+      "learning_rate": 1.0777710797005124e-06,
+      "loss": 0.2556,
+      "step": 4842
+    },
+    {
+      "epoch": 1.6170283806343906,
+      "grad_norm": 0.4647749919520915,
+      "learning_rate": 1.075964652865798e-06,
+      "loss": 0.2534,
+      "step": 4843
+    },
+    {
+      "epoch": 1.617362270450751,
+      "grad_norm": 0.4701260000940848,
+      "learning_rate": 1.074159558613197e-06,
+      "loss": 0.2497,
+      "step": 4844
+    },
+    {
+      "epoch": 1.6176961602671118,
+      "grad_norm": 0.47043248130665666,
+      "learning_rate": 1.072355797555713e-06,
+      "loss": 0.2538,
+      "step": 4845
+    },
+    {
+      "epoch": 1.6180300500834726,
+      "grad_norm": 0.4779221660832495,
+      "learning_rate": 1.0705533703058928e-06,
+      "loss": 0.2465,
+      "step": 4846
+    },
+    {
+      "epoch": 1.618363939899833,
+      "grad_norm": 0.4989686476300649,
+      "learning_rate": 1.0687522774758318e-06,
+      "loss": 0.2476,
+      "step": 4847
+    },
+    {
+      "epoch": 1.6186978297161936,
+      "grad_norm": 0.46508483162678277,
+      "learning_rate": 1.0669525196771695e-06,
+      "loss": 0.2539,
+      "step": 4848
+    },
+    {
+      "epoch": 1.619031719532554,
+      "grad_norm": 0.46207287231371136,
+      "learning_rate": 1.0651540975210966e-06,
+      "loss": 0.2494,
+      "step": 4849
+    },
+    {
+      "epoch": 1.6193656093489148,
+      "grad_norm": 0.47288183337273815,
+      "learning_rate": 1.063357011618349e-06,
+      "loss": 0.253,
+      "step": 4850
+    },
+    {
+      "epoch": 1.6196994991652756,
+      "grad_norm": 0.49098451569446583,
+      "learning_rate": 1.061561262579206e-06,
+      "loss": 0.2502,
+      "step": 4851
+    },
+    {
+      "epoch": 1.620033388981636,
+      "grad_norm": 0.5043745556588018,
+      "learning_rate": 1.059766851013495e-06,
+      "loss": 0.2503,
+      "step": 4852
+    },
+    {
+      "epoch": 1.6203672787979966,
+      "grad_norm": 0.4569878354870167,
+      "learning_rate": 1.0579737775305881e-06,
+      "loss": 0.2489,
+      "step": 4853
+    },
+    {
+      "epoch": 1.6207011686143573,
+      "grad_norm": 0.4858659199347927,
+      "learning_rate": 1.0561820427394054e-06,
+      "loss": 0.258,
+      "step": 4854
+    },
+    {
+      "epoch": 1.6210350584307178,
+      "grad_norm": 0.46054673461907786,
+      "learning_rate": 1.0543916472484105e-06,
+      "loss": 0.2385,
+      "step": 4855
+    },
+    {
+      "epoch": 1.6213689482470786,
+      "grad_norm": 0.45881039570693455,
+      "learning_rate": 1.052602591665612e-06,
+      "loss": 0.2377,
+      "step": 4856
+    },
+    {
+      "epoch": 1.621702838063439,
+      "grad_norm": 0.4632574185864471,
+      "learning_rate": 1.0508148765985621e-06,
+      "loss": 0.2549,
+      "step": 4857
+    },
+    {
+      "epoch": 1.6220367278797996,
+      "grad_norm": 0.46750095058865987,
+      "learning_rate": 1.0490285026543612e-06,
+      "loss": 0.2424,
+      "step": 4858
+    },
+    {
+      "epoch": 1.6223706176961603,
+      "grad_norm": 0.4516967531282782,
+      "learning_rate": 1.0472434704396544e-06,
+      "loss": 0.2407,
+      "step": 4859
+    },
+    {
+      "epoch": 1.6227045075125208,
+      "grad_norm": 0.46272073596275387,
+      "learning_rate": 1.0454597805606282e-06,
+      "loss": 0.2492,
+      "step": 4860
+    },
+    {
+      "epoch": 1.6230383973288816,
+      "grad_norm": 0.4485989526139971,
+      "learning_rate": 1.0436774336230137e-06,
+      "loss": 0.2349,
+      "step": 4861
+    },
+    {
+      "epoch": 1.623372287145242,
+      "grad_norm": 0.47338018908328017,
+      "learning_rate": 1.0418964302320854e-06,
+      "loss": 0.2356,
+      "step": 4862
+    },
+    {
+      "epoch": 1.6237061769616026,
+      "grad_norm": 0.46240804724624185,
+      "learning_rate": 1.0401167709926663e-06,
+      "loss": 0.2558,
+      "step": 4863
+    },
+    {
+      "epoch": 1.6240400667779633,
+      "grad_norm": 0.48194718148390925,
+      "learning_rate": 1.038338456509118e-06,
+      "loss": 0.2539,
+      "step": 4864
+    },
+    {
+      "epoch": 1.6243739565943238,
+      "grad_norm": 0.4588857979305098,
+      "learning_rate": 1.0365614873853463e-06,
+      "loss": 0.2435,
+      "step": 4865
+    },
+    {
+      "epoch": 1.6247078464106846,
+      "grad_norm": 0.4715371208347117,
+      "learning_rate": 1.0347858642247998e-06,
+      "loss": 0.2551,
+      "step": 4866
+    },
+    {
+      "epoch": 1.625041736227045,
+      "grad_norm": 0.4496393916718381,
+      "learning_rate": 1.033011587630472e-06,
+      "loss": 0.2363,
+      "step": 4867
+    },
+    {
+      "epoch": 1.6253756260434056,
+      "grad_norm": 0.46343359508158743,
+      "learning_rate": 1.0312386582048994e-06,
+      "loss": 0.2531,
+      "step": 4868
+    },
+    {
+      "epoch": 1.6257095158597663,
+      "grad_norm": 0.4894193747132886,
+      "learning_rate": 1.0294670765501592e-06,
+      "loss": 0.2564,
+      "step": 4869
+    },
+    {
+      "epoch": 1.626043405676127,
+      "grad_norm": 0.48566881006444007,
+      "learning_rate": 1.0276968432678702e-06,
+      "loss": 0.2622,
+      "step": 4870
+    },
+    {
+      "epoch": 1.6263772954924876,
+      "grad_norm": 0.47224493903419157,
+      "learning_rate": 1.0259279589591936e-06,
+      "loss": 0.2504,
+      "step": 4871
+    },
+    {
+      "epoch": 1.626711185308848,
+      "grad_norm": 0.4812637311016245,
+      "learning_rate": 1.0241604242248365e-06,
+      "loss": 0.2434,
+      "step": 4872
+    },
+    {
+      "epoch": 1.6270450751252086,
+      "grad_norm": 0.46820355085649296,
+      "learning_rate": 1.022394239665042e-06,
+      "loss": 0.2571,
+      "step": 4873
+    },
+    {
+      "epoch": 1.6273789649415693,
+      "grad_norm": 0.4763150747531845,
+      "learning_rate": 1.0206294058795974e-06,
+      "loss": 0.254,
+      "step": 4874
+    },
+    {
+      "epoch": 1.62771285475793,
+      "grad_norm": 0.4708134960632262,
+      "learning_rate": 1.0188659234678327e-06,
+      "loss": 0.24,
+      "step": 4875
+    },
+    {
+      "epoch": 1.6280467445742905,
+      "grad_norm": 0.4749485783426798,
+      "learning_rate": 1.0171037930286154e-06,
+      "loss": 0.2603,
+      "step": 4876
+    },
+    {
+      "epoch": 1.628380634390651,
+      "grad_norm": 0.4509518586474006,
+      "learning_rate": 1.0153430151603583e-06,
+      "loss": 0.235,
+      "step": 4877
+    },
+    {
+      "epoch": 1.6287145242070116,
+      "grad_norm": 0.48120165369628776,
+      "learning_rate": 1.013583590461011e-06,
+      "loss": 0.2582,
+      "step": 4878
+    },
+    {
+      "epoch": 1.6290484140233723,
+      "grad_norm": 0.4562192302910925,
+      "learning_rate": 1.0118255195280646e-06,
+      "loss": 0.2521,
+      "step": 4879
+    },
+    {
+      "epoch": 1.629382303839733,
+      "grad_norm": 0.47243139531708606,
+      "learning_rate": 1.010068802958551e-06,
+      "loss": 0.2574,
+      "step": 4880
+    },
+    {
+      "epoch": 1.6297161936560935,
+      "grad_norm": 0.4609198045445573,
+      "learning_rate": 1.0083134413490431e-06,
+      "loss": 0.2518,
+      "step": 4881
+    },
+    {
+      "epoch": 1.630050083472454,
+      "grad_norm": 0.47745440260582594,
+      "learning_rate": 1.0065594352956526e-06,
+      "loss": 0.2498,
+      "step": 4882
+    },
+    {
+      "epoch": 1.6303839732888146,
+      "grad_norm": 0.4557872618371228,
+      "learning_rate": 1.0048067853940286e-06,
+      "loss": 0.2414,
+      "step": 4883
+    },
+    {
+      "epoch": 1.6307178631051753,
+      "grad_norm": 0.4565132153673048,
+      "learning_rate": 1.0030554922393648e-06,
+      "loss": 0.2471,
+      "step": 4884
+    },
+    {
+      "epoch": 1.631051752921536,
+      "grad_norm": 0.46269549581438746,
+      "learning_rate": 1.001305556426389e-06,
+      "loss": 0.2493,
+      "step": 4885
+    },
+    {
+      "epoch": 1.6313856427378965,
+      "grad_norm": 0.4627667348229286,
+      "learning_rate": 9.99556978549373e-07,
+      "loss": 0.242,
+      "step": 4886
+    },
+    {
+      "epoch": 1.631719532554257,
+      "grad_norm": 0.4755233701557241,
+      "learning_rate": 9.978097592021236e-07,
+      "loss": 0.2413,
+      "step": 4887
+    },
+    {
+      "epoch": 1.6320534223706176,
+      "grad_norm": 0.4768997961771595,
+      "learning_rate": 9.96063898977987e-07,
+      "loss": 0.2498,
+      "step": 4888
+    },
+    {
+      "epoch": 1.6323873121869783,
+      "grad_norm": 0.45552454636914186,
+      "learning_rate": 9.943193984698469e-07,
+      "loss": 0.2453,
+      "step": 4889
+    },
+    {
+      "epoch": 1.632721202003339,
+      "grad_norm": 0.47105562021185315,
+      "learning_rate": 9.9257625827013e-07,
+      "loss": 0.2433,
+      "step": 4890
+    },
+    {
+      "epoch": 1.6330550918196995,
+      "grad_norm": 0.472841161356641,
+      "learning_rate": 9.90834478970794e-07,
+      "loss": 0.2513,
+      "step": 4891
+    },
+    {
+      "epoch": 1.63338898163606,
+      "grad_norm": 0.46900736515602254,
+      "learning_rate": 9.890940611633416e-07,
+      "loss": 0.2439,
+      "step": 4892
+    },
+    {
+      "epoch": 1.6337228714524206,
+      "grad_norm": 0.4612071012068493,
+      "learning_rate": 9.873550054388082e-07,
+      "loss": 0.246,
+      "step": 4893
+    },
+    {
+      "epoch": 1.6340567612687813,
+      "grad_norm": 0.4792340563095537,
+      "learning_rate": 9.856173123877667e-07,
+      "loss": 0.2652,
+      "step": 4894
+    },
+    {
+      "epoch": 1.634390651085142,
+      "grad_norm": 0.4582438028363659,
+      "learning_rate": 9.838809826003315e-07,
+      "loss": 0.2424,
+      "step": 4895
+    },
+    {
+      "epoch": 1.6347245409015025,
+      "grad_norm": 0.4638202559759776,
+      "learning_rate": 9.8214601666615e-07,
+      "loss": 0.2451,
+      "step": 4896
+    },
+    {
+      "epoch": 1.635058430717863,
+      "grad_norm": 0.4569279141401279,
+      "learning_rate": 9.80412415174407e-07,
+      "loss": 0.2452,
+      "step": 4897
+    },
+    {
+      "epoch": 1.6353923205342236,
+      "grad_norm": 0.46535563338618335,
+      "learning_rate": 9.786801787138245e-07,
+      "loss": 0.2519,
+      "step": 4898
+    },
+    {
+      "epoch": 1.6357262103505843,
+      "grad_norm": 0.4521123654364428,
+      "learning_rate": 9.769493078726622e-07,
+      "loss": 0.242,
+      "step": 4899
+    },
+    {
+      "epoch": 1.636060100166945,
+      "grad_norm": 0.4667062627102446,
+      "learning_rate": 9.752198032387133e-07,
+      "loss": 0.2432,
+      "step": 4900
+    },
+    {
+      "epoch": 1.6363939899833055,
+      "grad_norm": 0.4676111158884403,
+      "learning_rate": 9.734916653993104e-07,
+      "loss": 0.252,
+      "step": 4901
+    },
+    {
+      "epoch": 1.636727879799666,
+      "grad_norm": 0.4596657179719377,
+      "learning_rate": 9.7176489494132e-07,
+      "loss": 0.2362,
+      "step": 4902
+    },
+    {
+      "epoch": 1.6370617696160266,
+      "grad_norm": 0.46067125562693506,
+      "learning_rate": 9.700394924511424e-07,
+      "loss": 0.2455,
+      "step": 4903
+    },
+    {
+      "epoch": 1.6373956594323873,
+      "grad_norm": 0.475080656259677,
+      "learning_rate": 9.68315458514718e-07,
+      "loss": 0.2481,
+      "step": 4904
+    },
+    {
+      "epoch": 1.637729549248748,
+      "grad_norm": 0.48224630338585267,
+      "learning_rate": 9.665927937175184e-07,
+      "loss": 0.2514,
+      "step": 4905
+    },
+    {
+      "epoch": 1.6380634390651085,
+      "grad_norm": 0.47949393472541924,
+      "learning_rate": 9.648714986445523e-07,
+      "loss": 0.2523,
+      "step": 4906
+    },
+    {
+      "epoch": 1.638397328881469,
+      "grad_norm": 0.47820669235970414,
+      "learning_rate": 9.631515738803604e-07,
+      "loss": 0.2547,
+      "step": 4907
+    },
+    {
+      "epoch": 1.6387312186978296,
+      "grad_norm": 0.4757327523219286,
+      "learning_rate": 9.614330200090221e-07,
+      "loss": 0.2478,
+      "step": 4908
+    },
+    {
+      "epoch": 1.6390651085141903,
+      "grad_norm": 0.47070907741945284,
+      "learning_rate": 9.597158376141513e-07,
+      "loss": 0.2573,
+      "step": 4909
+    },
+    {
+      "epoch": 1.639398998330551,
+      "grad_norm": 0.47803862441126505,
+      "learning_rate": 9.580000272788915e-07,
+      "loss": 0.2423,
+      "step": 4910
+    },
+    {
+      "epoch": 1.6397328881469115,
+      "grad_norm": 0.45913013146266524,
+      "learning_rate": 9.56285589585924e-07,
+      "loss": 0.2429,
+      "step": 4911
+    },
+    {
+      "epoch": 1.640066777963272,
+      "grad_norm": 0.45694516650876316,
+      "learning_rate": 9.545725251174616e-07,
+      "loss": 0.2388,
+      "step": 4912
+    },
+    {
+      "epoch": 1.6404006677796326,
+      "grad_norm": 0.4449524985140172,
+      "learning_rate": 9.528608344552548e-07,
+      "loss": 0.2419,
+      "step": 4913
+    },
+    {
+      "epoch": 1.6407345575959933,
+      "grad_norm": 0.4802755652413974,
+      "learning_rate": 9.51150518180583e-07,
+      "loss": 0.2482,
+      "step": 4914
+    },
+    {
+      "epoch": 1.641068447412354,
+      "grad_norm": 0.4571352948186702,
+      "learning_rate": 9.494415768742616e-07,
+      "loss": 0.2346,
+      "step": 4915
+    },
+    {
+      "epoch": 1.6414023372287145,
+      "grad_norm": 0.47478338201279185,
+      "learning_rate": 9.477340111166361e-07,
+      "loss": 0.254,
+      "step": 4916
+    },
+    {
+      "epoch": 1.641736227045075,
+      "grad_norm": 0.4693123820351987,
+      "learning_rate": 9.460278214875884e-07,
+      "loss": 0.2534,
+      "step": 4917
+    },
+    {
+      "epoch": 1.6420701168614358,
+      "grad_norm": 0.4809256454333158,
+      "learning_rate": 9.443230085665339e-07,
+      "loss": 0.2468,
+      "step": 4918
+    },
+    {
+      "epoch": 1.6424040066777963,
+      "grad_norm": 0.4632073680363049,
+      "learning_rate": 9.426195729324161e-07,
+      "loss": 0.2469,
+      "step": 4919
+    },
+    {
+      "epoch": 1.642737896494157,
+      "grad_norm": 0.46042985452089663,
+      "learning_rate": 9.409175151637134e-07,
+      "loss": 0.2359,
+      "step": 4920
+    },
+    {
+      "epoch": 1.6430717863105175,
+      "grad_norm": 0.4640487841794795,
+      "learning_rate": 9.392168358384346e-07,
+      "loss": 0.2406,
+      "step": 4921
+    },
+    {
+      "epoch": 1.643405676126878,
+      "grad_norm": 0.45621737586279043,
+      "learning_rate": 9.375175355341232e-07,
+      "loss": 0.2391,
+      "step": 4922
+    },
+    {
+      "epoch": 1.6437395659432388,
+      "grad_norm": 0.44800154899370676,
+      "learning_rate": 9.358196148278525e-07,
+      "loss": 0.2426,
+      "step": 4923
+    },
+    {
+      "epoch": 1.6440734557595993,
+      "grad_norm": 0.4852409825819238,
+      "learning_rate": 9.34123074296227e-07,
+      "loss": 0.2576,
+      "step": 4924
+    },
+    {
+      "epoch": 1.64440734557596,
+      "grad_norm": 0.4567852558403836,
+      "learning_rate": 9.324279145153825e-07,
+      "loss": 0.2482,
+      "step": 4925
+    },
+    {
+      "epoch": 1.6447412353923205,
+      "grad_norm": 0.4664590682752434,
+      "learning_rate": 9.307341360609873e-07,
+      "loss": 0.2446,
+      "step": 4926
+    },
+    {
+      "epoch": 1.645075125208681,
+      "grad_norm": 0.45559022681496886,
+      "learning_rate": 9.290417395082413e-07,
+      "loss": 0.2444,
+      "step": 4927
+    },
+    {
+      "epoch": 1.6454090150250418,
+      "grad_norm": 0.44863013961347553,
+      "learning_rate": 9.27350725431872e-07,
+      "loss": 0.2395,
+      "step": 4928
+    },
+    {
+      "epoch": 1.6457429048414023,
+      "grad_norm": 0.47715696611050523,
+      "learning_rate": 9.25661094406139e-07,
+      "loss": 0.2518,
+      "step": 4929
+    },
+    {
+      "epoch": 1.646076794657763,
+      "grad_norm": 0.48167664930758675,
+      "learning_rate": 9.239728470048315e-07,
+      "loss": 0.258,
+      "step": 4930
+    },
+    {
+      "epoch": 1.6464106844741235,
+      "grad_norm": 0.4604362169133484,
+      "learning_rate": 9.222859838012721e-07,
+      "loss": 0.2469,
+      "step": 4931
+    },
+    {
+      "epoch": 1.646744574290484,
+      "grad_norm": 0.48012531936692315,
+      "learning_rate": 9.206005053683093e-07,
+      "loss": 0.2506,
+      "step": 4932
+    },
+    {
+      "epoch": 1.6470784641068448,
+      "grad_norm": 0.49554110290195946,
+      "learning_rate": 9.189164122783222e-07,
+      "loss": 0.2463,
+      "step": 4933
+    },
+    {
+      "epoch": 1.6474123539232055,
+      "grad_norm": 0.4649347975929347,
+      "learning_rate": 9.172337051032204e-07,
+      "loss": 0.2418,
+      "step": 4934
+    },
+    {
+      "epoch": 1.647746243739566,
+      "grad_norm": 0.48417684964527313,
+      "learning_rate": 9.15552384414442e-07,
+      "loss": 0.2461,
+      "step": 4935
+    },
+    {
+      "epoch": 1.6480801335559265,
+      "grad_norm": 0.47633059784699244,
+      "learning_rate": 9.138724507829577e-07,
+      "loss": 0.2505,
+      "step": 4936
+    },
+    {
+      "epoch": 1.648414023372287,
+      "grad_norm": 0.45838805223901635,
+      "learning_rate": 9.121939047792622e-07,
+      "loss": 0.2296,
+      "step": 4937
+    },
+    {
+      "epoch": 1.6487479131886478,
+      "grad_norm": 0.46132545765027866,
+      "learning_rate": 9.105167469733811e-07,
+      "loss": 0.2419,
+      "step": 4938
+    },
+    {
+      "epoch": 1.6490818030050085,
+      "grad_norm": 0.467456766098229,
+      "learning_rate": 9.088409779348673e-07,
+      "loss": 0.2494,
+      "step": 4939
+    },
+    {
+      "epoch": 1.649415692821369,
+      "grad_norm": 0.4833320987342185,
+      "learning_rate": 9.07166598232806e-07,
+      "loss": 0.2508,
+      "step": 4940
+    },
+    {
+      "epoch": 1.6497495826377295,
+      "grad_norm": 0.46326224167356483,
+      "learning_rate": 9.054936084358062e-07,
+      "loss": 0.2483,
+      "step": 4941
+    },
+    {
+      "epoch": 1.65008347245409,
+      "grad_norm": 0.47223452361873847,
+      "learning_rate": 9.03822009112006e-07,
+      "loss": 0.244,
+      "step": 4942
+    },
+    {
+      "epoch": 1.6504173622704508,
+      "grad_norm": 0.4761762906304975,
+      "learning_rate": 9.021518008290741e-07,
+      "loss": 0.2317,
+      "step": 4943
+    },
+    {
+      "epoch": 1.6507512520868115,
+      "grad_norm": 0.48123736035766285,
+      "learning_rate": 9.004829841542018e-07,
+      "loss": 0.2615,
+      "step": 4944
+    },
+    {
+      "epoch": 1.651085141903172,
+      "grad_norm": 0.4501171434936968,
+      "learning_rate": 8.988155596541131e-07,
+      "loss": 0.2408,
+      "step": 4945
+    },
+    {
+      "epoch": 1.6514190317195325,
+      "grad_norm": 0.48329660038526984,
+      "learning_rate": 8.971495278950559e-07,
+      "loss": 0.252,
+      "step": 4946
+    },
+    {
+      "epoch": 1.651752921535893,
+      "grad_norm": 0.46222117166723037,
+      "learning_rate": 8.954848894428053e-07,
+      "loss": 0.2354,
+      "step": 4947
+    },
+    {
+      "epoch": 1.6520868113522538,
+      "grad_norm": 0.46552867162565614,
+      "learning_rate": 8.938216448626635e-07,
+      "loss": 0.2471,
+      "step": 4948
+    },
+    {
+      "epoch": 1.6524207011686145,
+      "grad_norm": 0.4600558775275079,
+      "learning_rate": 8.921597947194621e-07,
+      "loss": 0.2398,
+      "step": 4949
+    },
+    {
+      "epoch": 1.652754590984975,
+      "grad_norm": 0.4715886447806753,
+      "learning_rate": 8.904993395775552e-07,
+      "loss": 0.2501,
+      "step": 4950
+    },
+    {
+      "epoch": 1.6530884808013355,
+      "grad_norm": 0.4505248283100601,
+      "learning_rate": 8.888402800008239e-07,
+      "loss": 0.2383,
+      "step": 4951
+    },
+    {
+      "epoch": 1.653422370617696,
+      "grad_norm": 0.4798320435216762,
+      "learning_rate": 8.871826165526781e-07,
+      "loss": 0.2567,
+      "step": 4952
+    },
+    {
+      "epoch": 1.6537562604340568,
+      "grad_norm": 0.47726668910435616,
+      "learning_rate": 8.855263497960504e-07,
+      "loss": 0.2596,
+      "step": 4953
+    },
+    {
+      "epoch": 1.6540901502504175,
+      "grad_norm": 0.46352034168788253,
+      "learning_rate": 8.838714802934017e-07,
+      "loss": 0.2513,
+      "step": 4954
+    },
+    {
+      "epoch": 1.654424040066778,
+      "grad_norm": 0.47921640539773924,
+      "learning_rate": 8.822180086067161e-07,
+      "loss": 0.2429,
+      "step": 4955
+    },
+    {
+      "epoch": 1.6547579298831385,
+      "grad_norm": 0.48441695837572335,
+      "learning_rate": 8.805659352975049e-07,
+      "loss": 0.2511,
+      "step": 4956
+    },
+    {
+      "epoch": 1.655091819699499,
+      "grad_norm": 0.46805684845311546,
+      "learning_rate": 8.789152609268015e-07,
+      "loss": 0.2506,
+      "step": 4957
+    },
+    {
+      "epoch": 1.6554257095158598,
+      "grad_norm": 0.46839955438686376,
+      "learning_rate": 8.772659860551691e-07,
+      "loss": 0.2548,
+      "step": 4958
+    },
+    {
+      "epoch": 1.6557595993322205,
+      "grad_norm": 0.4687524938251378,
+      "learning_rate": 8.756181112426909e-07,
+      "loss": 0.2463,
+      "step": 4959
+    },
+    {
+      "epoch": 1.656093489148581,
+      "grad_norm": 0.4728896864361125,
+      "learning_rate": 8.739716370489787e-07,
+      "loss": 0.2627,
+      "step": 4960
+    },
+    {
+      "epoch": 1.6564273789649415,
+      "grad_norm": 0.4666871731754719,
+      "learning_rate": 8.723265640331652e-07,
+      "loss": 0.2429,
+      "step": 4961
+    },
+    {
+      "epoch": 1.656761268781302,
+      "grad_norm": 0.4752142579221166,
+      "learning_rate": 8.706828927539079e-07,
+      "loss": 0.2489,
+      "step": 4962
+    },
+    {
+      "epoch": 1.6570951585976628,
+      "grad_norm": 0.4684947553918931,
+      "learning_rate": 8.690406237693916e-07,
+      "loss": 0.2467,
+      "step": 4963
+    },
+    {
+      "epoch": 1.6574290484140235,
+      "grad_norm": 0.4645926526693745,
+      "learning_rate": 8.673997576373205e-07,
+      "loss": 0.2459,
+      "step": 4964
+    },
+    {
+      "epoch": 1.657762938230384,
+      "grad_norm": 0.47546524501524107,
+      "learning_rate": 8.657602949149252e-07,
+      "loss": 0.2539,
+      "step": 4965
+    },
+    {
+      "epoch": 1.6580968280467445,
+      "grad_norm": 0.4747000807852713,
+      "learning_rate": 8.641222361589563e-07,
+      "loss": 0.2558,
+      "step": 4966
+    },
+    {
+      "epoch": 1.658430717863105,
+      "grad_norm": 0.5076241374991348,
+      "learning_rate": 8.624855819256922e-07,
+      "loss": 0.2503,
+      "step": 4967
+    },
+    {
+      "epoch": 1.6587646076794658,
+      "grad_norm": 0.46004535666171015,
+      "learning_rate": 8.608503327709333e-07,
+      "loss": 0.2391,
+      "step": 4968
+    },
+    {
+      "epoch": 1.6590984974958265,
+      "grad_norm": 0.49196893529908076,
+      "learning_rate": 8.592164892500005e-07,
+      "loss": 0.2475,
+      "step": 4969
+    },
+    {
+      "epoch": 1.659432387312187,
+      "grad_norm": 0.45405743036956275,
+      "learning_rate": 8.575840519177386e-07,
+      "loss": 0.2512,
+      "step": 4970
+    },
+    {
+      "epoch": 1.6597662771285475,
+      "grad_norm": 0.4498165709164128,
+      "learning_rate": 8.559530213285133e-07,
+      "loss": 0.2376,
+      "step": 4971
+    },
+    {
+      "epoch": 1.660100166944908,
+      "grad_norm": 0.46090932247948446,
+      "learning_rate": 8.543233980362175e-07,
+      "loss": 0.243,
+      "step": 4972
+    },
+    {
+      "epoch": 1.6604340567612688,
+      "grad_norm": 0.4752167783328063,
+      "learning_rate": 8.526951825942608e-07,
+      "loss": 0.2519,
+      "step": 4973
+    },
+    {
+      "epoch": 1.6607679465776295,
+      "grad_norm": 0.476115428703703,
+      "learning_rate": 8.510683755555771e-07,
+      "loss": 0.2479,
+      "step": 4974
+    },
+    {
+      "epoch": 1.66110183639399,
+      "grad_norm": 0.4666506771059333,
+      "learning_rate": 8.494429774726198e-07,
+      "loss": 0.2452,
+      "step": 4975
+    },
+    {
+      "epoch": 1.6614357262103505,
+      "grad_norm": 0.48556914903423154,
+      "learning_rate": 8.478189888973676e-07,
+      "loss": 0.2577,
+      "step": 4976
+    },
+    {
+      "epoch": 1.661769616026711,
+      "grad_norm": 0.46130148453848513,
+      "learning_rate": 8.461964103813203e-07,
+      "loss": 0.2403,
+      "step": 4977
+    },
+    {
+      "epoch": 1.6621035058430718,
+      "grad_norm": 0.47739094361729234,
+      "learning_rate": 8.44575242475495e-07,
+      "loss": 0.2471,
+      "step": 4978
+    },
+    {
+      "epoch": 1.6624373956594325,
+      "grad_norm": 0.4511096477550543,
+      "learning_rate": 8.429554857304318e-07,
+      "loss": 0.2326,
+      "step": 4979
+    },
+    {
+      "epoch": 1.662771285475793,
+      "grad_norm": 0.4841338421193107,
+      "learning_rate": 8.413371406961918e-07,
+      "loss": 0.2489,
+      "step": 4980
+    },
+    {
+      "epoch": 1.6631051752921535,
+      "grad_norm": 0.47402471946529856,
+      "learning_rate": 8.397202079223577e-07,
+      "loss": 0.2439,
+      "step": 4981
+    },
+    {
+      "epoch": 1.663439065108514,
+      "grad_norm": 0.4699029537083943,
+      "learning_rate": 8.381046879580307e-07,
+      "loss": 0.2548,
+      "step": 4982
+    },
+    {
+      "epoch": 1.6637729549248748,
+      "grad_norm": 0.46860745528654624,
+      "learning_rate": 8.364905813518331e-07,
+      "loss": 0.2538,
+      "step": 4983
+    },
+    {
+      "epoch": 1.6641068447412355,
+      "grad_norm": 0.47867468115979955,
+      "learning_rate": 8.348778886519065e-07,
+      "loss": 0.2405,
+      "step": 4984
+    },
+    {
+      "epoch": 1.664440734557596,
+      "grad_norm": 0.46777937534266273,
+      "learning_rate": 8.332666104059134e-07,
+      "loss": 0.2534,
+      "step": 4985
+    },
+    {
+      "epoch": 1.6647746243739565,
+      "grad_norm": 0.47251336571418756,
+      "learning_rate": 8.316567471610371e-07,
+      "loss": 0.2479,
+      "step": 4986
+    },
+    {
+      "epoch": 1.6651085141903172,
+      "grad_norm": 0.4750499335928575,
+      "learning_rate": 8.300482994639775e-07,
+      "loss": 0.2518,
+      "step": 4987
+    },
+    {
+      "epoch": 1.6654424040066778,
+      "grad_norm": 0.4798535795229569,
+      "learning_rate": 8.284412678609555e-07,
+      "loss": 0.2367,
+      "step": 4988
+    },
+    {
+      "epoch": 1.6657762938230385,
+      "grad_norm": 0.46318099426372356,
+      "learning_rate": 8.268356528977089e-07,
+      "loss": 0.2423,
+      "step": 4989
+    },
+    {
+      "epoch": 1.666110183639399,
+      "grad_norm": 0.46747578216819513,
+      "learning_rate": 8.25231455119499e-07,
+      "loss": 0.2467,
+      "step": 4990
+    },
+    {
+      "epoch": 1.6664440734557595,
+      "grad_norm": 0.49975035843433036,
+      "learning_rate": 8.23628675071102e-07,
+      "loss": 0.255,
+      "step": 4991
+    },
+    {
+      "epoch": 1.6667779632721202,
+      "grad_norm": 0.45165734536051017,
+      "learning_rate": 8.220273132968137e-07,
+      "loss": 0.2469,
+      "step": 4992
+    },
+    {
+      "epoch": 1.6671118530884808,
+      "grad_norm": 0.46528403431593895,
+      "learning_rate": 8.204273703404463e-07,
+      "loss": 0.2544,
+      "step": 4993
+    },
+    {
+      "epoch": 1.6674457429048415,
+      "grad_norm": 0.45421870243038714,
+      "learning_rate": 8.188288467453348e-07,
+      "loss": 0.2417,
+      "step": 4994
+    },
+    {
+      "epoch": 1.667779632721202,
+      "grad_norm": 0.46231132359491356,
+      "learning_rate": 8.172317430543292e-07,
+      "loss": 0.2525,
+      "step": 4995
+    },
+    {
+      "epoch": 1.6681135225375625,
+      "grad_norm": 0.47044550599312424,
+      "learning_rate": 8.156360598097979e-07,
+      "loss": 0.2453,
+      "step": 4996
+    },
+    {
+      "epoch": 1.6684474123539232,
+      "grad_norm": 0.4598152474595619,
+      "learning_rate": 8.140417975536258e-07,
+      "loss": 0.25,
+      "step": 4997
+    },
+    {
+      "epoch": 1.6687813021702838,
+      "grad_norm": 0.466002285400396,
+      "learning_rate": 8.124489568272148e-07,
+      "loss": 0.2472,
+      "step": 4998
+    },
+    {
+      "epoch": 1.6691151919866445,
+      "grad_norm": 0.45284126301746735,
+      "learning_rate": 8.108575381714883e-07,
+      "loss": 0.244,
+      "step": 4999
+    },
+    {
+      "epoch": 1.669449081803005,
+      "grad_norm": 0.48490427003566994,
+      "learning_rate": 8.092675421268825e-07,
+      "loss": 0.2548,
+      "step": 5000
+    },
+    {
+      "epoch": 1.6697829716193655,
+      "grad_norm": 0.4672179334124505,
+      "learning_rate": 8.076789692333503e-07,
+      "loss": 0.2438,
+      "step": 5001
+    },
+    {
+      "epoch": 1.6701168614357262,
+      "grad_norm": 0.4614802136278703,
+      "learning_rate": 8.060918200303658e-07,
+      "loss": 0.2374,
+      "step": 5002
+    },
+    {
+      "epoch": 1.670450751252087,
+      "grad_norm": 0.48682527623131977,
+      "learning_rate": 8.045060950569139e-07,
+      "loss": 0.2483,
+      "step": 5003
+    },
+    {
+      "epoch": 1.6707846410684475,
+      "grad_norm": 0.46328283650493507,
+      "learning_rate": 8.029217948515006e-07,
+      "loss": 0.2503,
+      "step": 5004
+    },
+    {
+      "epoch": 1.671118530884808,
+      "grad_norm": 0.46464645642852187,
+      "learning_rate": 8.013389199521459e-07,
+      "loss": 0.2522,
+      "step": 5005
+    },
+    {
+      "epoch": 1.6714524207011685,
+      "grad_norm": 0.4721344459220853,
+      "learning_rate": 7.997574708963852e-07,
+      "loss": 0.2398,
+      "step": 5006
+    },
+    {
+      "epoch": 1.6717863105175292,
+      "grad_norm": 0.4579193727245352,
+      "learning_rate": 7.981774482212695e-07,
+      "loss": 0.2365,
+      "step": 5007
+    },
+    {
+      "epoch": 1.67212020033389,
+      "grad_norm": 0.4732887019175692,
+      "learning_rate": 7.965988524633683e-07,
+      "loss": 0.2568,
+      "step": 5008
+    },
+    {
+      "epoch": 1.6724540901502505,
+      "grad_norm": 0.49056713393584306,
+      "learning_rate": 7.950216841587638e-07,
+      "loss": 0.24,
+      "step": 5009
+    },
+    {
+      "epoch": 1.672787979966611,
+      "grad_norm": 0.48456313930211925,
+      "learning_rate": 7.934459438430531e-07,
+      "loss": 0.2393,
+      "step": 5010
+    },
+    {
+      "epoch": 1.6731218697829715,
+      "grad_norm": 0.48014180640632775,
+      "learning_rate": 7.918716320513514e-07,
+      "loss": 0.2565,
+      "step": 5011
+    },
+    {
+      "epoch": 1.6734557595993322,
+      "grad_norm": 0.4711865223041657,
+      "learning_rate": 7.902987493182846e-07,
+      "loss": 0.2426,
+      "step": 5012
+    },
+    {
+      "epoch": 1.673789649415693,
+      "grad_norm": 0.472709011103895,
+      "learning_rate": 7.887272961779979e-07,
+      "loss": 0.2446,
+      "step": 5013
+    },
+    {
+      "epoch": 1.6741235392320535,
+      "grad_norm": 0.48422080608316703,
+      "learning_rate": 7.871572731641481e-07,
+      "loss": 0.2418,
+      "step": 5014
+    },
+    {
+      "epoch": 1.674457429048414,
+      "grad_norm": 0.4761122933797272,
+      "learning_rate": 7.855886808099062e-07,
+      "loss": 0.2571,
+      "step": 5015
+    },
+    {
+      "epoch": 1.6747913188647745,
+      "grad_norm": 0.4700570179764089,
+      "learning_rate": 7.840215196479567e-07,
+      "loss": 0.2438,
+      "step": 5016
+    },
+    {
+      "epoch": 1.6751252086811352,
+      "grad_norm": 0.47025960647320664,
+      "learning_rate": 7.824557902105024e-07,
+      "loss": 0.2527,
+      "step": 5017
+    },
+    {
+      "epoch": 1.675459098497496,
+      "grad_norm": 0.46270113957792497,
+      "learning_rate": 7.808914930292544e-07,
+      "loss": 0.2394,
+      "step": 5018
+    },
+    {
+      "epoch": 1.6757929883138565,
+      "grad_norm": 0.48500093646748715,
+      "learning_rate": 7.793286286354418e-07,
+      "loss": 0.2555,
+      "step": 5019
+    },
+    {
+      "epoch": 1.676126878130217,
+      "grad_norm": 0.47826202933921313,
+      "learning_rate": 7.77767197559805e-07,
+      "loss": 0.2546,
+      "step": 5020
+    },
+    {
+      "epoch": 1.6764607679465775,
+      "grad_norm": 0.47407534335531915,
+      "learning_rate": 7.76207200332596e-07,
+      "loss": 0.2544,
+      "step": 5021
+    },
+    {
+      "epoch": 1.6767946577629382,
+      "grad_norm": 0.47150093628069456,
+      "learning_rate": 7.746486374835843e-07,
+      "loss": 0.2536,
+      "step": 5022
+    },
+    {
+      "epoch": 1.677128547579299,
+      "grad_norm": 0.462027788473639,
+      "learning_rate": 7.730915095420489e-07,
+      "loss": 0.245,
+      "step": 5023
+    },
+    {
+      "epoch": 1.6774624373956595,
+      "grad_norm": 0.4649191177930882,
+      "learning_rate": 7.71535817036782e-07,
+      "loss": 0.2393,
+      "step": 5024
+    },
+    {
+      "epoch": 1.67779632721202,
+      "grad_norm": 0.4849481650777474,
+      "learning_rate": 7.699815604960892e-07,
+      "loss": 0.2507,
+      "step": 5025
+    },
+    {
+      "epoch": 1.6781302170283805,
+      "grad_norm": 0.4644209573050273,
+      "learning_rate": 7.684287404477864e-07,
+      "loss": 0.2422,
+      "step": 5026
+    },
+    {
+      "epoch": 1.6784641068447412,
+      "grad_norm": 0.46119881775329435,
+      "learning_rate": 7.66877357419204e-07,
+      "loss": 0.2398,
+      "step": 5027
+    },
+    {
+      "epoch": 1.678797996661102,
+      "grad_norm": 0.46236042178484227,
+      "learning_rate": 7.653274119371862e-07,
+      "loss": 0.2394,
+      "step": 5028
+    },
+    {
+      "epoch": 1.6791318864774625,
+      "grad_norm": 0.4723085063029113,
+      "learning_rate": 7.637789045280841e-07,
+      "loss": 0.2394,
+      "step": 5029
+    },
+    {
+      "epoch": 1.679465776293823,
+      "grad_norm": 0.45681522076877495,
+      "learning_rate": 7.622318357177633e-07,
+      "loss": 0.2433,
+      "step": 5030
+    },
+    {
+      "epoch": 1.6797996661101835,
+      "grad_norm": 0.48335212312730713,
+      "learning_rate": 7.606862060315989e-07,
+      "loss": 0.2514,
+      "step": 5031
+    },
+    {
+      "epoch": 1.6801335559265442,
+      "grad_norm": 0.4736701627836829,
+      "learning_rate": 7.591420159944813e-07,
+      "loss": 0.242,
+      "step": 5032
+    },
+    {
+      "epoch": 1.680467445742905,
+      "grad_norm": 0.46956932488014397,
+      "learning_rate": 7.575992661308085e-07,
+      "loss": 0.2321,
+      "step": 5033
+    },
+    {
+      "epoch": 1.6808013355592655,
+      "grad_norm": 0.471765721166744,
+      "learning_rate": 7.5605795696449e-07,
+      "loss": 0.2554,
+      "step": 5034
+    },
+    {
+      "epoch": 1.681135225375626,
+      "grad_norm": 0.48000926530432564,
+      "learning_rate": 7.545180890189441e-07,
+      "loss": 0.2514,
+      "step": 5035
+    },
+    {
+      "epoch": 1.6814691151919865,
+      "grad_norm": 0.47341723597699376,
+      "learning_rate": 7.529796628171071e-07,
+      "loss": 0.2497,
+      "step": 5036
+    },
+    {
+      "epoch": 1.6818030050083472,
+      "grad_norm": 0.4600007862621662,
+      "learning_rate": 7.514426788814178e-07,
+      "loss": 0.2432,
+      "step": 5037
+    },
+    {
+      "epoch": 1.682136894824708,
+      "grad_norm": 0.4702744010502343,
+      "learning_rate": 7.499071377338279e-07,
+      "loss": 0.2543,
+      "step": 5038
+    },
+    {
+      "epoch": 1.6824707846410685,
+      "grad_norm": 0.46601904772860836,
+      "learning_rate": 7.483730398958001e-07,
+      "loss": 0.2352,
+      "step": 5039
+    },
+    {
+      "epoch": 1.682804674457429,
+      "grad_norm": 0.46439790041028695,
+      "learning_rate": 7.468403858883038e-07,
+      "loss": 0.244,
+      "step": 5040
+    },
+    {
+      "epoch": 1.6831385642737895,
+      "grad_norm": 0.48161299864481594,
+      "learning_rate": 7.453091762318232e-07,
+      "loss": 0.2468,
+      "step": 5041
+    },
+    {
+      "epoch": 1.6834724540901502,
+      "grad_norm": 0.47230340182080127,
+      "learning_rate": 7.437794114463482e-07,
+      "loss": 0.2491,
+      "step": 5042
+    },
+    {
+      "epoch": 1.683806343906511,
+      "grad_norm": 0.4736835455022473,
+      "learning_rate": 7.422510920513792e-07,
+      "loss": 0.2548,
+      "step": 5043
+    },
+    {
+      "epoch": 1.6841402337228715,
+      "grad_norm": 0.4754812088898494,
+      "learning_rate": 7.407242185659236e-07,
+      "loss": 0.2464,
+      "step": 5044
+    },
+    {
+      "epoch": 1.684474123539232,
+      "grad_norm": 0.4755260730684596,
+      "learning_rate": 7.391987915085014e-07,
+      "loss": 0.2407,
+      "step": 5045
+    },
+    {
+      "epoch": 1.6848080133555925,
+      "grad_norm": 0.4556290855717323,
+      "learning_rate": 7.376748113971404e-07,
+      "loss": 0.2442,
+      "step": 5046
+    },
+    {
+      "epoch": 1.6851419031719532,
+      "grad_norm": 0.47464565333191566,
+      "learning_rate": 7.361522787493763e-07,
+      "loss": 0.2458,
+      "step": 5047
+    },
+    {
+      "epoch": 1.685475792988314,
+      "grad_norm": 0.4540874348894393,
+      "learning_rate": 7.346311940822526e-07,
+      "loss": 0.2438,
+      "step": 5048
+    },
+    {
+      "epoch": 1.6858096828046745,
+      "grad_norm": 0.4658373521445874,
+      "learning_rate": 7.331115579123205e-07,
+      "loss": 0.2431,
+      "step": 5049
+    },
+    {
+      "epoch": 1.686143572621035,
+      "grad_norm": 0.4531834044323251,
+      "learning_rate": 7.315933707556433e-07,
+      "loss": 0.2436,
+      "step": 5050
+    },
+    {
+      "epoch": 1.6864774624373957,
+      "grad_norm": 0.4614241877536614,
+      "learning_rate": 7.300766331277881e-07,
+      "loss": 0.2505,
+      "step": 5051
+    },
+    {
+      "epoch": 1.6868113522537562,
+      "grad_norm": 0.48357111509355233,
+      "learning_rate": 7.285613455438301e-07,
+      "loss": 0.2422,
+      "step": 5052
+    },
+    {
+      "epoch": 1.687145242070117,
+      "grad_norm": 0.476913125845272,
+      "learning_rate": 7.270475085183554e-07,
+      "loss": 0.2524,
+      "step": 5053
+    },
+    {
+      "epoch": 1.6874791318864775,
+      "grad_norm": 0.4718340020136781,
+      "learning_rate": 7.255351225654527e-07,
+      "loss": 0.2448,
+      "step": 5054
+    },
+    {
+      "epoch": 1.687813021702838,
+      "grad_norm": 0.4779094598410625,
+      "learning_rate": 7.240241881987231e-07,
+      "loss": 0.2436,
+      "step": 5055
+    },
+    {
+      "epoch": 1.6881469115191987,
+      "grad_norm": 0.46448034042788494,
+      "learning_rate": 7.225147059312704e-07,
+      "loss": 0.2393,
+      "step": 5056
+    },
+    {
+      "epoch": 1.6884808013355592,
+      "grad_norm": 0.4807911320620965,
+      "learning_rate": 7.210066762757079e-07,
+      "loss": 0.2536,
+      "step": 5057
+    },
+    {
+      "epoch": 1.68881469115192,
+      "grad_norm": 0.4545526362207757,
+      "learning_rate": 7.195000997441532e-07,
+      "loss": 0.2543,
+      "step": 5058
+    },
+    {
+      "epoch": 1.6891485809682805,
+      "grad_norm": 0.4587793014545801,
+      "learning_rate": 7.179949768482337e-07,
+      "loss": 0.2466,
+      "step": 5059
+    },
+    {
+      "epoch": 1.689482470784641,
+      "grad_norm": 0.46481246577979346,
+      "learning_rate": 7.164913080990804e-07,
+      "loss": 0.2437,
+      "step": 5060
+    },
+    {
+      "epoch": 1.6898163606010017,
+      "grad_norm": 0.47341603496910856,
+      "learning_rate": 7.149890940073306e-07,
+      "loss": 0.2498,
+      "step": 5061
+    },
+    {
+      "epoch": 1.6901502504173622,
+      "grad_norm": 0.4862587241239024,
+      "learning_rate": 7.134883350831301e-07,
+      "loss": 0.2563,
+      "step": 5062
+    },
+    {
+      "epoch": 1.690484140233723,
+      "grad_norm": 0.4713070568685044,
+      "learning_rate": 7.119890318361278e-07,
+      "loss": 0.2511,
+      "step": 5063
+    },
+    {
+      "epoch": 1.6908180300500835,
+      "grad_norm": 0.464247827581248,
+      "learning_rate": 7.104911847754803e-07,
+      "loss": 0.2432,
+      "step": 5064
+    },
+    {
+      "epoch": 1.691151919866444,
+      "grad_norm": 0.4642248747361159,
+      "learning_rate": 7.08994794409848e-07,
+      "loss": 0.2405,
+      "step": 5065
+    },
+    {
+      "epoch": 1.6914858096828047,
+      "grad_norm": 0.4796844646422036,
+      "learning_rate": 7.074998612473977e-07,
+      "loss": 0.2434,
+      "step": 5066
+    },
+    {
+      "epoch": 1.6918196994991654,
+      "grad_norm": 0.4685673002137572,
+      "learning_rate": 7.060063857957994e-07,
+      "loss": 0.2484,
+      "step": 5067
+    },
+    {
+      "epoch": 1.692153589315526,
+      "grad_norm": 0.4657586093937777,
+      "learning_rate": 7.045143685622324e-07,
+      "loss": 0.2454,
+      "step": 5068
+    },
+    {
+      "epoch": 1.6924874791318865,
+      "grad_norm": 0.4577791740557612,
+      "learning_rate": 7.030238100533754e-07,
+      "loss": 0.2373,
+      "step": 5069
+    },
+    {
+      "epoch": 1.692821368948247,
+      "grad_norm": 0.4604910094074669,
+      "learning_rate": 7.015347107754161e-07,
+      "loss": 0.2474,
+      "step": 5070
+    },
+    {
+      "epoch": 1.6931552587646077,
+      "grad_norm": 0.47411109458583395,
+      "learning_rate": 7.000470712340451e-07,
+      "loss": 0.2467,
+      "step": 5071
+    },
+    {
+      "epoch": 1.6934891485809684,
+      "grad_norm": 0.4623978020665636,
+      "learning_rate": 6.98560891934455e-07,
+      "loss": 0.2361,
+      "step": 5072
+    },
+    {
+      "epoch": 1.693823038397329,
+      "grad_norm": 0.4752815269012213,
+      "learning_rate": 6.97076173381347e-07,
+      "loss": 0.2559,
+      "step": 5073
+    },
+    {
+      "epoch": 1.6941569282136895,
+      "grad_norm": 0.481125053487027,
+      "learning_rate": 6.955929160789227e-07,
+      "loss": 0.2486,
+      "step": 5074
+    },
+    {
+      "epoch": 1.69449081803005,
+      "grad_norm": 0.45834311584953513,
+      "learning_rate": 6.941111205308893e-07,
+      "loss": 0.2428,
+      "step": 5075
+    },
+    {
+      "epoch": 1.6948247078464107,
+      "grad_norm": 0.47063478341117976,
+      "learning_rate": 6.926307872404548e-07,
+      "loss": 0.2491,
+      "step": 5076
+    },
+    {
+      "epoch": 1.6951585976627714,
+      "grad_norm": 0.47409958637037874,
+      "learning_rate": 6.911519167103353e-07,
+      "loss": 0.2486,
+      "step": 5077
+    },
+    {
+      "epoch": 1.695492487479132,
+      "grad_norm": 0.4751326589804299,
+      "learning_rate": 6.896745094427454e-07,
+      "loss": 0.2547,
+      "step": 5078
+    },
+    {
+      "epoch": 1.6958263772954925,
+      "grad_norm": 0.49220799565759743,
+      "learning_rate": 6.881985659394075e-07,
+      "loss": 0.2427,
+      "step": 5079
+    },
+    {
+      "epoch": 1.696160267111853,
+      "grad_norm": 0.4810568492946589,
+      "learning_rate": 6.867240867015429e-07,
+      "loss": 0.2473,
+      "step": 5080
+    },
+    {
+      "epoch": 1.6964941569282137,
+      "grad_norm": 0.4743272398482338,
+      "learning_rate": 6.852510722298761e-07,
+      "loss": 0.2466,
+      "step": 5081
+    },
+    {
+      "epoch": 1.6968280467445744,
+      "grad_norm": 0.45899029033410926,
+      "learning_rate": 6.837795230246375e-07,
+      "loss": 0.248,
+      "step": 5082
+    },
+    {
+      "epoch": 1.697161936560935,
+      "grad_norm": 0.47392665549330204,
+      "learning_rate": 6.823094395855562e-07,
+      "loss": 0.2544,
+      "step": 5083
+    },
+    {
+      "epoch": 1.6974958263772955,
+      "grad_norm": 0.47499461294135914,
+      "learning_rate": 6.808408224118651e-07,
+      "loss": 0.2515,
+      "step": 5084
+    },
+    {
+      "epoch": 1.697829716193656,
+      "grad_norm": 0.4590856785508998,
+      "learning_rate": 6.793736720022981e-07,
+      "loss": 0.2403,
+      "step": 5085
+    },
+    {
+      "epoch": 1.6981636060100167,
+      "grad_norm": 0.4712260624383991,
+      "learning_rate": 6.779079888550933e-07,
+      "loss": 0.2463,
+      "step": 5086
+    },
+    {
+      "epoch": 1.6984974958263774,
+      "grad_norm": 0.4787817536277574,
+      "learning_rate": 6.764437734679891e-07,
+      "loss": 0.2517,
+      "step": 5087
+    },
+    {
+      "epoch": 1.698831385642738,
+      "grad_norm": 0.46653627431543154,
+      "learning_rate": 6.749810263382256e-07,
+      "loss": 0.2414,
+      "step": 5088
+    },
+    {
+      "epoch": 1.6991652754590985,
+      "grad_norm": 0.46789676115306583,
+      "learning_rate": 6.735197479625433e-07,
+      "loss": 0.2516,
+      "step": 5089
+    },
+    {
+      "epoch": 1.699499165275459,
+      "grad_norm": 0.4558910692585483,
+      "learning_rate": 6.72059938837184e-07,
+      "loss": 0.246,
+      "step": 5090
+    },
+    {
+      "epoch": 1.6998330550918197,
+      "grad_norm": 0.458323532573891,
+      "learning_rate": 6.706015994578935e-07,
+      "loss": 0.2456,
+      "step": 5091
+    },
+    {
+      "epoch": 1.7001669449081804,
+      "grad_norm": 0.46023372198408186,
+      "learning_rate": 6.691447303199156e-07,
+      "loss": 0.2399,
+      "step": 5092
+    },
+    {
+      "epoch": 1.700500834724541,
+      "grad_norm": 0.4826539942865911,
+      "learning_rate": 6.67689331917995e-07,
+      "loss": 0.2516,
+      "step": 5093
+    },
+    {
+      "epoch": 1.7008347245409015,
+      "grad_norm": 0.47336427423035593,
+      "learning_rate": 6.66235404746376e-07,
+      "loss": 0.2577,
+      "step": 5094
+    },
+    {
+      "epoch": 1.701168614357262,
+      "grad_norm": 0.48598206824309725,
+      "learning_rate": 6.647829492988072e-07,
+      "loss": 0.2502,
+      "step": 5095
+    },
+    {
+      "epoch": 1.7015025041736227,
+      "grad_norm": 0.4624798197430895,
+      "learning_rate": 6.633319660685345e-07,
+      "loss": 0.2378,
+      "step": 5096
+    },
+    {
+      "epoch": 1.7018363939899834,
+      "grad_norm": 0.45481008741546364,
+      "learning_rate": 6.618824555483044e-07,
+      "loss": 0.2359,
+      "step": 5097
+    },
+    {
+      "epoch": 1.702170283806344,
+      "grad_norm": 0.48097941006170863,
+      "learning_rate": 6.604344182303629e-07,
+      "loss": 0.2528,
+      "step": 5098
+    },
+    {
+      "epoch": 1.7025041736227045,
+      "grad_norm": 0.46347017909565025,
+      "learning_rate": 6.589878546064544e-07,
+      "loss": 0.2407,
+      "step": 5099
+    },
+    {
+      "epoch": 1.702838063439065,
+      "grad_norm": 0.4590497475299015,
+      "learning_rate": 6.575427651678267e-07,
+      "loss": 0.2469,
+      "step": 5100
+    },
+    {
+      "epoch": 1.7031719532554257,
+      "grad_norm": 0.4875483895827171,
+      "learning_rate": 6.560991504052244e-07,
+      "loss": 0.2524,
+      "step": 5101
+    },
+    {
+      "epoch": 1.7035058430717864,
+      "grad_norm": 0.48274588637055665,
+      "learning_rate": 6.546570108088907e-07,
+      "loss": 0.2575,
+      "step": 5102
+    },
+    {
+      "epoch": 1.703839732888147,
+      "grad_norm": 0.4944165909289753,
+      "learning_rate": 6.532163468685682e-07,
+      "loss": 0.2601,
+      "step": 5103
+    },
+    {
+      "epoch": 1.7041736227045075,
+      "grad_norm": 0.49095277495785855,
+      "learning_rate": 6.517771590734995e-07,
+      "loss": 0.2523,
+      "step": 5104
+    },
+    {
+      "epoch": 1.704507512520868,
+      "grad_norm": 0.4798042493109746,
+      "learning_rate": 6.50339447912427e-07,
+      "loss": 0.2419,
+      "step": 5105
+    },
+    {
+      "epoch": 1.7048414023372287,
+      "grad_norm": 0.4675841749686726,
+      "learning_rate": 6.489032138735884e-07,
+      "loss": 0.2505,
+      "step": 5106
+    },
+    {
+      "epoch": 1.7051752921535894,
+      "grad_norm": 0.4767007810657495,
+      "learning_rate": 6.474684574447221e-07,
+      "loss": 0.2519,
+      "step": 5107
+    },
+    {
+      "epoch": 1.70550918196995,
+      "grad_norm": 0.45964536834447106,
+      "learning_rate": 6.46035179113062e-07,
+      "loss": 0.2501,
+      "step": 5108
+    },
+    {
+      "epoch": 1.7058430717863104,
+      "grad_norm": 0.45915173324899,
+      "learning_rate": 6.446033793653445e-07,
+      "loss": 0.2383,
+      "step": 5109
+    },
+    {
+      "epoch": 1.706176961602671,
+      "grad_norm": 0.4590951727151981,
+      "learning_rate": 6.431730586878004e-07,
+      "loss": 0.2485,
+      "step": 5110
+    },
+    {
+      "epoch": 1.7065108514190317,
+      "grad_norm": 0.4614062429149388,
+      "learning_rate": 6.417442175661598e-07,
+      "loss": 0.2431,
+      "step": 5111
+    },
+    {
+      "epoch": 1.7068447412353924,
+      "grad_norm": 0.4639738148154111,
+      "learning_rate": 6.403168564856472e-07,
+      "loss": 0.2453,
+      "step": 5112
+    },
+    {
+      "epoch": 1.707178631051753,
+      "grad_norm": 0.4647921652265915,
+      "learning_rate": 6.388909759309892e-07,
+      "loss": 0.2568,
+      "step": 5113
+    },
+    {
+      "epoch": 1.7075125208681134,
+      "grad_norm": 0.4740719967711552,
+      "learning_rate": 6.374665763864085e-07,
+      "loss": 0.2432,
+      "step": 5114
+    },
+    {
+      "epoch": 1.707846410684474,
+      "grad_norm": 0.4773259748383645,
+      "learning_rate": 6.360436583356222e-07,
+      "loss": 0.2522,
+      "step": 5115
+    },
+    {
+      "epoch": 1.7081803005008347,
+      "grad_norm": 0.47870890365355934,
+      "learning_rate": 6.34622222261847e-07,
+      "loss": 0.2484,
+      "step": 5116
+    },
+    {
+      "epoch": 1.7085141903171954,
+      "grad_norm": 0.47944732899301046,
+      "learning_rate": 6.332022686477929e-07,
+      "loss": 0.2428,
+      "step": 5117
+    },
+    {
+      "epoch": 1.708848080133556,
+      "grad_norm": 0.43497630726524705,
+      "learning_rate": 6.317837979756713e-07,
+      "loss": 0.2276,
+      "step": 5118
+    },
+    {
+      "epoch": 1.7091819699499164,
+      "grad_norm": 0.4906468157106463,
+      "learning_rate": 6.303668107271871e-07,
+      "loss": 0.2488,
+      "step": 5119
+    },
+    {
+      "epoch": 1.7095158597662772,
+      "grad_norm": 0.4655087262435791,
+      "learning_rate": 6.289513073835396e-07,
+      "loss": 0.2433,
+      "step": 5120
+    },
+    {
+      "epoch": 1.7098497495826377,
+      "grad_norm": 0.44461023355328316,
+      "learning_rate": 6.275372884254299e-07,
+      "loss": 0.2298,
+      "step": 5121
+    },
+    {
+      "epoch": 1.7101836393989984,
+      "grad_norm": 0.4631463973561414,
+      "learning_rate": 6.261247543330484e-07,
+      "loss": 0.2555,
+      "step": 5122
+    },
+    {
+      "epoch": 1.710517529215359,
+      "grad_norm": 0.446916190225191,
+      "learning_rate": 6.247137055860869e-07,
+      "loss": 0.2404,
+      "step": 5123
+    },
+    {
+      "epoch": 1.7108514190317194,
+      "grad_norm": 0.4789603843477625,
+      "learning_rate": 6.233041426637293e-07,
+      "loss": 0.2471,
+      "step": 5124
+    },
+    {
+      "epoch": 1.7111853088480802,
+      "grad_norm": 0.47064010986116683,
+      "learning_rate": 6.218960660446554e-07,
+      "loss": 0.245,
+      "step": 5125
+    },
+    {
+      "epoch": 1.7115191986644407,
+      "grad_norm": 0.4736770717787985,
+      "learning_rate": 6.204894762070407e-07,
+      "loss": 0.2464,
+      "step": 5126
+    },
+    {
+      "epoch": 1.7118530884808014,
+      "grad_norm": 0.45842290078100617,
+      "learning_rate": 6.190843736285568e-07,
+      "loss": 0.2533,
+      "step": 5127
+    },
+    {
+      "epoch": 1.712186978297162,
+      "grad_norm": 0.46539797453915754,
+      "learning_rate": 6.176807587863681e-07,
+      "loss": 0.239,
+      "step": 5128
+    },
+    {
+      "epoch": 1.7125208681135224,
+      "grad_norm": 0.45333186003620185,
+      "learning_rate": 6.162786321571351e-07,
+      "loss": 0.2472,
+      "step": 5129
+    },
+    {
+      "epoch": 1.7128547579298832,
+      "grad_norm": 0.4622336378782702,
+      "learning_rate": 6.148779942170141e-07,
+      "loss": 0.2542,
+      "step": 5130
+    },
+    {
+      "epoch": 1.7131886477462437,
+      "grad_norm": 0.4512486835377363,
+      "learning_rate": 6.134788454416529e-07,
+      "loss": 0.2466,
+      "step": 5131
+    },
+    {
+      "epoch": 1.7135225375626044,
+      "grad_norm": 0.4806201795460655,
+      "learning_rate": 6.120811863061965e-07,
+      "loss": 0.257,
+      "step": 5132
+    },
+    {
+      "epoch": 1.713856427378965,
+      "grad_norm": 0.4508574760475429,
+      "learning_rate": 6.106850172852819e-07,
+      "loss": 0.243,
+      "step": 5133
+    },
+    {
+      "epoch": 1.7141903171953254,
+      "grad_norm": 0.4746530905484952,
+      "learning_rate": 6.092903388530413e-07,
+      "loss": 0.2334,
+      "step": 5134
+    },
+    {
+      "epoch": 1.7145242070116862,
+      "grad_norm": 0.4725881857890721,
+      "learning_rate": 6.078971514830989e-07,
+      "loss": 0.2517,
+      "step": 5135
+    },
+    {
+      "epoch": 1.714858096828047,
+      "grad_norm": 0.46881609264102736,
+      "learning_rate": 6.065054556485761e-07,
+      "loss": 0.2375,
+      "step": 5136
+    },
+    {
+      "epoch": 1.7151919866444074,
+      "grad_norm": 0.4687686260841863,
+      "learning_rate": 6.051152518220838e-07,
+      "loss": 0.2431,
+      "step": 5137
+    },
+    {
+      "epoch": 1.715525876460768,
+      "grad_norm": 0.45672726441781436,
+      "learning_rate": 6.037265404757292e-07,
+      "loss": 0.2521,
+      "step": 5138
+    },
+    {
+      "epoch": 1.7158597662771284,
+      "grad_norm": 0.4819128737865506,
+      "learning_rate": 6.023393220811119e-07,
+      "loss": 0.257,
+      "step": 5139
+    },
+    {
+      "epoch": 1.7161936560934892,
+      "grad_norm": 0.469627234913836,
+      "learning_rate": 6.009535971093216e-07,
+      "loss": 0.2421,
+      "step": 5140
+    },
+    {
+      "epoch": 1.71652754590985,
+      "grad_norm": 0.47390076817464644,
+      "learning_rate": 5.995693660309466e-07,
+      "loss": 0.2561,
+      "step": 5141
+    },
+    {
+      "epoch": 1.7168614357262104,
+      "grad_norm": 0.46219873731185324,
+      "learning_rate": 5.981866293160627e-07,
+      "loss": 0.2429,
+      "step": 5142
+    },
+    {
+      "epoch": 1.717195325542571,
+      "grad_norm": 0.4824463706348088,
+      "learning_rate": 5.968053874342405e-07,
+      "loss": 0.2447,
+      "step": 5143
+    },
+    {
+      "epoch": 1.7175292153589314,
+      "grad_norm": 0.48300442875760274,
+      "learning_rate": 5.954256408545417e-07,
+      "loss": 0.2432,
+      "step": 5144
+    },
+    {
+      "epoch": 1.7178631051752922,
+      "grad_norm": 0.4655707995544866,
+      "learning_rate": 5.94047390045523e-07,
+      "loss": 0.2506,
+      "step": 5145
+    },
+    {
+      "epoch": 1.718196994991653,
+      "grad_norm": 0.4780844220364181,
+      "learning_rate": 5.926706354752293e-07,
+      "loss": 0.259,
+      "step": 5146
+    },
+    {
+      "epoch": 1.7185308848080134,
+      "grad_norm": 0.4656418197374167,
+      "learning_rate": 5.912953776112007e-07,
+      "loss": 0.2411,
+      "step": 5147
+    },
+    {
+      "epoch": 1.718864774624374,
+      "grad_norm": 0.4891358113641871,
+      "learning_rate": 5.89921616920468e-07,
+      "loss": 0.2485,
+      "step": 5148
+    },
+    {
+      "epoch": 1.7191986644407344,
+      "grad_norm": 0.48197037058347314,
+      "learning_rate": 5.885493538695508e-07,
+      "loss": 0.2456,
+      "step": 5149
+    },
+    {
+      "epoch": 1.7195325542570952,
+      "grad_norm": 0.4573659995512331,
+      "learning_rate": 5.871785889244657e-07,
+      "loss": 0.2449,
+      "step": 5150
+    },
+    {
+      "epoch": 1.719866444073456,
+      "grad_norm": 0.4594844032144234,
+      "learning_rate": 5.85809322550716e-07,
+      "loss": 0.2457,
+      "step": 5151
+    },
+    {
+      "epoch": 1.7202003338898164,
+      "grad_norm": 0.47107177768848274,
+      "learning_rate": 5.84441555213297e-07,
+      "loss": 0.2577,
+      "step": 5152
+    },
+    {
+      "epoch": 1.720534223706177,
+      "grad_norm": 0.468428475507681,
+      "learning_rate": 5.830752873766948e-07,
+      "loss": 0.2476,
+      "step": 5153
+    },
+    {
+      "epoch": 1.7208681135225374,
+      "grad_norm": 0.4567071756907123,
+      "learning_rate": 5.817105195048878e-07,
+      "loss": 0.2414,
+      "step": 5154
+    },
+    {
+      "epoch": 1.7212020033388982,
+      "grad_norm": 0.4655145968146217,
+      "learning_rate": 5.803472520613457e-07,
+      "loss": 0.2442,
+      "step": 5155
+    },
+    {
+      "epoch": 1.721535893155259,
+      "grad_norm": 0.48278429888227964,
+      "learning_rate": 5.789854855090249e-07,
+      "loss": 0.2409,
+      "step": 5156
+    },
+    {
+      "epoch": 1.7218697829716194,
+      "grad_norm": 0.46836483718059574,
+      "learning_rate": 5.776252203103755e-07,
+      "loss": 0.2435,
+      "step": 5157
+    },
+    {
+      "epoch": 1.72220367278798,
+      "grad_norm": 0.4713704476708414,
+      "learning_rate": 5.762664569273346e-07,
+      "loss": 0.2465,
+      "step": 5158
+    },
+    {
+      "epoch": 1.7225375626043404,
+      "grad_norm": 0.4408222004112837,
+      "learning_rate": 5.749091958213338e-07,
+      "loss": 0.2329,
+      "step": 5159
+    },
+    {
+      "epoch": 1.7228714524207012,
+      "grad_norm": 0.4746791484447419,
+      "learning_rate": 5.735534374532908e-07,
+      "loss": 0.2427,
+      "step": 5160
+    },
+    {
+      "epoch": 1.723205342237062,
+      "grad_norm": 0.4726554004973202,
+      "learning_rate": 5.721991822836137e-07,
+      "loss": 0.2471,
+      "step": 5161
+    },
+    {
+      "epoch": 1.7235392320534224,
+      "grad_norm": 0.48783625263040453,
+      "learning_rate": 5.708464307722006e-07,
+      "loss": 0.2514,
+      "step": 5162
+    },
+    {
+      "epoch": 1.723873121869783,
+      "grad_norm": 0.4574040512418704,
+      "learning_rate": 5.694951833784384e-07,
+      "loss": 0.238,
+      "step": 5163
+    },
+    {
+      "epoch": 1.7242070116861434,
+      "grad_norm": 0.4709131745192135,
+      "learning_rate": 5.681454405612063e-07,
+      "loss": 0.2476,
+      "step": 5164
+    },
+    {
+      "epoch": 1.7245409015025042,
+      "grad_norm": 0.4653766627564802,
+      "learning_rate": 5.667972027788682e-07,
+      "loss": 0.2442,
+      "step": 5165
+    },
+    {
+      "epoch": 1.724874791318865,
+      "grad_norm": 0.4691292160034917,
+      "learning_rate": 5.654504704892782e-07,
+      "loss": 0.2497,
+      "step": 5166
+    },
+    {
+      "epoch": 1.7252086811352254,
+      "grad_norm": 0.47584677789590635,
+      "learning_rate": 5.641052441497797e-07,
+      "loss": 0.2525,
+      "step": 5167
+    },
+    {
+      "epoch": 1.725542570951586,
+      "grad_norm": 0.46554622992509365,
+      "learning_rate": 5.627615242172057e-07,
+      "loss": 0.2446,
+      "step": 5168
+    },
+    {
+      "epoch": 1.7258764607679464,
+      "grad_norm": 0.4738080222082383,
+      "learning_rate": 5.614193111478761e-07,
+      "loss": 0.2551,
+      "step": 5169
+    },
+    {
+      "epoch": 1.7262103505843072,
+      "grad_norm": 0.4540355986542076,
+      "learning_rate": 5.600786053975993e-07,
+      "loss": 0.2509,
+      "step": 5170
+    },
+    {
+      "epoch": 1.726544240400668,
+      "grad_norm": 0.4703318494957613,
+      "learning_rate": 5.587394074216712e-07,
+      "loss": 0.246,
+      "step": 5171
+    },
+    {
+      "epoch": 1.7268781302170284,
+      "grad_norm": 0.47443302495857304,
+      "learning_rate": 5.574017176748769e-07,
+      "loss": 0.2611,
+      "step": 5172
+    },
+    {
+      "epoch": 1.727212020033389,
+      "grad_norm": 0.47197636936050813,
+      "learning_rate": 5.560655366114909e-07,
+      "loss": 0.2452,
+      "step": 5173
+    },
+    {
+      "epoch": 1.7275459098497494,
+      "grad_norm": 0.4586604864950804,
+      "learning_rate": 5.547308646852717e-07,
+      "loss": 0.2277,
+      "step": 5174
+    },
+    {
+      "epoch": 1.7278797996661102,
+      "grad_norm": 0.46804035258564103,
+      "learning_rate": 5.533977023494675e-07,
+      "loss": 0.2343,
+      "step": 5175
+    },
+    {
+      "epoch": 1.728213689482471,
+      "grad_norm": 0.4602559191985946,
+      "learning_rate": 5.520660500568115e-07,
+      "loss": 0.2482,
+      "step": 5176
+    },
+    {
+      "epoch": 1.7285475792988314,
+      "grad_norm": 0.5028121018189675,
+      "learning_rate": 5.507359082595287e-07,
+      "loss": 0.2534,
+      "step": 5177
+    },
+    {
+      "epoch": 1.728881469115192,
+      "grad_norm": 0.4568506256351656,
+      "learning_rate": 5.494072774093273e-07,
+      "loss": 0.2396,
+      "step": 5178
+    },
+    {
+      "epoch": 1.7292153589315524,
+      "grad_norm": 0.4851305423013494,
+      "learning_rate": 5.48080157957403e-07,
+      "loss": 0.2571,
+      "step": 5179
+    },
+    {
+      "epoch": 1.7295492487479132,
+      "grad_norm": 0.48504745566983376,
+      "learning_rate": 5.467545503544381e-07,
+      "loss": 0.2414,
+      "step": 5180
+    },
+    {
+      "epoch": 1.729883138564274,
+      "grad_norm": 0.46223180775857486,
+      "learning_rate": 5.454304550506029e-07,
+      "loss": 0.2467,
+      "step": 5181
+    },
+    {
+      "epoch": 1.7302170283806344,
+      "grad_norm": 0.4575434343930806,
+      "learning_rate": 5.441078724955545e-07,
+      "loss": 0.2414,
+      "step": 5182
+    },
+    {
+      "epoch": 1.730550918196995,
+      "grad_norm": 0.46209764682806376,
+      "learning_rate": 5.427868031384342e-07,
+      "loss": 0.2426,
+      "step": 5183
+    },
+    {
+      "epoch": 1.7308848080133556,
+      "grad_norm": 0.47931239179313984,
+      "learning_rate": 5.414672474278698e-07,
+      "loss": 0.2529,
+      "step": 5184
+    },
+    {
+      "epoch": 1.7312186978297162,
+      "grad_norm": 0.4762864184462101,
+      "learning_rate": 5.401492058119751e-07,
+      "loss": 0.2353,
+      "step": 5185
+    },
+    {
+      "epoch": 1.731552587646077,
+      "grad_norm": 0.4793504999419486,
+      "learning_rate": 5.388326787383518e-07,
+      "loss": 0.2538,
+      "step": 5186
+    },
+    {
+      "epoch": 1.7318864774624374,
+      "grad_norm": 0.47083212462574786,
+      "learning_rate": 5.375176666540848e-07,
+      "loss": 0.2403,
+      "step": 5187
+    },
+    {
+      "epoch": 1.732220367278798,
+      "grad_norm": 0.4595181956068547,
+      "learning_rate": 5.362041700057452e-07,
+      "loss": 0.2361,
+      "step": 5188
+    },
+    {
+      "epoch": 1.7325542570951586,
+      "grad_norm": 0.5062853414607508,
+      "learning_rate": 5.348921892393905e-07,
+      "loss": 0.2529,
+      "step": 5189
+    },
+    {
+      "epoch": 1.7328881469115192,
+      "grad_norm": 0.4757317328621939,
+      "learning_rate": 5.33581724800561e-07,
+      "loss": 0.26,
+      "step": 5190
+    },
+    {
+      "epoch": 1.7332220367278799,
+      "grad_norm": 0.48700345834997805,
+      "learning_rate": 5.322727771342856e-07,
+      "loss": 0.2507,
+      "step": 5191
+    },
+    {
+      "epoch": 1.7335559265442404,
+      "grad_norm": 0.46664249379081946,
+      "learning_rate": 5.309653466850756e-07,
+      "loss": 0.2547,
+      "step": 5192
+    },
+    {
+      "epoch": 1.733889816360601,
+      "grad_norm": 0.466380118775154,
+      "learning_rate": 5.296594338969274e-07,
+      "loss": 0.2504,
+      "step": 5193
+    },
+    {
+      "epoch": 1.7342237061769616,
+      "grad_norm": 0.48328866878269416,
+      "learning_rate": 5.283550392133208e-07,
+      "loss": 0.2532,
+      "step": 5194
+    },
+    {
+      "epoch": 1.7345575959933222,
+      "grad_norm": 0.4818095901866241,
+      "learning_rate": 5.270521630772241e-07,
+      "loss": 0.2533,
+      "step": 5195
+    },
+    {
+      "epoch": 1.7348914858096829,
+      "grad_norm": 0.454323109038423,
+      "learning_rate": 5.257508059310862e-07,
+      "loss": 0.2356,
+      "step": 5196
+    },
+    {
+      "epoch": 1.7352253756260434,
+      "grad_norm": 0.47844086846561446,
+      "learning_rate": 5.244509682168397e-07,
+      "loss": 0.2447,
+      "step": 5197
+    },
+    {
+      "epoch": 1.735559265442404,
+      "grad_norm": 0.447013819571352,
+      "learning_rate": 5.231526503759055e-07,
+      "loss": 0.2386,
+      "step": 5198
+    },
+    {
+      "epoch": 1.7358931552587646,
+      "grad_norm": 0.4574325268062427,
+      "learning_rate": 5.218558528491829e-07,
+      "loss": 0.2368,
+      "step": 5199
+    },
+    {
+      "epoch": 1.7362270450751254,
+      "grad_norm": 0.49793220335885113,
+      "learning_rate": 5.205605760770605e-07,
+      "loss": 0.2468,
+      "step": 5200
+    },
+    {
+      "epoch": 1.7365609348914859,
+      "grad_norm": 0.4673252320573378,
+      "learning_rate": 5.192668204994056e-07,
+      "loss": 0.2485,
+      "step": 5201
+    },
+    {
+      "epoch": 1.7368948247078464,
+      "grad_norm": 0.4701210451679904,
+      "learning_rate": 5.179745865555713e-07,
+      "loss": 0.2347,
+      "step": 5202
+    },
+    {
+      "epoch": 1.737228714524207,
+      "grad_norm": 0.48491957440551686,
+      "learning_rate": 5.166838746843933e-07,
+      "loss": 0.2499,
+      "step": 5203
+    },
+    {
+      "epoch": 1.7375626043405676,
+      "grad_norm": 0.47742353741201,
+      "learning_rate": 5.153946853241914e-07,
+      "loss": 0.2516,
+      "step": 5204
+    },
+    {
+      "epoch": 1.7378964941569284,
+      "grad_norm": 0.4551317118174587,
+      "learning_rate": 5.141070189127668e-07,
+      "loss": 0.2353,
+      "step": 5205
+    },
+    {
+      "epoch": 1.7382303839732889,
+      "grad_norm": 0.48401364854188733,
+      "learning_rate": 5.128208758874059e-07,
+      "loss": 0.2515,
+      "step": 5206
+    },
+    {
+      "epoch": 1.7385642737896494,
+      "grad_norm": 0.47995535592231486,
+      "learning_rate": 5.115362566848747e-07,
+      "loss": 0.2477,
+      "step": 5207
+    },
+    {
+      "epoch": 1.73889816360601,
+      "grad_norm": 0.4702627745884613,
+      "learning_rate": 5.10253161741423e-07,
+      "loss": 0.2533,
+      "step": 5208
+    },
+    {
+      "epoch": 1.7392320534223706,
+      "grad_norm": 0.4908909819625363,
+      "learning_rate": 5.089715914927851e-07,
+      "loss": 0.2561,
+      "step": 5209
+    },
+    {
+      "epoch": 1.7395659432387314,
+      "grad_norm": 0.48122970061150555,
+      "learning_rate": 5.076915463741744e-07,
+      "loss": 0.2486,
+      "step": 5210
+    },
+    {
+      "epoch": 1.7398998330550919,
+      "grad_norm": 0.4642105405718951,
+      "learning_rate": 5.064130268202877e-07,
+      "loss": 0.2427,
+      "step": 5211
+    },
+    {
+      "epoch": 1.7402337228714524,
+      "grad_norm": 0.4994472777090362,
+      "learning_rate": 5.051360332653032e-07,
+      "loss": 0.2544,
+      "step": 5212
+    },
+    {
+      "epoch": 1.740567612687813,
+      "grad_norm": 0.4551400708965084,
+      "learning_rate": 5.038605661428809e-07,
+      "loss": 0.2561,
+      "step": 5213
+    },
+    {
+      "epoch": 1.7409015025041736,
+      "grad_norm": 0.4563695169169142,
+      "learning_rate": 5.025866258861639e-07,
+      "loss": 0.2358,
+      "step": 5214
+    },
+    {
+      "epoch": 1.7412353923205344,
+      "grad_norm": 0.4761146668552975,
+      "learning_rate": 5.013142129277759e-07,
+      "loss": 0.2487,
+      "step": 5215
+    },
+    {
+      "epoch": 1.7415692821368949,
+      "grad_norm": 0.4801413060644081,
+      "learning_rate": 5.000433276998218e-07,
+      "loss": 0.2544,
+      "step": 5216
+    },
+    {
+      "epoch": 1.7419031719532554,
+      "grad_norm": 0.4487951735332803,
+      "learning_rate": 4.987739706338862e-07,
+      "loss": 0.2464,
+      "step": 5217
+    },
+    {
+      "epoch": 1.742237061769616,
+      "grad_norm": 0.4701453773493022,
+      "learning_rate": 4.975061421610377e-07,
+      "loss": 0.2493,
+      "step": 5218
+    },
+    {
+      "epoch": 1.7425709515859766,
+      "grad_norm": 0.49847527747365056,
+      "learning_rate": 4.962398427118243e-07,
+      "loss": 0.2611,
+      "step": 5219
+    },
+    {
+      "epoch": 1.7429048414023374,
+      "grad_norm": 0.47660782529531226,
+      "learning_rate": 4.949750727162738e-07,
+      "loss": 0.2466,
+      "step": 5220
+    },
+    {
+      "epoch": 1.7432387312186979,
+      "grad_norm": 0.46486323430757837,
+      "learning_rate": 4.937118326038964e-07,
+      "loss": 0.2464,
+      "step": 5221
+    },
+    {
+      "epoch": 1.7435726210350584,
+      "grad_norm": 0.4723069042160636,
+      "learning_rate": 4.924501228036793e-07,
+      "loss": 0.2462,
+      "step": 5222
+    },
+    {
+      "epoch": 1.743906510851419,
+      "grad_norm": 0.4570261060867376,
+      "learning_rate": 4.911899437440964e-07,
+      "loss": 0.2379,
+      "step": 5223
+    },
+    {
+      "epoch": 1.7442404006677796,
+      "grad_norm": 0.4531059668740549,
+      "learning_rate": 4.899312958530966e-07,
+      "loss": 0.2469,
+      "step": 5224
+    },
+    {
+      "epoch": 1.7445742904841404,
+      "grad_norm": 0.4614053775035508,
+      "learning_rate": 4.886741795581101e-07,
+      "loss": 0.2498,
+      "step": 5225
+    },
+    {
+      "epoch": 1.7449081803005009,
+      "grad_norm": 0.45329729860927737,
+      "learning_rate": 4.874185952860471e-07,
+      "loss": 0.2511,
+      "step": 5226
+    },
+    {
+      "epoch": 1.7452420701168614,
+      "grad_norm": 0.46224246940822494,
+      "learning_rate": 4.861645434632967e-07,
+      "loss": 0.2418,
+      "step": 5227
+    },
+    {
+      "epoch": 1.745575959933222,
+      "grad_norm": 0.47237876172862037,
+      "learning_rate": 4.849120245157296e-07,
+      "loss": 0.2454,
+      "step": 5228
+    },
+    {
+      "epoch": 1.7459098497495826,
+      "grad_norm": 0.4755230865829938,
+      "learning_rate": 4.836610388686947e-07,
+      "loss": 0.2558,
+      "step": 5229
+    },
+    {
+      "epoch": 1.7462437395659434,
+      "grad_norm": 0.4925178233300828,
+      "learning_rate": 4.824115869470197e-07,
+      "loss": 0.2464,
+      "step": 5230
+    },
+    {
+      "epoch": 1.7465776293823039,
+      "grad_norm": 0.4841310710942091,
+      "learning_rate": 4.81163669175011e-07,
+      "loss": 0.2568,
+      "step": 5231
+    },
+    {
+      "epoch": 1.7469115191986644,
+      "grad_norm": 0.47063056471568193,
+      "learning_rate": 4.799172859764567e-07,
+      "loss": 0.2408,
+      "step": 5232
+    },
+    {
+      "epoch": 1.747245409015025,
+      "grad_norm": 0.46542695096625053,
+      "learning_rate": 4.786724377746221e-07,
+      "loss": 0.2532,
+      "step": 5233
+    },
+    {
+      "epoch": 1.7475792988313856,
+      "grad_norm": 0.458557840532002,
+      "learning_rate": 4.774291249922508e-07,
+      "loss": 0.2461,
+      "step": 5234
+    },
+    {
+      "epoch": 1.7479131886477464,
+      "grad_norm": 0.452272109947127,
+      "learning_rate": 4.76187348051565e-07,
+      "loss": 0.2377,
+      "step": 5235
+    },
+    {
+      "epoch": 1.7482470784641069,
+      "grad_norm": 0.4582185713904869,
+      "learning_rate": 4.749471073742656e-07,
+      "loss": 0.2548,
+      "step": 5236
+    },
+    {
+      "epoch": 1.7485809682804674,
+      "grad_norm": 0.4796352820570309,
+      "learning_rate": 4.7370840338153313e-07,
+      "loss": 0.2433,
+      "step": 5237
+    },
+    {
+      "epoch": 1.748914858096828,
+      "grad_norm": 0.45501055596484163,
+      "learning_rate": 4.7247123649402427e-07,
+      "loss": 0.2385,
+      "step": 5238
+    },
+    {
+      "epoch": 1.7492487479131886,
+      "grad_norm": 0.45214973518448603,
+      "learning_rate": 4.7123560713187353e-07,
+      "loss": 0.2427,
+      "step": 5239
+    },
+    {
+      "epoch": 1.7495826377295494,
+      "grad_norm": 0.4467523398442794,
+      "learning_rate": 4.700015157146959e-07,
+      "loss": 0.2567,
+      "step": 5240
+    },
+    {
+      "epoch": 1.7499165275459099,
+      "grad_norm": 0.49056506937439087,
+      "learning_rate": 4.68768962661581e-07,
+      "loss": 0.2531,
+      "step": 5241
+    },
+    {
+      "epoch": 1.7502504173622704,
+      "grad_norm": 0.4585927043170065,
+      "learning_rate": 4.6753794839109966e-07,
+      "loss": 0.2395,
+      "step": 5242
+    },
+    {
+      "epoch": 1.750584307178631,
+      "grad_norm": 0.47068832606834254,
+      "learning_rate": 4.663084733212958e-07,
+      "loss": 0.2492,
+      "step": 5243
+    },
+    {
+      "epoch": 1.7509181969949916,
+      "grad_norm": 0.4563173909457415,
+      "learning_rate": 4.6508053786969375e-07,
+      "loss": 0.2435,
+      "step": 5244
+    },
+    {
+      "epoch": 1.7512520868113524,
+      "grad_norm": 0.4986658002128787,
+      "learning_rate": 4.638541424532933e-07,
+      "loss": 0.249,
+      "step": 5245
+    },
+    {
+      "epoch": 1.7515859766277129,
+      "grad_norm": 0.47010988536975645,
+      "learning_rate": 4.6262928748857296e-07,
+      "loss": 0.251,
+      "step": 5246
+    },
+    {
+      "epoch": 1.7519198664440734,
+      "grad_norm": 0.47543900841206255,
+      "learning_rate": 4.614059733914872e-07,
+      "loss": 0.2584,
+      "step": 5247
+    },
+    {
+      "epoch": 1.752253756260434,
+      "grad_norm": 0.45999642510259175,
+      "learning_rate": 4.6018420057746537e-07,
+      "loss": 0.2381,
+      "step": 5248
+    },
+    {
+      "epoch": 1.7525876460767946,
+      "grad_norm": 0.45196869698752834,
+      "learning_rate": 4.5896396946141784e-07,
+      "loss": 0.2359,
+      "step": 5249
+    },
+    {
+      "epoch": 1.7529215358931554,
+      "grad_norm": 0.4780765735493111,
+      "learning_rate": 4.577452804577259e-07,
+      "loss": 0.2503,
+      "step": 5250
+    },
+    {
+      "epoch": 1.7532554257095159,
+      "grad_norm": 0.4776739515146518,
+      "learning_rate": 4.56528133980253e-07,
+      "loss": 0.2403,
+      "step": 5251
+    },
+    {
+      "epoch": 1.7535893155258764,
+      "grad_norm": 0.48336354355711647,
+      "learning_rate": 4.5531253044233393e-07,
+      "loss": 0.2508,
+      "step": 5252
+    },
+    {
+      "epoch": 1.7539232053422371,
+      "grad_norm": 0.4540144725297971,
+      "learning_rate": 4.5409847025678246e-07,
+      "loss": 0.2419,
+      "step": 5253
+    },
+    {
+      "epoch": 1.7542570951585976,
+      "grad_norm": 0.4616288289904927,
+      "learning_rate": 4.5288595383588484e-07,
+      "loss": 0.2441,
+      "step": 5254
+    },
+    {
+      "epoch": 1.7545909849749584,
+      "grad_norm": 0.4703261518758878,
+      "learning_rate": 4.5167498159140845e-07,
+      "loss": 0.2523,
+      "step": 5255
+    },
+    {
+      "epoch": 1.7549248747913189,
+      "grad_norm": 0.4574509558824663,
+      "learning_rate": 4.504655539345909e-07,
+      "loss": 0.2389,
+      "step": 5256
+    },
+    {
+      "epoch": 1.7552587646076794,
+      "grad_norm": 0.4756939311512544,
+      "learning_rate": 4.4925767127614927e-07,
+      "loss": 0.2556,
+      "step": 5257
+    },
+    {
+      "epoch": 1.7555926544240401,
+      "grad_norm": 0.45464040120583554,
+      "learning_rate": 4.4805133402627323e-07,
+      "loss": 0.2388,
+      "step": 5258
+    },
+    {
+      "epoch": 1.7559265442404006,
+      "grad_norm": 0.46964497555761753,
+      "learning_rate": 4.4684654259462836e-07,
+      "loss": 0.2359,
+      "step": 5259
+    },
+    {
+      "epoch": 1.7562604340567614,
+      "grad_norm": 0.4463986307419851,
+      "learning_rate": 4.4564329739035684e-07,
+      "loss": 0.2415,
+      "step": 5260
+    },
+    {
+      "epoch": 1.7565943238731219,
+      "grad_norm": 0.4637000231249989,
+      "learning_rate": 4.4444159882207406e-07,
+      "loss": 0.2391,
+      "step": 5261
+    },
+    {
+      "epoch": 1.7569282136894824,
+      "grad_norm": 0.4829351877510723,
+      "learning_rate": 4.432414472978702e-07,
+      "loss": 0.2569,
+      "step": 5262
+    },
+    {
+      "epoch": 1.757262103505843,
+      "grad_norm": 0.46324769176740865,
+      "learning_rate": 4.4204284322530933e-07,
+      "loss": 0.2496,
+      "step": 5263
+    },
+    {
+      "epoch": 1.7575959933222036,
+      "grad_norm": 0.4871122037408097,
+      "learning_rate": 4.408457870114341e-07,
+      "loss": 0.2488,
+      "step": 5264
+    },
+    {
+      "epoch": 1.7579298831385644,
+      "grad_norm": 0.47114010838127535,
+      "learning_rate": 4.3965027906275604e-07,
+      "loss": 0.2358,
+      "step": 5265
+    },
+    {
+      "epoch": 1.7582637729549249,
+      "grad_norm": 0.4643372040945664,
+      "learning_rate": 4.3845631978526536e-07,
+      "loss": 0.2408,
+      "step": 5266
+    },
+    {
+      "epoch": 1.7585976627712854,
+      "grad_norm": 0.45483570237475807,
+      "learning_rate": 4.3726390958442386e-07,
+      "loss": 0.2464,
+      "step": 5267
+    },
+    {
+      "epoch": 1.758931552587646,
+      "grad_norm": 0.48446646065548343,
+      "learning_rate": 4.3607304886516655e-07,
+      "loss": 0.2466,
+      "step": 5268
+    },
+    {
+      "epoch": 1.7592654424040068,
+      "grad_norm": 0.4607223618134017,
+      "learning_rate": 4.3488373803190545e-07,
+      "loss": 0.2452,
+      "step": 5269
+    },
+    {
+      "epoch": 1.7595993322203674,
+      "grad_norm": 0.4697251136955839,
+      "learning_rate": 4.336959774885241e-07,
+      "loss": 0.2507,
+      "step": 5270
+    },
+    {
+      "epoch": 1.7599332220367279,
+      "grad_norm": 0.4839587590976735,
+      "learning_rate": 4.325097676383788e-07,
+      "loss": 0.2544,
+      "step": 5271
+    },
+    {
+      "epoch": 1.7602671118530884,
+      "grad_norm": 0.4799059891120207,
+      "learning_rate": 4.3132510888430015e-07,
+      "loss": 0.2461,
+      "step": 5272
+    },
+    {
+      "epoch": 1.760601001669449,
+      "grad_norm": 0.4905624034445771,
+      "learning_rate": 4.3014200162859233e-07,
+      "loss": 0.249,
+      "step": 5273
+    },
+    {
+      "epoch": 1.7609348914858098,
+      "grad_norm": 0.470515676043973,
+      "learning_rate": 4.28960446273034e-07,
+      "loss": 0.2453,
+      "step": 5274
+    },
+    {
+      "epoch": 1.7612687813021703,
+      "grad_norm": 0.47757281551535025,
+      "learning_rate": 4.277804432188748e-07,
+      "loss": 0.2425,
+      "step": 5275
+    },
+    {
+      "epoch": 1.7616026711185309,
+      "grad_norm": 0.48278002901485717,
+      "learning_rate": 4.2660199286683636e-07,
+      "loss": 0.2499,
+      "step": 5276
+    },
+    {
+      "epoch": 1.7619365609348914,
+      "grad_norm": 0.4552197700450515,
+      "learning_rate": 4.2542509561711407e-07,
+      "loss": 0.2455,
+      "step": 5277
+    },
+    {
+      "epoch": 1.762270450751252,
+      "grad_norm": 0.4806491895082474,
+      "learning_rate": 4.242497518693783e-07,
+      "loss": 0.2528,
+      "step": 5278
+    },
+    {
+      "epoch": 1.7626043405676128,
+      "grad_norm": 0.49330127715296523,
+      "learning_rate": 4.230759620227681e-07,
+      "loss": 0.2611,
+      "step": 5279
+    },
+    {
+      "epoch": 1.7629382303839733,
+      "grad_norm": 0.480604123117753,
+      "learning_rate": 4.2190372647589736e-07,
+      "loss": 0.251,
+      "step": 5280
+    },
+    {
+      "epoch": 1.7632721202003339,
+      "grad_norm": 0.4828839711660939,
+      "learning_rate": 4.2073304562684934e-07,
+      "loss": 0.2473,
+      "step": 5281
+    },
+    {
+      "epoch": 1.7636060100166944,
+      "grad_norm": 0.48353026734716814,
+      "learning_rate": 4.1956391987318266e-07,
+      "loss": 0.2514,
+      "step": 5282
+    },
+    {
+      "epoch": 1.763939899833055,
+      "grad_norm": 0.46071954312446023,
+      "learning_rate": 4.1839634961192697e-07,
+      "loss": 0.2446,
+      "step": 5283
+    },
+    {
+      "epoch": 1.7642737896494158,
+      "grad_norm": 0.48891045199884453,
+      "learning_rate": 4.1723033523958235e-07,
+      "loss": 0.2436,
+      "step": 5284
+    },
+    {
+      "epoch": 1.7646076794657763,
+      "grad_norm": 0.48948905839377516,
+      "learning_rate": 4.160658771521214e-07,
+      "loss": 0.2519,
+      "step": 5285
+    },
+    {
+      "epoch": 1.7649415692821369,
+      "grad_norm": 0.4576753229489727,
+      "learning_rate": 4.149029757449874e-07,
+      "loss": 0.2439,
+      "step": 5286
+    },
+    {
+      "epoch": 1.7652754590984974,
+      "grad_norm": 0.4608675290599738,
+      "learning_rate": 4.137416314130965e-07,
+      "loss": 0.2355,
+      "step": 5287
+    },
+    {
+      "epoch": 1.765609348914858,
+      "grad_norm": 0.48596552504607515,
+      "learning_rate": 4.125818445508351e-07,
+      "loss": 0.2472,
+      "step": 5288
+    },
+    {
+      "epoch": 1.7659432387312188,
+      "grad_norm": 0.47803271256325014,
+      "learning_rate": 4.1142361555206066e-07,
+      "loss": 0.2496,
+      "step": 5289
+    },
+    {
+      "epoch": 1.7662771285475793,
+      "grad_norm": 0.48471356633382306,
+      "learning_rate": 4.10266944810101e-07,
+      "loss": 0.2479,
+      "step": 5290
+    },
+    {
+      "epoch": 1.7666110183639399,
+      "grad_norm": 0.47248971335826173,
+      "learning_rate": 4.0911183271775625e-07,
+      "loss": 0.2458,
+      "step": 5291
+    },
+    {
+      "epoch": 1.7669449081803004,
+      "grad_norm": 0.4458157831116004,
+      "learning_rate": 4.079582796672976e-07,
+      "loss": 0.232,
+      "step": 5292
+    },
+    {
+      "epoch": 1.767278797996661,
+      "grad_norm": 0.48751597346079906,
+      "learning_rate": 4.06806286050464e-07,
+      "loss": 0.256,
+      "step": 5293
+    },
+    {
+      "epoch": 1.7676126878130218,
+      "grad_norm": 0.4722492839707816,
+      "learning_rate": 4.0565585225846806e-07,
+      "loss": 0.2405,
+      "step": 5294
+    },
+    {
+      "epoch": 1.7679465776293823,
+      "grad_norm": 0.467284096785122,
+      "learning_rate": 4.0450697868198885e-07,
+      "loss": 0.2412,
+      "step": 5295
+    },
+    {
+      "epoch": 1.7682804674457429,
+      "grad_norm": 0.4478973272913406,
+      "learning_rate": 4.033596657111799e-07,
+      "loss": 0.228,
+      "step": 5296
+    },
+    {
+      "epoch": 1.7686143572621034,
+      "grad_norm": 0.4611195063774751,
+      "learning_rate": 4.022139137356623e-07,
+      "loss": 0.2441,
+      "step": 5297
+    },
+    {
+      "epoch": 1.768948247078464,
+      "grad_norm": 0.46666024506338005,
+      "learning_rate": 4.010697231445276e-07,
+      "loss": 0.2412,
+      "step": 5298
+    },
+    {
+      "epoch": 1.7692821368948248,
+      "grad_norm": 0.4680627336189031,
+      "learning_rate": 3.999270943263356e-07,
+      "loss": 0.2466,
+      "step": 5299
+    },
+    {
+      "epoch": 1.7696160267111853,
+      "grad_norm": 0.4750671392308169,
+      "learning_rate": 3.987860276691186e-07,
+      "loss": 0.2498,
+      "step": 5300
+    },
+    {
+      "epoch": 1.7699499165275459,
+      "grad_norm": 0.46050919970321114,
+      "learning_rate": 3.9764652356037715e-07,
+      "loss": 0.2444,
+      "step": 5301
+    },
+    {
+      "epoch": 1.7702838063439064,
+      "grad_norm": 0.45878032234642374,
+      "learning_rate": 3.965085823870807e-07,
+      "loss": 0.2423,
+      "step": 5302
+    },
+    {
+      "epoch": 1.770617696160267,
+      "grad_norm": 0.45603296683614913,
+      "learning_rate": 3.9537220453566784e-07,
+      "loss": 0.2352,
+      "step": 5303
+    },
+    {
+      "epoch": 1.7709515859766278,
+      "grad_norm": 0.46204734012180687,
+      "learning_rate": 3.9423739039204546e-07,
+      "loss": 0.2575,
+      "step": 5304
+    },
+    {
+      "epoch": 1.7712854757929883,
+      "grad_norm": 0.4556346750376226,
+      "learning_rate": 3.9310414034159306e-07,
+      "loss": 0.2429,
+      "step": 5305
+    },
+    {
+      "epoch": 1.7716193656093489,
+      "grad_norm": 0.4658497999922338,
+      "learning_rate": 3.9197245476915556e-07,
+      "loss": 0.2521,
+      "step": 5306
+    },
+    {
+      "epoch": 1.7719532554257094,
+      "grad_norm": 0.4733820266780628,
+      "learning_rate": 3.908423340590456e-07,
+      "loss": 0.2462,
+      "step": 5307
+    },
+    {
+      "epoch": 1.77228714524207,
+      "grad_norm": 0.4762457738035972,
+      "learning_rate": 3.8971377859504944e-07,
+      "loss": 0.2449,
+      "step": 5308
+    },
+    {
+      "epoch": 1.7726210350584308,
+      "grad_norm": 0.4808836146543555,
+      "learning_rate": 3.885867887604161e-07,
+      "loss": 0.2541,
+      "step": 5309
+    },
+    {
+      "epoch": 1.7729549248747913,
+      "grad_norm": 0.4782689350142458,
+      "learning_rate": 3.8746136493786725e-07,
+      "loss": 0.2498,
+      "step": 5310
+    },
+    {
+      "epoch": 1.7732888146911518,
+      "grad_norm": 0.4551458593358314,
+      "learning_rate": 3.8633750750959097e-07,
+      "loss": 0.2479,
+      "step": 5311
+    },
+    {
+      "epoch": 1.7736227045075124,
+      "grad_norm": 0.46387628880968984,
+      "learning_rate": 3.8521521685724204e-07,
+      "loss": 0.2498,
+      "step": 5312
+    },
+    {
+      "epoch": 1.773956594323873,
+      "grad_norm": 0.4612862599419894,
+      "learning_rate": 3.8409449336194505e-07,
+      "loss": 0.2417,
+      "step": 5313
+    },
+    {
+      "epoch": 1.7742904841402338,
+      "grad_norm": 0.45791409582290166,
+      "learning_rate": 3.829753374042927e-07,
+      "loss": 0.2494,
+      "step": 5314
+    },
+    {
+      "epoch": 1.7746243739565943,
+      "grad_norm": 0.43889168476269985,
+      "learning_rate": 3.818577493643444e-07,
+      "loss": 0.2373,
+      "step": 5315
+    },
+    {
+      "epoch": 1.7749582637729548,
+      "grad_norm": 0.4688815669515016,
+      "learning_rate": 3.807417296216254e-07,
+      "loss": 0.2414,
+      "step": 5316
+    },
+    {
+      "epoch": 1.7752921535893156,
+      "grad_norm": 0.47253126939209705,
+      "learning_rate": 3.7962727855513305e-07,
+      "loss": 0.2434,
+      "step": 5317
+    },
+    {
+      "epoch": 1.775626043405676,
+      "grad_norm": 0.45154222978264574,
+      "learning_rate": 3.785143965433269e-07,
+      "loss": 0.2353,
+      "step": 5318
+    },
+    {
+      "epoch": 1.7759599332220368,
+      "grad_norm": 0.4720535022278631,
+      "learning_rate": 3.7740308396413794e-07,
+      "loss": 0.2517,
+      "step": 5319
+    },
+    {
+      "epoch": 1.7762938230383973,
+      "grad_norm": 0.4567681925522677,
+      "learning_rate": 3.76293341194961e-07,
+      "loss": 0.2477,
+      "step": 5320
+    },
+    {
+      "epoch": 1.7766277128547578,
+      "grad_norm": 0.46827874718231033,
+      "learning_rate": 3.75185168612659e-07,
+      "loss": 0.2392,
+      "step": 5321
+    },
+    {
+      "epoch": 1.7769616026711186,
+      "grad_norm": 0.44219151885173746,
+      "learning_rate": 3.7407856659356147e-07,
+      "loss": 0.2325,
+      "step": 5322
+    },
+    {
+      "epoch": 1.777295492487479,
+      "grad_norm": 0.46089880825948615,
+      "learning_rate": 3.7297353551346507e-07,
+      "loss": 0.2334,
+      "step": 5323
+    },
+    {
+      "epoch": 1.7776293823038398,
+      "grad_norm": 0.476762748567693,
+      "learning_rate": 3.718700757476323e-07,
+      "loss": 0.2549,
+      "step": 5324
+    },
+    {
+      "epoch": 1.7779632721202003,
+      "grad_norm": 0.4630799048737172,
+      "learning_rate": 3.7076818767079304e-07,
+      "loss": 0.2453,
+      "step": 5325
+    },
+    {
+      "epoch": 1.7782971619365608,
+      "grad_norm": 0.48293411833739636,
+      "learning_rate": 3.696678716571428e-07,
+      "loss": 0.251,
+      "step": 5326
+    },
+    {
+      "epoch": 1.7786310517529216,
+      "grad_norm": 0.4794313168430625,
+      "learning_rate": 3.6856912808034164e-07,
+      "loss": 0.2443,
+      "step": 5327
+    },
+    {
+      "epoch": 1.778964941569282,
+      "grad_norm": 0.46984007130270256,
+      "learning_rate": 3.6747195731351935e-07,
+      "loss": 0.2435,
+      "step": 5328
+    },
+    {
+      "epoch": 1.7792988313856428,
+      "grad_norm": 0.483662759179581,
+      "learning_rate": 3.6637635972926853e-07,
+      "loss": 0.2517,
+      "step": 5329
+    },
+    {
+      "epoch": 1.7796327212020033,
+      "grad_norm": 0.4640972671669119,
+      "learning_rate": 3.652823356996488e-07,
+      "loss": 0.2482,
+      "step": 5330
+    },
+    {
+      "epoch": 1.7799666110183638,
+      "grad_norm": 0.48004862193004233,
+      "learning_rate": 3.641898855961834e-07,
+      "loss": 0.2461,
+      "step": 5331
+    },
+    {
+      "epoch": 1.7803005008347246,
+      "grad_norm": 0.4786472925025637,
+      "learning_rate": 3.6309900978986454e-07,
+      "loss": 0.2358,
+      "step": 5332
+    },
+    {
+      "epoch": 1.7806343906510853,
+      "grad_norm": 0.46785452833648417,
+      "learning_rate": 3.6200970865114705e-07,
+      "loss": 0.2475,
+      "step": 5333
+    },
+    {
+      "epoch": 1.7809682804674458,
+      "grad_norm": 0.461545385825162,
+      "learning_rate": 3.6092198254995326e-07,
+      "loss": 0.2535,
+      "step": 5334
+    },
+    {
+      "epoch": 1.7813021702838063,
+      "grad_norm": 0.4494120674114874,
+      "learning_rate": 3.598358318556683e-07,
+      "loss": 0.2386,
+      "step": 5335
+    },
+    {
+      "epoch": 1.7816360601001668,
+      "grad_norm": 0.4815817154758275,
+      "learning_rate": 3.5875125693714265e-07,
+      "loss": 0.2463,
+      "step": 5336
+    },
+    {
+      "epoch": 1.7819699499165276,
+      "grad_norm": 0.4934531660140036,
+      "learning_rate": 3.576682581626939e-07,
+      "loss": 0.2469,
+      "step": 5337
+    },
+    {
+      "epoch": 1.7823038397328883,
+      "grad_norm": 0.48882157106970325,
+      "learning_rate": 3.5658683590010237e-07,
+      "loss": 0.2526,
+      "step": 5338
+    },
+    {
+      "epoch": 1.7826377295492488,
+      "grad_norm": 0.467341605031373,
+      "learning_rate": 3.555069905166131e-07,
+      "loss": 0.2484,
+      "step": 5339
+    },
+    {
+      "epoch": 1.7829716193656093,
+      "grad_norm": 0.48150434804847225,
+      "learning_rate": 3.5442872237893554e-07,
+      "loss": 0.2553,
+      "step": 5340
+    },
+    {
+      "epoch": 1.7833055091819698,
+      "grad_norm": 0.4879780302110648,
+      "learning_rate": 3.5335203185324406e-07,
+      "loss": 0.2506,
+      "step": 5341
+    },
+    {
+      "epoch": 1.7836393989983306,
+      "grad_norm": 0.4558464089476179,
+      "learning_rate": 3.5227691930517895e-07,
+      "loss": 0.2392,
+      "step": 5342
+    },
+    {
+      "epoch": 1.7839732888146913,
+      "grad_norm": 0.47516069727539595,
+      "learning_rate": 3.5120338509984143e-07,
+      "loss": 0.2465,
+      "step": 5343
+    },
+    {
+      "epoch": 1.7843071786310518,
+      "grad_norm": 0.44171306904241153,
+      "learning_rate": 3.5013142960179824e-07,
+      "loss": 0.2354,
+      "step": 5344
+    },
+    {
+      "epoch": 1.7846410684474123,
+      "grad_norm": 0.4606300375185207,
+      "learning_rate": 3.490610531750788e-07,
+      "loss": 0.2444,
+      "step": 5345
+    },
+    {
+      "epoch": 1.7849749582637728,
+      "grad_norm": 0.46894003838055787,
+      "learning_rate": 3.4799225618318e-07,
+      "loss": 0.2412,
+      "step": 5346
+    },
+    {
+      "epoch": 1.7853088480801336,
+      "grad_norm": 0.4725904540817893,
+      "learning_rate": 3.469250389890583e-07,
+      "loss": 0.2463,
+      "step": 5347
+    },
+    {
+      "epoch": 1.7856427378964943,
+      "grad_norm": 0.4723744008729653,
+      "learning_rate": 3.4585940195513537e-07,
+      "loss": 0.2525,
+      "step": 5348
+    },
+    {
+      "epoch": 1.7859766277128548,
+      "grad_norm": 0.4726319855416695,
+      "learning_rate": 3.447953454432951e-07,
+      "loss": 0.2524,
+      "step": 5349
+    },
+    {
+      "epoch": 1.7863105175292153,
+      "grad_norm": 0.47936332805304066,
+      "learning_rate": 3.437328698148862e-07,
+      "loss": 0.2535,
+      "step": 5350
+    },
+    {
+      "epoch": 1.7866444073455758,
+      "grad_norm": 0.47871007352327183,
+      "learning_rate": 3.426719754307206e-07,
+      "loss": 0.2463,
+      "step": 5351
+    },
+    {
+      "epoch": 1.7869782971619366,
+      "grad_norm": 0.47114105443561927,
+      "learning_rate": 3.416126626510724e-07,
+      "loss": 0.2566,
+      "step": 5352
+    },
+    {
+      "epoch": 1.7873121869782973,
+      "grad_norm": 0.47840401007334366,
+      "learning_rate": 3.405549318356788e-07,
+      "loss": 0.2443,
+      "step": 5353
+    },
+    {
+      "epoch": 1.7876460767946578,
+      "grad_norm": 0.4708635377610315,
+      "learning_rate": 3.3949878334373796e-07,
+      "loss": 0.2486,
+      "step": 5354
+    },
+    {
+      "epoch": 1.7879799666110183,
+      "grad_norm": 0.4738761819005699,
+      "learning_rate": 3.384442175339148e-07,
+      "loss": 0.2471,
+      "step": 5355
+    },
+    {
+      "epoch": 1.7883138564273788,
+      "grad_norm": 0.46647729833496127,
+      "learning_rate": 3.373912347643332e-07,
+      "loss": 0.239,
+      "step": 5356
+    },
+    {
+      "epoch": 1.7886477462437396,
+      "grad_norm": 0.47612892745444,
+      "learning_rate": 3.3633983539258065e-07,
+      "loss": 0.2475,
+      "step": 5357
+    },
+    {
+      "epoch": 1.7889816360601003,
+      "grad_norm": 0.489601510212642,
+      "learning_rate": 3.3529001977570695e-07,
+      "loss": 0.256,
+      "step": 5358
+    },
+    {
+      "epoch": 1.7893155258764608,
+      "grad_norm": 0.46310856441675774,
+      "learning_rate": 3.342417882702237e-07,
+      "loss": 0.2468,
+      "step": 5359
+    },
+    {
+      "epoch": 1.7896494156928213,
+      "grad_norm": 0.4710069037993586,
+      "learning_rate": 3.331951412321066e-07,
+      "loss": 0.2385,
+      "step": 5360
+    },
+    {
+      "epoch": 1.7899833055091818,
+      "grad_norm": 0.46331818415299436,
+      "learning_rate": 3.321500790167903e-07,
+      "loss": 0.2421,
+      "step": 5361
+    },
+    {
+      "epoch": 1.7903171953255426,
+      "grad_norm": 0.45402252993889824,
+      "learning_rate": 3.311066019791731e-07,
+      "loss": 0.2363,
+      "step": 5362
+    },
+    {
+      "epoch": 1.7906510851419033,
+      "grad_norm": 0.48523506826558094,
+      "learning_rate": 3.3006471047361267e-07,
+      "loss": 0.254,
+      "step": 5363
+    },
+    {
+      "epoch": 1.7909849749582638,
+      "grad_norm": 0.45405257102552626,
+      "learning_rate": 3.2902440485393204e-07,
+      "loss": 0.2453,
+      "step": 5364
+    },
+    {
+      "epoch": 1.7913188647746243,
+      "grad_norm": 0.49657624377576975,
+      "learning_rate": 3.2798568547341315e-07,
+      "loss": 0.2462,
+      "step": 5365
+    },
+    {
+      "epoch": 1.7916527545909848,
+      "grad_norm": 0.47387448502146906,
+      "learning_rate": 3.269485526847993e-07,
+      "loss": 0.2451,
+      "step": 5366
+    },
+    {
+      "epoch": 1.7919866444073456,
+      "grad_norm": 0.45823996412991214,
+      "learning_rate": 3.2591300684029494e-07,
+      "loss": 0.2467,
+      "step": 5367
+    },
+    {
+      "epoch": 1.7923205342237063,
+      "grad_norm": 0.4557853175667333,
+      "learning_rate": 3.2487904829156644e-07,
+      "loss": 0.2399,
+      "step": 5368
+    },
+    {
+      "epoch": 1.7926544240400668,
+      "grad_norm": 0.47571585623208623,
+      "learning_rate": 3.23846677389742e-07,
+      "loss": 0.2496,
+      "step": 5369
+    },
+    {
+      "epoch": 1.7929883138564273,
+      "grad_norm": 0.48645141718718043,
+      "learning_rate": 3.228158944854082e-07,
+      "loss": 0.2575,
+      "step": 5370
+    },
+    {
+      "epoch": 1.7933222036727878,
+      "grad_norm": 0.47655929085240645,
+      "learning_rate": 3.21786699928614e-07,
+      "loss": 0.2453,
+      "step": 5371
+    },
+    {
+      "epoch": 1.7936560934891486,
+      "grad_norm": 0.4696412757025309,
+      "learning_rate": 3.207590940688682e-07,
+      "loss": 0.2494,
+      "step": 5372
+    },
+    {
+      "epoch": 1.7939899833055093,
+      "grad_norm": 0.45339530880405343,
+      "learning_rate": 3.1973307725514047e-07,
+      "loss": 0.2445,
+      "step": 5373
+    },
+    {
+      "epoch": 1.7943238731218698,
+      "grad_norm": 0.47137660611567356,
+      "learning_rate": 3.187086498358616e-07,
+      "loss": 0.2582,
+      "step": 5374
+    },
+    {
+      "epoch": 1.7946577629382303,
+      "grad_norm": 0.46611049490014894,
+      "learning_rate": 3.1768581215892046e-07,
+      "loss": 0.2601,
+      "step": 5375
+    },
+    {
+      "epoch": 1.7949916527545908,
+      "grad_norm": 0.4635302072261996,
+      "learning_rate": 3.166645645716687e-07,
+      "loss": 0.2491,
+      "step": 5376
+    },
+    {
+      "epoch": 1.7953255425709516,
+      "grad_norm": 0.45298180653712766,
+      "learning_rate": 3.156449074209156e-07,
+      "loss": 0.2446,
+      "step": 5377
+    },
+    {
+      "epoch": 1.7956594323873123,
+      "grad_norm": 0.4539114282445554,
+      "learning_rate": 3.14626841052933e-07,
+      "loss": 0.2421,
+      "step": 5378
+    },
+    {
+      "epoch": 1.7959933222036728,
+      "grad_norm": 0.48171852836421997,
+      "learning_rate": 3.1361036581344985e-07,
+      "loss": 0.2497,
+      "step": 5379
+    },
+    {
+      "epoch": 1.7963272120200333,
+      "grad_norm": 0.47410857414202295,
+      "learning_rate": 3.125954820476562e-07,
+      "loss": 0.2507,
+      "step": 5380
+    },
+    {
+      "epoch": 1.7966611018363938,
+      "grad_norm": 0.46803674414373836,
+      "learning_rate": 3.1158219010020087e-07,
+      "loss": 0.234,
+      "step": 5381
+    },
+    {
+      "epoch": 1.7969949916527546,
+      "grad_norm": 0.46402733757005604,
+      "learning_rate": 3.105704903151929e-07,
+      "loss": 0.2405,
+      "step": 5382
+    },
+    {
+      "epoch": 1.7973288814691153,
+      "grad_norm": 0.463023682185276,
+      "learning_rate": 3.095603830362004e-07,
+      "loss": 0.2482,
+      "step": 5383
+    },
+    {
+      "epoch": 1.7976627712854758,
+      "grad_norm": 0.473704513316659,
+      "learning_rate": 3.0855186860625054e-07,
+      "loss": 0.2498,
+      "step": 5384
+    },
+    {
+      "epoch": 1.7979966611018363,
+      "grad_norm": 0.47665709870808926,
+      "learning_rate": 3.075449473678299e-07,
+      "loss": 0.2556,
+      "step": 5385
+    },
+    {
+      "epoch": 1.798330550918197,
+      "grad_norm": 0.46701378145940775,
+      "learning_rate": 3.0653961966288283e-07,
+      "loss": 0.2465,
+      "step": 5386
+    },
+    {
+      "epoch": 1.7986644407345576,
+      "grad_norm": 0.47789050162383395,
+      "learning_rate": 3.0553588583281446e-07,
+      "loss": 0.259,
+      "step": 5387
+    },
+    {
+      "epoch": 1.7989983305509183,
+      "grad_norm": 0.4572315280863874,
+      "learning_rate": 3.045337462184877e-07,
+      "loss": 0.2413,
+      "step": 5388
+    },
+    {
+      "epoch": 1.7993322203672788,
+      "grad_norm": 0.4592783789316767,
+      "learning_rate": 3.0353320116022267e-07,
+      "loss": 0.2479,
+      "step": 5389
+    },
+    {
+      "epoch": 1.7996661101836393,
+      "grad_norm": 0.4589288910754346,
+      "learning_rate": 3.0253425099779963e-07,
+      "loss": 0.2391,
+      "step": 5390
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.47911956046552323,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.253,
+      "step": 5391
+    },
+    {
+      "epoch": 1.8003338898163606,
+      "grad_norm": 0.4710110657888732,
+      "learning_rate": 3.0054113671689365e-07,
+      "loss": 0.2492,
+      "step": 5392
+    },
+    {
+      "epoch": 1.8006677796327213,
+      "grad_norm": 0.4604139370608681,
+      "learning_rate": 2.995469732752615e-07,
+      "loss": 0.2441,
+      "step": 5393
+    },
+    {
+      "epoch": 1.8010016694490818,
+      "grad_norm": 0.46900480078410894,
+      "learning_rate": 2.985544060831741e-07,
+      "loss": 0.2509,
+      "step": 5394
+    },
+    {
+      "epoch": 1.8013355592654423,
+      "grad_norm": 0.49442913904012115,
+      "learning_rate": 2.9756343547770127e-07,
+      "loss": 0.2617,
+      "step": 5395
+    },
+    {
+      "epoch": 1.801669449081803,
+      "grad_norm": 0.46962820362778845,
+      "learning_rate": 2.965740617953733e-07,
+      "loss": 0.2498,
+      "step": 5396
+    },
+    {
+      "epoch": 1.8020033388981636,
+      "grad_norm": 0.45686022553170064,
+      "learning_rate": 2.9558628537217583e-07,
+      "loss": 0.2449,
+      "step": 5397
+    },
+    {
+      "epoch": 1.8023372287145243,
+      "grad_norm": 0.4637304676288355,
+      "learning_rate": 2.9460010654355165e-07,
+      "loss": 0.2387,
+      "step": 5398
+    },
+    {
+      "epoch": 1.8026711185308848,
+      "grad_norm": 0.4582275691730836,
+      "learning_rate": 2.936155256444023e-07,
+      "loss": 0.2438,
+      "step": 5399
+    },
+    {
+      "epoch": 1.8030050083472453,
+      "grad_norm": 0.4664115299310696,
+      "learning_rate": 2.926325430090865e-07,
+      "loss": 0.2434,
+      "step": 5400
+    },
+    {
+      "epoch": 1.803338898163606,
+      "grad_norm": 0.4669903612723537,
+      "learning_rate": 2.916511589714216e-07,
+      "loss": 0.2444,
+      "step": 5401
+    },
+    {
+      "epoch": 1.8036727879799668,
+      "grad_norm": 0.4855395158298133,
+      "learning_rate": 2.9067137386467823e-07,
+      "loss": 0.2507,
+      "step": 5402
+    },
+    {
+      "epoch": 1.8040066777963273,
+      "grad_norm": 0.482565744310392,
+      "learning_rate": 2.896931880215881e-07,
+      "loss": 0.246,
+      "step": 5403
+    },
+    {
+      "epoch": 1.8043405676126878,
+      "grad_norm": 0.46748986302128015,
+      "learning_rate": 2.8871660177433646e-07,
+      "loss": 0.2534,
+      "step": 5404
+    },
+    {
+      "epoch": 1.8046744574290483,
+      "grad_norm": 0.46491799671223044,
+      "learning_rate": 2.877416154545681e-07,
+      "loss": 0.2501,
+      "step": 5405
+    },
+    {
+      "epoch": 1.805008347245409,
+      "grad_norm": 0.47618148388574255,
+      "learning_rate": 2.8676822939338375e-07,
+      "loss": 0.2542,
+      "step": 5406
+    },
+    {
+      "epoch": 1.8053422370617698,
+      "grad_norm": 0.47778234267364067,
+      "learning_rate": 2.8579644392133896e-07,
+      "loss": 0.2578,
+      "step": 5407
+    },
+    {
+      "epoch": 1.8056761268781303,
+      "grad_norm": 0.5027342165493799,
+      "learning_rate": 2.8482625936844755e-07,
+      "loss": 0.2588,
+      "step": 5408
+    },
+    {
+      "epoch": 1.8060100166944908,
+      "grad_norm": 0.47351917990988823,
+      "learning_rate": 2.8385767606417814e-07,
+      "loss": 0.2483,
+      "step": 5409
+    },
+    {
+      "epoch": 1.8063439065108513,
+      "grad_norm": 0.4772817755961298,
+      "learning_rate": 2.828906943374593e-07,
+      "loss": 0.2483,
+      "step": 5410
+    },
+    {
+      "epoch": 1.806677796327212,
+      "grad_norm": 0.475085945087576,
+      "learning_rate": 2.819253145166712e-07,
+      "loss": 0.2413,
+      "step": 5411
+    },
+    {
+      "epoch": 1.8070116861435728,
+      "grad_norm": 0.47743676011397335,
+      "learning_rate": 2.80961536929652e-07,
+      "loss": 0.2464,
+      "step": 5412
+    },
+    {
+      "epoch": 1.8073455759599333,
+      "grad_norm": 0.45958906401099275,
+      "learning_rate": 2.7999936190369615e-07,
+      "loss": 0.2426,
+      "step": 5413
+    },
+    {
+      "epoch": 1.8076794657762938,
+      "grad_norm": 0.4926948721206024,
+      "learning_rate": 2.7903878976555165e-07,
+      "loss": 0.2558,
+      "step": 5414
+    },
+    {
+      "epoch": 1.8080133555926543,
+      "grad_norm": 0.4689545385830566,
+      "learning_rate": 2.7807982084142595e-07,
+      "loss": 0.2618,
+      "step": 5415
+    },
+    {
+      "epoch": 1.808347245409015,
+      "grad_norm": 0.4724427652952268,
+      "learning_rate": 2.7712245545697903e-07,
+      "loss": 0.2484,
+      "step": 5416
+    },
+    {
+      "epoch": 1.8086811352253758,
+      "grad_norm": 0.46190315542018073,
+      "learning_rate": 2.7616669393732643e-07,
+      "loss": 0.244,
+      "step": 5417
+    },
+    {
+      "epoch": 1.8090150250417363,
+      "grad_norm": 0.46970600522785866,
+      "learning_rate": 2.7521253660704075e-07,
+      "loss": 0.245,
+      "step": 5418
+    },
+    {
+      "epoch": 1.8093489148580968,
+      "grad_norm": 0.48655871413749446,
+      "learning_rate": 2.742599837901494e-07,
+      "loss": 0.2491,
+      "step": 5419
+    },
+    {
+      "epoch": 1.8096828046744573,
+      "grad_norm": 0.46911357342133403,
+      "learning_rate": 2.733090358101337e-07,
+      "loss": 0.2488,
+      "step": 5420
+    },
+    {
+      "epoch": 1.810016694490818,
+      "grad_norm": 0.4577890771828799,
+      "learning_rate": 2.723596929899308e-07,
+      "loss": 0.2449,
+      "step": 5421
+    },
+    {
+      "epoch": 1.8103505843071788,
+      "grad_norm": 0.455904023034579,
+      "learning_rate": 2.7141195565193237e-07,
+      "loss": 0.2374,
+      "step": 5422
+    },
+    {
+      "epoch": 1.8106844741235393,
+      "grad_norm": 0.45998302432111837,
+      "learning_rate": 2.704658241179847e-07,
+      "loss": 0.2372,
+      "step": 5423
+    },
+    {
+      "epoch": 1.8110183639398998,
+      "grad_norm": 0.4579866019150119,
+      "learning_rate": 2.6952129870939035e-07,
+      "loss": 0.2444,
+      "step": 5424
+    },
+    {
+      "epoch": 1.8113522537562603,
+      "grad_norm": 0.45837831935505435,
+      "learning_rate": 2.685783797469049e-07,
+      "loss": 0.2537,
+      "step": 5425
+    },
+    {
+      "epoch": 1.811686143572621,
+      "grad_norm": 0.46350690509189385,
+      "learning_rate": 2.6763706755073726e-07,
+      "loss": 0.2431,
+      "step": 5426
+    },
+    {
+      "epoch": 1.8120200333889818,
+      "grad_norm": 0.44350919426165386,
+      "learning_rate": 2.666973624405544e-07,
+      "loss": 0.2372,
+      "step": 5427
+    },
+    {
+      "epoch": 1.8123539232053423,
+      "grad_norm": 0.47414462368156757,
+      "learning_rate": 2.657592647354734e-07,
+      "loss": 0.2547,
+      "step": 5428
+    },
+    {
+      "epoch": 1.8126878130217028,
+      "grad_norm": 0.4723570061889481,
+      "learning_rate": 2.6482277475406935e-07,
+      "loss": 0.2267,
+      "step": 5429
+    },
+    {
+      "epoch": 1.8130217028380633,
+      "grad_norm": 0.46360182440568565,
+      "learning_rate": 2.6388789281436844e-07,
+      "loss": 0.2449,
+      "step": 5430
+    },
+    {
+      "epoch": 1.813355592654424,
+      "grad_norm": 0.4867710555472745,
+      "learning_rate": 2.629546192338517e-07,
+      "loss": 0.2548,
+      "step": 5431
+    },
+    {
+      "epoch": 1.8136894824707848,
+      "grad_norm": 0.494896855112373,
+      "learning_rate": 2.620229543294528e-07,
+      "loss": 0.2609,
+      "step": 5432
+    },
+    {
+      "epoch": 1.8140233722871453,
+      "grad_norm": 0.4695103978096548,
+      "learning_rate": 2.6109289841756257e-07,
+      "loss": 0.2462,
+      "step": 5433
+    },
+    {
+      "epoch": 1.8143572621035058,
+      "grad_norm": 0.4956776176656017,
+      "learning_rate": 2.601644518140217e-07,
+      "loss": 0.2468,
+      "step": 5434
+    },
+    {
+      "epoch": 1.8146911519198663,
+      "grad_norm": 0.4712213027739387,
+      "learning_rate": 2.5923761483412635e-07,
+      "loss": 0.2575,
+      "step": 5435
+    },
+    {
+      "epoch": 1.815025041736227,
+      "grad_norm": 0.4709553859919981,
+      "learning_rate": 2.583123877926258e-07,
+      "loss": 0.249,
+      "step": 5436
+    },
+    {
+      "epoch": 1.8153589315525878,
+      "grad_norm": 0.46930339723434317,
+      "learning_rate": 2.573887710037215e-07,
+      "loss": 0.2535,
+      "step": 5437
+    },
+    {
+      "epoch": 1.8156928213689483,
+      "grad_norm": 0.4597604603777485,
+      "learning_rate": 2.5646676478107034e-07,
+      "loss": 0.2475,
+      "step": 5438
+    },
+    {
+      "epoch": 1.8160267111853088,
+      "grad_norm": 0.4480661706116194,
+      "learning_rate": 2.5554636943777965e-07,
+      "loss": 0.2421,
+      "step": 5439
+    },
+    {
+      "epoch": 1.8163606010016693,
+      "grad_norm": 0.47336072685734043,
+      "learning_rate": 2.546275852864116e-07,
+      "loss": 0.2505,
+      "step": 5440
+    },
+    {
+      "epoch": 1.81669449081803,
+      "grad_norm": 0.4605392817310272,
+      "learning_rate": 2.5371041263897945e-07,
+      "loss": 0.2369,
+      "step": 5441
+    },
+    {
+      "epoch": 1.8170283806343908,
+      "grad_norm": 0.4783868899261674,
+      "learning_rate": 2.527948518069517e-07,
+      "loss": 0.2396,
+      "step": 5442
+    },
+    {
+      "epoch": 1.8173622704507513,
+      "grad_norm": 0.4630728408745935,
+      "learning_rate": 2.51880903101247e-07,
+      "loss": 0.2444,
+      "step": 5443
+    },
+    {
+      "epoch": 1.8176961602671118,
+      "grad_norm": 0.45834976699743935,
+      "learning_rate": 2.509685668322387e-07,
+      "loss": 0.2323,
+      "step": 5444
+    },
+    {
+      "epoch": 1.8180300500834723,
+      "grad_norm": 0.4717765681769888,
+      "learning_rate": 2.500578433097511e-07,
+      "loss": 0.2501,
+      "step": 5445
+    },
+    {
+      "epoch": 1.818363939899833,
+      "grad_norm": 0.46266545396030084,
+      "learning_rate": 2.491487328430603e-07,
+      "loss": 0.2408,
+      "step": 5446
+    },
+    {
+      "epoch": 1.8186978297161938,
+      "grad_norm": 0.47873886873337546,
+      "learning_rate": 2.482412357408964e-07,
+      "loss": 0.2445,
+      "step": 5447
+    },
+    {
+      "epoch": 1.8190317195325543,
+      "grad_norm": 0.45987336385339334,
+      "learning_rate": 2.473353523114408e-07,
+      "loss": 0.2461,
+      "step": 5448
+    },
+    {
+      "epoch": 1.8193656093489148,
+      "grad_norm": 0.4727865819744004,
+      "learning_rate": 2.464310828623262e-07,
+      "loss": 0.2491,
+      "step": 5449
+    },
+    {
+      "epoch": 1.8196994991652755,
+      "grad_norm": 0.4832574846206866,
+      "learning_rate": 2.4552842770063756e-07,
+      "loss": 0.2501,
+      "step": 5450
+    },
+    {
+      "epoch": 1.820033388981636,
+      "grad_norm": 0.4661970292417355,
+      "learning_rate": 2.44627387132913e-07,
+      "loss": 0.2374,
+      "step": 5451
+    },
+    {
+      "epoch": 1.8203672787979968,
+      "grad_norm": 0.4955013829349986,
+      "learning_rate": 2.4372796146514056e-07,
+      "loss": 0.2496,
+      "step": 5452
+    },
+    {
+      "epoch": 1.8207011686143573,
+      "grad_norm": 0.45681352117336066,
+      "learning_rate": 2.428301510027614e-07,
+      "loss": 0.2348,
+      "step": 5453
+    },
+    {
+      "epoch": 1.8210350584307178,
+      "grad_norm": 0.47619037808476994,
+      "learning_rate": 2.419339560506667e-07,
+      "loss": 0.258,
+      "step": 5454
+    },
+    {
+      "epoch": 1.8213689482470785,
+      "grad_norm": 0.45782607381643975,
+      "learning_rate": 2.4103937691319856e-07,
+      "loss": 0.2395,
+      "step": 5455
+    },
+    {
+      "epoch": 1.821702838063439,
+      "grad_norm": 0.5168927627351607,
+      "learning_rate": 2.401464138941528e-07,
+      "loss": 0.2604,
+      "step": 5456
+    },
+    {
+      "epoch": 1.8220367278797998,
+      "grad_norm": 0.45864433653944114,
+      "learning_rate": 2.392550672967753e-07,
+      "loss": 0.247,
+      "step": 5457
+    },
+    {
+      "epoch": 1.8223706176961603,
+      "grad_norm": 0.44792431044410747,
+      "learning_rate": 2.3836533742376212e-07,
+      "loss": 0.2409,
+      "step": 5458
+    },
+    {
+      "epoch": 1.8227045075125208,
+      "grad_norm": 0.46236654110468683,
+      "learning_rate": 2.3747722457725996e-07,
+      "loss": 0.2443,
+      "step": 5459
+    },
+    {
+      "epoch": 1.8230383973288815,
+      "grad_norm": 0.4808554903142352,
+      "learning_rate": 2.3659072905886815e-07,
+      "loss": 0.2576,
+      "step": 5460
+    },
+    {
+      "epoch": 1.823372287145242,
+      "grad_norm": 0.4767140020923188,
+      "learning_rate": 2.357058511696375e-07,
+      "loss": 0.2472,
+      "step": 5461
+    },
+    {
+      "epoch": 1.8237061769616028,
+      "grad_norm": 0.4551167660197069,
+      "learning_rate": 2.34822591210066e-07,
+      "loss": 0.2375,
+      "step": 5462
+    },
+    {
+      "epoch": 1.8240400667779633,
+      "grad_norm": 0.4574429985675856,
+      "learning_rate": 2.3394094948010539e-07,
+      "loss": 0.2369,
+      "step": 5463
+    },
+    {
+      "epoch": 1.8243739565943238,
+      "grad_norm": 0.4783898188874311,
+      "learning_rate": 2.33060926279155e-07,
+      "loss": 0.2472,
+      "step": 5464
+    },
+    {
+      "epoch": 1.8247078464106845,
+      "grad_norm": 0.4635460730602953,
+      "learning_rate": 2.3218252190606804e-07,
+      "loss": 0.2435,
+      "step": 5465
+    },
+    {
+      "epoch": 1.8250417362270452,
+      "grad_norm": 0.4715437224163109,
+      "learning_rate": 2.313057366591448e-07,
+      "loss": 0.2517,
+      "step": 5466
+    },
+    {
+      "epoch": 1.8253756260434058,
+      "grad_norm": 0.4680499690521324,
+      "learning_rate": 2.3043057083613818e-07,
+      "loss": 0.2526,
+      "step": 5467
+    },
+    {
+      "epoch": 1.8257095158597663,
+      "grad_norm": 0.47185448987174283,
+      "learning_rate": 2.2955702473424824e-07,
+      "loss": 0.2605,
+      "step": 5468
+    },
+    {
+      "epoch": 1.8260434056761268,
+      "grad_norm": 0.43881354313513643,
+      "learning_rate": 2.2868509865012767e-07,
+      "loss": 0.2345,
+      "step": 5469
+    },
+    {
+      "epoch": 1.8263772954924875,
+      "grad_norm": 0.48326777061670123,
+      "learning_rate": 2.2781479287987908e-07,
+      "loss": 0.2523,
+      "step": 5470
+    },
+    {
+      "epoch": 1.8267111853088482,
+      "grad_norm": 0.4547953773590844,
+      "learning_rate": 2.2694610771905267e-07,
+      "loss": 0.2444,
+      "step": 5471
+    },
+    {
+      "epoch": 1.8270450751252088,
+      "grad_norm": 0.47091362298093553,
+      "learning_rate": 2.260790434626492e-07,
+      "loss": 0.24,
+      "step": 5472
+    },
+    {
+      "epoch": 1.8273789649415693,
+      "grad_norm": 0.45591511030283466,
+      "learning_rate": 2.2521360040511974e-07,
+      "loss": 0.2579,
+      "step": 5473
+    },
+    {
+      "epoch": 1.8277128547579298,
+      "grad_norm": 0.46857658358475446,
+      "learning_rate": 2.2434977884036424e-07,
+      "loss": 0.2419,
+      "step": 5474
+    },
+    {
+      "epoch": 1.8280467445742905,
+      "grad_norm": 0.49678978500757215,
+      "learning_rate": 2.2348757906173245e-07,
+      "loss": 0.2546,
+      "step": 5475
+    },
+    {
+      "epoch": 1.8283806343906512,
+      "grad_norm": 0.47511173883186786,
+      "learning_rate": 2.2262700136202241e-07,
+      "loss": 0.2524,
+      "step": 5476
+    },
+    {
+      "epoch": 1.8287145242070117,
+      "grad_norm": 0.4681091623256702,
+      "learning_rate": 2.217680460334809e-07,
+      "loss": 0.2517,
+      "step": 5477
+    },
+    {
+      "epoch": 1.8290484140233723,
+      "grad_norm": 0.454941506979999,
+      "learning_rate": 2.2091071336780622e-07,
+      "loss": 0.2454,
+      "step": 5478
+    },
+    {
+      "epoch": 1.8293823038397328,
+      "grad_norm": 0.46350853884600995,
+      "learning_rate": 2.2005500365614386e-07,
+      "loss": 0.2549,
+      "step": 5479
+    },
+    {
+      "epoch": 1.8297161936560935,
+      "grad_norm": 0.49733605491679655,
+      "learning_rate": 2.1920091718908853e-07,
+      "loss": 0.2633,
+      "step": 5480
+    },
+    {
+      "epoch": 1.8300500834724542,
+      "grad_norm": 0.47086609973860655,
+      "learning_rate": 2.1834845425668327e-07,
+      "loss": 0.2474,
+      "step": 5481
+    },
+    {
+      "epoch": 1.8303839732888147,
+      "grad_norm": 0.4664364094706561,
+      "learning_rate": 2.1749761514841871e-07,
+      "loss": 0.2382,
+      "step": 5482
+    },
+    {
+      "epoch": 1.8307178631051753,
+      "grad_norm": 0.4630768344270842,
+      "learning_rate": 2.1664840015323818e-07,
+      "loss": 0.2331,
+      "step": 5483
+    },
+    {
+      "epoch": 1.8310517529215358,
+      "grad_norm": 0.4850035113591518,
+      "learning_rate": 2.1580080955952875e-07,
+      "loss": 0.2602,
+      "step": 5484
+    },
+    {
+      "epoch": 1.8313856427378965,
+      "grad_norm": 0.4780966722889159,
+      "learning_rate": 2.1495484365512797e-07,
+      "loss": 0.2509,
+      "step": 5485
+    },
+    {
+      "epoch": 1.8317195325542572,
+      "grad_norm": 0.4733653547795717,
+      "learning_rate": 2.141105027273227e-07,
+      "loss": 0.255,
+      "step": 5486
+    },
+    {
+      "epoch": 1.8320534223706177,
+      "grad_norm": 0.45286593602911956,
+      "learning_rate": 2.1326778706284468e-07,
+      "loss": 0.2482,
+      "step": 5487
+    },
+    {
+      "epoch": 1.8323873121869783,
+      "grad_norm": 0.4593129178381757,
+      "learning_rate": 2.1242669694787776e-07,
+      "loss": 0.2468,
+      "step": 5488
+    },
+    {
+      "epoch": 1.8327212020033388,
+      "grad_norm": 0.4622242295706907,
+      "learning_rate": 2.1158723266805125e-07,
+      "loss": 0.2397,
+      "step": 5489
+    },
+    {
+      "epoch": 1.8330550918196995,
+      "grad_norm": 0.47560781294014665,
+      "learning_rate": 2.1074939450844267e-07,
+      "loss": 0.2508,
+      "step": 5490
+    },
+    {
+      "epoch": 1.8333889816360602,
+      "grad_norm": 0.4626215124167977,
+      "learning_rate": 2.0991318275357663e-07,
+      "loss": 0.2443,
+      "step": 5491
+    },
+    {
+      "epoch": 1.8337228714524207,
+      "grad_norm": 0.44089762897352475,
+      "learning_rate": 2.0907859768742823e-07,
+      "loss": 0.2458,
+      "step": 5492
+    },
+    {
+      "epoch": 1.8340567612687813,
+      "grad_norm": 0.4931380892626743,
+      "learning_rate": 2.0824563959341682e-07,
+      "loss": 0.2483,
+      "step": 5493
+    },
+    {
+      "epoch": 1.8343906510851418,
+      "grad_norm": 0.4738465755854713,
+      "learning_rate": 2.074143087544106e-07,
+      "loss": 0.2475,
+      "step": 5494
+    },
+    {
+      "epoch": 1.8347245409015025,
+      "grad_norm": 0.45991025906344857,
+      "learning_rate": 2.0658460545272653e-07,
+      "loss": 0.2351,
+      "step": 5495
+    },
+    {
+      "epoch": 1.8350584307178632,
+      "grad_norm": 0.47286786097048666,
+      "learning_rate": 2.0575652997012586e-07,
+      "loss": 0.2426,
+      "step": 5496
+    },
+    {
+      "epoch": 1.8353923205342237,
+      "grad_norm": 0.4704075864139457,
+      "learning_rate": 2.0493008258782033e-07,
+      "loss": 0.2544,
+      "step": 5497
+    },
+    {
+      "epoch": 1.8357262103505843,
+      "grad_norm": 0.4712896727245379,
+      "learning_rate": 2.0410526358646654e-07,
+      "loss": 0.2516,
+      "step": 5498
+    },
+    {
+      "epoch": 1.8360601001669448,
+      "grad_norm": 0.4750501323013859,
+      "learning_rate": 2.0328207324616877e-07,
+      "loss": 0.2557,
+      "step": 5499
+    },
+    {
+      "epoch": 1.8363939899833055,
+      "grad_norm": 0.45989725167150663,
+      "learning_rate": 2.024605118464773e-07,
+      "loss": 0.2444,
+      "step": 5500
+    },
+    {
+      "epoch": 1.8367278797996662,
+      "grad_norm": 0.4549317647093536,
+      "learning_rate": 2.0164057966639116e-07,
+      "loss": 0.2344,
+      "step": 5501
+    },
+    {
+      "epoch": 1.8370617696160267,
+      "grad_norm": 0.45057076635249355,
+      "learning_rate": 2.0082227698435485e-07,
+      "loss": 0.239,
+      "step": 5502
+    },
+    {
+      "epoch": 1.8373956594323873,
+      "grad_norm": 0.46595408121131193,
+      "learning_rate": 2.0000560407825999e-07,
+      "loss": 0.2436,
+      "step": 5503
+    },
+    {
+      "epoch": 1.8377295492487478,
+      "grad_norm": 0.456965318596994,
+      "learning_rate": 1.9919056122544467e-07,
+      "loss": 0.2434,
+      "step": 5504
+    },
+    {
+      "epoch": 1.8380634390651085,
+      "grad_norm": 0.4668358738647557,
+      "learning_rate": 1.9837714870269198e-07,
+      "loss": 0.2443,
+      "step": 5505
+    },
+    {
+      "epoch": 1.8383973288814692,
+      "grad_norm": 0.46497621426001845,
+      "learning_rate": 1.9756536678623429e-07,
+      "loss": 0.2484,
+      "step": 5506
+    },
+    {
+      "epoch": 1.8387312186978297,
+      "grad_norm": 0.4625038924918306,
+      "learning_rate": 1.9675521575174717e-07,
+      "loss": 0.2498,
+      "step": 5507
+    },
+    {
+      "epoch": 1.8390651085141902,
+      "grad_norm": 0.4772604744059744,
+      "learning_rate": 1.9594669587435445e-07,
+      "loss": 0.2356,
+      "step": 5508
+    },
+    {
+      "epoch": 1.8393989983305508,
+      "grad_norm": 0.458197359115479,
+      "learning_rate": 1.9513980742862482e-07,
+      "loss": 0.2437,
+      "step": 5509
+    },
+    {
+      "epoch": 1.8397328881469115,
+      "grad_norm": 0.4770968791564782,
+      "learning_rate": 1.943345506885741e-07,
+      "loss": 0.2487,
+      "step": 5510
+    },
+    {
+      "epoch": 1.8400667779632722,
+      "grad_norm": 0.4801833790454237,
+      "learning_rate": 1.9353092592766186e-07,
+      "loss": 0.244,
+      "step": 5511
+    },
+    {
+      "epoch": 1.8404006677796327,
+      "grad_norm": 0.47565518128459733,
+      "learning_rate": 1.9272893341879707e-07,
+      "loss": 0.2532,
+      "step": 5512
+    },
+    {
+      "epoch": 1.8407345575959932,
+      "grad_norm": 0.48969623928635103,
+      "learning_rate": 1.9192857343433069e-07,
+      "loss": 0.2441,
+      "step": 5513
+    },
+    {
+      "epoch": 1.8410684474123538,
+      "grad_norm": 0.47086736249612093,
+      "learning_rate": 1.911298462460609e-07,
+      "loss": 0.2514,
+      "step": 5514
+    },
+    {
+      "epoch": 1.8414023372287145,
+      "grad_norm": 0.47276190364915666,
+      "learning_rate": 1.9033275212523238e-07,
+      "loss": 0.2451,
+      "step": 5515
+    },
+    {
+      "epoch": 1.8417362270450752,
+      "grad_norm": 0.47421041492833677,
+      "learning_rate": 1.8953729134253362e-07,
+      "loss": 0.2501,
+      "step": 5516
+    },
+    {
+      "epoch": 1.8420701168614357,
+      "grad_norm": 0.4780834459756555,
+      "learning_rate": 1.8874346416809853e-07,
+      "loss": 0.2544,
+      "step": 5517
+    },
+    {
+      "epoch": 1.8424040066777962,
+      "grad_norm": 0.48303016529668275,
+      "learning_rate": 1.8795127087150644e-07,
+      "loss": 0.2475,
+      "step": 5518
+    },
+    {
+      "epoch": 1.842737896494157,
+      "grad_norm": 0.47215676968594283,
+      "learning_rate": 1.871607117217833e-07,
+      "loss": 0.2392,
+      "step": 5519
+    },
+    {
+      "epoch": 1.8430717863105175,
+      "grad_norm": 0.47771232309636985,
+      "learning_rate": 1.863717869873982e-07,
+      "loss": 0.2444,
+      "step": 5520
+    },
+    {
+      "epoch": 1.8434056761268782,
+      "grad_norm": 0.46044813206353824,
+      "learning_rate": 1.8558449693626633e-07,
+      "loss": 0.2438,
+      "step": 5521
+    },
+    {
+      "epoch": 1.8437395659432387,
+      "grad_norm": 0.47327095251434026,
+      "learning_rate": 1.847988418357466e-07,
+      "loss": 0.245,
+      "step": 5522
+    },
+    {
+      "epoch": 1.8440734557595992,
+      "grad_norm": 0.44636713933486755,
+      "learning_rate": 1.8401482195264386e-07,
+      "loss": 0.244,
+      "step": 5523
+    },
+    {
+      "epoch": 1.84440734557596,
+      "grad_norm": 0.45347940766326406,
+      "learning_rate": 1.8323243755320741e-07,
+      "loss": 0.2487,
+      "step": 5524
+    },
+    {
+      "epoch": 1.8447412353923205,
+      "grad_norm": 0.4703352883469288,
+      "learning_rate": 1.824516889031308e-07,
+      "loss": 0.2484,
+      "step": 5525
+    },
+    {
+      "epoch": 1.8450751252086812,
+      "grad_norm": 0.4612157696593409,
+      "learning_rate": 1.8167257626755196e-07,
+      "loss": 0.2397,
+      "step": 5526
+    },
+    {
+      "epoch": 1.8454090150250417,
+      "grad_norm": 0.47188817156282736,
+      "learning_rate": 1.8089509991105315e-07,
+      "loss": 0.2408,
+      "step": 5527
+    },
+    {
+      "epoch": 1.8457429048414022,
+      "grad_norm": 0.47727015166612385,
+      "learning_rate": 1.8011926009766202e-07,
+      "loss": 0.246,
+      "step": 5528
+    },
+    {
+      "epoch": 1.846076794657763,
+      "grad_norm": 0.4447166514185811,
+      "learning_rate": 1.7934505709085004e-07,
+      "loss": 0.2365,
+      "step": 5529
+    },
+    {
+      "epoch": 1.8464106844741235,
+      "grad_norm": 0.46076655537517747,
+      "learning_rate": 1.7857249115353192e-07,
+      "loss": 0.2386,
+      "step": 5530
+    },
+    {
+      "epoch": 1.8467445742904842,
+      "grad_norm": 0.46640158746540084,
+      "learning_rate": 1.778015625480678e-07,
+      "loss": 0.2433,
+      "step": 5531
+    },
+    {
+      "epoch": 1.8470784641068447,
+      "grad_norm": 0.46160568119111767,
+      "learning_rate": 1.7703227153625934e-07,
+      "loss": 0.248,
+      "step": 5532
+    },
+    {
+      "epoch": 1.8474123539232052,
+      "grad_norm": 0.4846582235332713,
+      "learning_rate": 1.7626461837935592e-07,
+      "loss": 0.2604,
+      "step": 5533
+    },
+    {
+      "epoch": 1.847746243739566,
+      "grad_norm": 0.44576890705483363,
+      "learning_rate": 1.7549860333804735e-07,
+      "loss": 0.2388,
+      "step": 5534
+    },
+    {
+      "epoch": 1.8480801335559267,
+      "grad_norm": 0.4852702061047057,
+      "learning_rate": 1.7473422667246886e-07,
+      "loss": 0.2501,
+      "step": 5535
+    },
+    {
+      "epoch": 1.8484140233722872,
+      "grad_norm": 0.4903302610341483,
+      "learning_rate": 1.7397148864219782e-07,
+      "loss": 0.2539,
+      "step": 5536
+    },
+    {
+      "epoch": 1.8487479131886477,
+      "grad_norm": 0.4711505769331896,
+      "learning_rate": 1.7321038950625702e-07,
+      "loss": 0.2491,
+      "step": 5537
+    },
+    {
+      "epoch": 1.8490818030050082,
+      "grad_norm": 0.46896911642918865,
+      "learning_rate": 1.7245092952311193e-07,
+      "loss": 0.2501,
+      "step": 5538
+    },
+    {
+      "epoch": 1.849415692821369,
+      "grad_norm": 0.4666784860619274,
+      "learning_rate": 1.7169310895067126e-07,
+      "loss": 0.2484,
+      "step": 5539
+    },
+    {
+      "epoch": 1.8497495826377297,
+      "grad_norm": 0.45749749694530195,
+      "learning_rate": 1.7093692804628637e-07,
+      "loss": 0.2393,
+      "step": 5540
+    },
+    {
+      "epoch": 1.8500834724540902,
+      "grad_norm": 0.47407072680167706,
+      "learning_rate": 1.7018238706675237e-07,
+      "loss": 0.2383,
+      "step": 5541
+    },
+    {
+      "epoch": 1.8504173622704507,
+      "grad_norm": 0.477812548701259,
+      "learning_rate": 1.6942948626830824e-07,
+      "loss": 0.2451,
+      "step": 5542
+    },
+    {
+      "epoch": 1.8507512520868112,
+      "grad_norm": 0.4758009313081135,
+      "learning_rate": 1.6867822590663497e-07,
+      "loss": 0.2518,
+      "step": 5543
+    },
+    {
+      "epoch": 1.851085141903172,
+      "grad_norm": 0.5040759472009858,
+      "learning_rate": 1.6792860623685624e-07,
+      "loss": 0.2518,
+      "step": 5544
+    },
+    {
+      "epoch": 1.8514190317195327,
+      "grad_norm": 0.46960673144948645,
+      "learning_rate": 1.6718062751353902e-07,
+      "loss": 0.2561,
+      "step": 5545
+    },
+    {
+      "epoch": 1.8517529215358932,
+      "grad_norm": 0.4472081184319388,
+      "learning_rate": 1.664342899906929e-07,
+      "loss": 0.2374,
+      "step": 5546
+    },
+    {
+      "epoch": 1.8520868113522537,
+      "grad_norm": 0.49905094024970836,
+      "learning_rate": 1.656895939217712e-07,
+      "loss": 0.2428,
+      "step": 5547
+    },
+    {
+      "epoch": 1.8524207011686142,
+      "grad_norm": 0.4550601933705948,
+      "learning_rate": 1.6494653955966833e-07,
+      "loss": 0.255,
+      "step": 5548
+    },
+    {
+      "epoch": 1.852754590984975,
+      "grad_norm": 0.47304421900750476,
+      "learning_rate": 1.642051271567213e-07,
+      "loss": 0.2442,
+      "step": 5549
+    },
+    {
+      "epoch": 1.8530884808013357,
+      "grad_norm": 0.46728794714076627,
+      "learning_rate": 1.634653569647099e-07,
+      "loss": 0.2382,
+      "step": 5550
+    },
+    {
+      "epoch": 1.8534223706176962,
+      "grad_norm": 0.4708161241865143,
+      "learning_rate": 1.627272292348575e-07,
+      "loss": 0.242,
+      "step": 5551
+    },
+    {
+      "epoch": 1.8537562604340567,
+      "grad_norm": 0.4671130969400677,
+      "learning_rate": 1.6199074421782702e-07,
+      "loss": 0.2494,
+      "step": 5552
+    },
+    {
+      "epoch": 1.8540901502504172,
+      "grad_norm": 0.4397283848445979,
+      "learning_rate": 1.61255902163725e-07,
+      "loss": 0.2366,
+      "step": 5553
+    },
+    {
+      "epoch": 1.854424040066778,
+      "grad_norm": 0.44568602970902427,
+      "learning_rate": 1.605227033221013e-07,
+      "loss": 0.2325,
+      "step": 5554
+    },
+    {
+      "epoch": 1.8547579298831387,
+      "grad_norm": 0.45836338911311414,
+      "learning_rate": 1.5979114794194506e-07,
+      "loss": 0.2354,
+      "step": 5555
+    },
+    {
+      "epoch": 1.8550918196994992,
+      "grad_norm": 0.4918212590966045,
+      "learning_rate": 1.590612362716898e-07,
+      "loss": 0.257,
+      "step": 5556
+    },
+    {
+      "epoch": 1.8554257095158597,
+      "grad_norm": 0.47406413911923756,
+      "learning_rate": 1.5833296855920944e-07,
+      "loss": 0.2522,
+      "step": 5557
+    },
+    {
+      "epoch": 1.8557595993322202,
+      "grad_norm": 0.4584138574006161,
+      "learning_rate": 1.5760634505182004e-07,
+      "loss": 0.2455,
+      "step": 5558
+    },
+    {
+      "epoch": 1.856093489148581,
+      "grad_norm": 0.4733548008246845,
+      "learning_rate": 1.5688136599627757e-07,
+      "loss": 0.2494,
+      "step": 5559
+    },
+    {
+      "epoch": 1.8564273789649417,
+      "grad_norm": 0.45547514449285037,
+      "learning_rate": 1.561580316387834e-07,
+      "loss": 0.243,
+      "step": 5560
+    },
+    {
+      "epoch": 1.8567612687813022,
+      "grad_norm": 0.44732955568077226,
+      "learning_rate": 1.5543634222497718e-07,
+      "loss": 0.236,
+      "step": 5561
+    },
+    {
+      "epoch": 1.8570951585976627,
+      "grad_norm": 0.4735788184085831,
+      "learning_rate": 1.547162979999406e-07,
+      "loss": 0.2441,
+      "step": 5562
+    },
+    {
+      "epoch": 1.8574290484140232,
+      "grad_norm": 0.45555078232454393,
+      "learning_rate": 1.5399789920819808e-07,
+      "loss": 0.2465,
+      "step": 5563
+    },
+    {
+      "epoch": 1.857762938230384,
+      "grad_norm": 0.4599215512868056,
+      "learning_rate": 1.5328114609371224e-07,
+      "loss": 0.2456,
+      "step": 5564
+    },
+    {
+      "epoch": 1.8580968280467447,
+      "grad_norm": 0.46442537360394714,
+      "learning_rate": 1.5256603889989063e-07,
+      "loss": 0.2454,
+      "step": 5565
+    },
+    {
+      "epoch": 1.8584307178631052,
+      "grad_norm": 0.4740864754303273,
+      "learning_rate": 1.518525778695795e-07,
+      "loss": 0.2446,
+      "step": 5566
+    },
+    {
+      "epoch": 1.8587646076794657,
+      "grad_norm": 0.46929395872880386,
+      "learning_rate": 1.5114076324506567e-07,
+      "loss": 0.2402,
+      "step": 5567
+    },
+    {
+      "epoch": 1.8590984974958262,
+      "grad_norm": 0.47031069714512147,
+      "learning_rate": 1.504305952680779e-07,
+      "loss": 0.2487,
+      "step": 5568
+    },
+    {
+      "epoch": 1.859432387312187,
+      "grad_norm": 0.4664937152340864,
+      "learning_rate": 1.4972207417978669e-07,
+      "loss": 0.2521,
+      "step": 5569
+    },
+    {
+      "epoch": 1.8597662771285477,
+      "grad_norm": 0.4946654868644428,
+      "learning_rate": 1.4901520022080063e-07,
+      "loss": 0.2512,
+      "step": 5570
+    },
+    {
+      "epoch": 1.8601001669449082,
+      "grad_norm": 0.45494471880285187,
+      "learning_rate": 1.4830997363117218e-07,
+      "loss": 0.2442,
+      "step": 5571
+    },
+    {
+      "epoch": 1.8604340567612687,
+      "grad_norm": 0.4521422285758792,
+      "learning_rate": 1.4760639465039084e-07,
+      "loss": 0.2443,
+      "step": 5572
+    },
+    {
+      "epoch": 1.8607679465776292,
+      "grad_norm": 0.4834758397231174,
+      "learning_rate": 1.4690446351738942e-07,
+      "loss": 0.2533,
+      "step": 5573
+    },
+    {
+      "epoch": 1.86110183639399,
+      "grad_norm": 0.4751619412293637,
+      "learning_rate": 1.4620418047053998e-07,
+      "loss": 0.2469,
+      "step": 5574
+    },
+    {
+      "epoch": 1.8614357262103507,
+      "grad_norm": 0.4680318187820225,
+      "learning_rate": 1.4550554574765507e-07,
+      "loss": 0.2556,
+      "step": 5575
+    },
+    {
+      "epoch": 1.8617696160267112,
+      "grad_norm": 0.474306904447675,
+      "learning_rate": 1.4480855958598716e-07,
+      "loss": 0.2514,
+      "step": 5576
+    },
+    {
+      "epoch": 1.8621035058430717,
+      "grad_norm": 0.4630839167493572,
+      "learning_rate": 1.4411322222222913e-07,
+      "loss": 0.2391,
+      "step": 5577
+    },
+    {
+      "epoch": 1.8624373956594322,
+      "grad_norm": 0.4896146459995647,
+      "learning_rate": 1.4341953389251427e-07,
+      "loss": 0.2481,
+      "step": 5578
+    },
+    {
+      "epoch": 1.862771285475793,
+      "grad_norm": 0.4462856289992834,
+      "learning_rate": 1.4272749483241477e-07,
+      "loss": 0.2368,
+      "step": 5579
+    },
+    {
+      "epoch": 1.8631051752921537,
+      "grad_norm": 0.46915729492949304,
+      "learning_rate": 1.420371052769448e-07,
+      "loss": 0.2418,
+      "step": 5580
+    },
+    {
+      "epoch": 1.8634390651085142,
+      "grad_norm": 0.4644942893182027,
+      "learning_rate": 1.4134836546055686e-07,
+      "loss": 0.2466,
+      "step": 5581
+    },
+    {
+      "epoch": 1.8637729549248747,
+      "grad_norm": 0.4643325426838156,
+      "learning_rate": 1.4066127561714215e-07,
+      "loss": 0.247,
+      "step": 5582
+    },
+    {
+      "epoch": 1.8641068447412354,
+      "grad_norm": 0.47976884518874313,
+      "learning_rate": 1.3997583598003405e-07,
+      "loss": 0.2427,
+      "step": 5583
+    },
+    {
+      "epoch": 1.864440734557596,
+      "grad_norm": 0.4745015288917905,
+      "learning_rate": 1.3929204678200415e-07,
+      "loss": 0.2408,
+      "step": 5584
+    },
+    {
+      "epoch": 1.8647746243739567,
+      "grad_norm": 0.46689089614886775,
+      "learning_rate": 1.3860990825526334e-07,
+      "loss": 0.2395,
+      "step": 5585
+    },
+    {
+      "epoch": 1.8651085141903172,
+      "grad_norm": 0.4773853440617286,
+      "learning_rate": 1.3792942063146188e-07,
+      "loss": 0.2391,
+      "step": 5586
+    },
+    {
+      "epoch": 1.8654424040066777,
+      "grad_norm": 0.461076969099267,
+      "learning_rate": 1.3725058414169047e-07,
+      "loss": 0.2477,
+      "step": 5587
+    },
+    {
+      "epoch": 1.8657762938230384,
+      "grad_norm": 0.49232555190371613,
+      "learning_rate": 1.3657339901647914e-07,
+      "loss": 0.2565,
+      "step": 5588
+    },
+    {
+      "epoch": 1.866110183639399,
+      "grad_norm": 0.47465113319306257,
+      "learning_rate": 1.3589786548579563e-07,
+      "loss": 0.2512,
+      "step": 5589
+    },
+    {
+      "epoch": 1.8664440734557597,
+      "grad_norm": 0.4765881006911071,
+      "learning_rate": 1.352239837790481e-07,
+      "loss": 0.2524,
+      "step": 5590
+    },
+    {
+      "epoch": 1.8667779632721202,
+      "grad_norm": 0.427153323616612,
+      "learning_rate": 1.3455175412508237e-07,
+      "loss": 0.2239,
+      "step": 5591
+    },
+    {
+      "epoch": 1.8671118530884807,
+      "grad_norm": 0.46962299370964405,
+      "learning_rate": 1.3388117675218525e-07,
+      "loss": 0.2429,
+      "step": 5592
+    },
+    {
+      "epoch": 1.8674457429048414,
+      "grad_norm": 0.4530045930332671,
+      "learning_rate": 1.3321225188808075e-07,
+      "loss": 0.246,
+      "step": 5593
+    },
+    {
+      "epoch": 1.867779632721202,
+      "grad_norm": 0.46375795435474126,
+      "learning_rate": 1.3254497975993263e-07,
+      "loss": 0.2416,
+      "step": 5594
+    },
+    {
+      "epoch": 1.8681135225375627,
+      "grad_norm": 0.4604575284964545,
+      "learning_rate": 1.3187936059434303e-07,
+      "loss": 0.2422,
+      "step": 5595
+    },
+    {
+      "epoch": 1.8684474123539232,
+      "grad_norm": 0.4656435923313742,
+      "learning_rate": 1.3121539461735112e-07,
+      "loss": 0.2369,
+      "step": 5596
+    },
+    {
+      "epoch": 1.8687813021702837,
+      "grad_norm": 0.46592072805575174,
+      "learning_rate": 1.3055308205443928e-07,
+      "loss": 0.2421,
+      "step": 5597
+    },
+    {
+      "epoch": 1.8691151919866444,
+      "grad_norm": 0.48630249926977703,
+      "learning_rate": 1.2989242313052318e-07,
+      "loss": 0.2493,
+      "step": 5598
+    },
+    {
+      "epoch": 1.8694490818030052,
+      "grad_norm": 0.4729988460117708,
+      "learning_rate": 1.292334180699606e-07,
+      "loss": 0.2452,
+      "step": 5599
+    },
+    {
+      "epoch": 1.8697829716193657,
+      "grad_norm": 0.47565321542941724,
+      "learning_rate": 1.285760670965447e-07,
+      "loss": 0.2495,
+      "step": 5600
+    },
+    {
+      "epoch": 1.8701168614357262,
+      "grad_norm": 0.45090204870110484,
+      "learning_rate": 1.2792037043351023e-07,
+      "loss": 0.2464,
+      "step": 5601
+    },
+    {
+      "epoch": 1.8704507512520867,
+      "grad_norm": 0.45590874891289546,
+      "learning_rate": 1.2726632830352748e-07,
+      "loss": 0.2451,
+      "step": 5602
+    },
+    {
+      "epoch": 1.8707846410684474,
+      "grad_norm": 0.47996433849625914,
+      "learning_rate": 1.266139409287054e-07,
+      "loss": 0.2466,
+      "step": 5603
+    },
+    {
+      "epoch": 1.8711185308848082,
+      "grad_norm": 0.46956346446312636,
+      "learning_rate": 1.2596320853059175e-07,
+      "loss": 0.247,
+      "step": 5604
+    },
+    {
+      "epoch": 1.8714524207011687,
+      "grad_norm": 0.46457527347342803,
+      "learning_rate": 1.2531413133017146e-07,
+      "loss": 0.2385,
+      "step": 5605
+    },
+    {
+      "epoch": 1.8717863105175292,
+      "grad_norm": 0.4594565878058066,
+      "learning_rate": 1.246667095478693e-07,
+      "loss": 0.2417,
+      "step": 5606
+    },
+    {
+      "epoch": 1.8721202003338897,
+      "grad_norm": 0.48130739555561813,
+      "learning_rate": 1.2402094340354497e-07,
+      "loss": 0.2499,
+      "step": 5607
+    },
+    {
+      "epoch": 1.8724540901502504,
+      "grad_norm": 0.44805380463774114,
+      "learning_rate": 1.233768331164975e-07,
+      "loss": 0.2352,
+      "step": 5608
+    },
+    {
+      "epoch": 1.8727879799666112,
+      "grad_norm": 0.49135647760855383,
+      "learning_rate": 1.2273437890546357e-07,
+      "loss": 0.2541,
+      "step": 5609
+    },
+    {
+      "epoch": 1.8731218697829717,
+      "grad_norm": 0.4553003143209982,
+      "learning_rate": 1.22093580988617e-07,
+      "loss": 0.2502,
+      "step": 5610
+    },
+    {
+      "epoch": 1.8734557595993322,
+      "grad_norm": 0.47882887450074074,
+      "learning_rate": 1.214544395835704e-07,
+      "loss": 0.2437,
+      "step": 5611
+    },
+    {
+      "epoch": 1.8737896494156927,
+      "grad_norm": 0.45956315210555837,
+      "learning_rate": 1.2081695490737177e-07,
+      "loss": 0.2507,
+      "step": 5612
+    },
+    {
+      "epoch": 1.8741235392320534,
+      "grad_norm": 0.47447981684673546,
+      "learning_rate": 1.2018112717650797e-07,
+      "loss": 0.2378,
+      "step": 5613
+    },
+    {
+      "epoch": 1.8744574290484142,
+      "grad_norm": 0.4586166274560622,
+      "learning_rate": 1.1954695660690353e-07,
+      "loss": 0.2463,
+      "step": 5614
+    },
+    {
+      "epoch": 1.8747913188647747,
+      "grad_norm": 0.47228426724607764,
+      "learning_rate": 1.1891444341391778e-07,
+      "loss": 0.25,
+      "step": 5615
+    },
+    {
+      "epoch": 1.8751252086811352,
+      "grad_norm": 0.45386922781547895,
+      "learning_rate": 1.1828358781235117e-07,
+      "loss": 0.2503,
+      "step": 5616
+    },
+    {
+      "epoch": 1.8754590984974957,
+      "grad_norm": 0.4491601968517656,
+      "learning_rate": 1.176543900164373e-07,
+      "loss": 0.2354,
+      "step": 5617
+    },
+    {
+      "epoch": 1.8757929883138564,
+      "grad_norm": 0.4807939007719051,
+      "learning_rate": 1.1702685023984972e-07,
+      "loss": 0.25,
+      "step": 5618
+    },
+    {
+      "epoch": 1.8761268781302172,
+      "grad_norm": 0.4667139361566036,
+      "learning_rate": 1.1640096869569628e-07,
+      "loss": 0.2484,
+      "step": 5619
+    },
+    {
+      "epoch": 1.8764607679465777,
+      "grad_norm": 0.46761356024352574,
+      "learning_rate": 1.1577674559652419e-07,
+      "loss": 0.2468,
+      "step": 5620
+    },
+    {
+      "epoch": 1.8767946577629382,
+      "grad_norm": 0.482500184111278,
+      "learning_rate": 1.1515418115431554e-07,
+      "loss": 0.2554,
+      "step": 5621
+    },
+    {
+      "epoch": 1.8771285475792987,
+      "grad_norm": 0.4683607763843555,
+      "learning_rate": 1.1453327558049122e-07,
+      "loss": 0.2427,
+      "step": 5622
+    },
+    {
+      "epoch": 1.8774624373956594,
+      "grad_norm": 0.471495606871019,
+      "learning_rate": 1.1391402908590644e-07,
+      "loss": 0.2413,
+      "step": 5623
+    },
+    {
+      "epoch": 1.8777963272120202,
+      "grad_norm": 0.4464440648846218,
+      "learning_rate": 1.132964418808541e-07,
+      "loss": 0.2375,
+      "step": 5624
+    },
+    {
+      "epoch": 1.8781302170283807,
+      "grad_norm": 0.46492196420725557,
+      "learning_rate": 1.1268051417506476e-07,
+      "loss": 0.2453,
+      "step": 5625
+    },
+    {
+      "epoch": 1.8784641068447412,
+      "grad_norm": 0.46196200682318983,
+      "learning_rate": 1.120662461777039e-07,
+      "loss": 0.243,
+      "step": 5626
+    },
+    {
+      "epoch": 1.8787979966611017,
+      "grad_norm": 0.4700401391481696,
+      "learning_rate": 1.1145363809737297e-07,
+      "loss": 0.2507,
+      "step": 5627
+    },
+    {
+      "epoch": 1.8791318864774624,
+      "grad_norm": 0.45925431987314397,
+      "learning_rate": 1.1084269014211058e-07,
+      "loss": 0.2421,
+      "step": 5628
+    },
+    {
+      "epoch": 1.8794657762938232,
+      "grad_norm": 0.49036440176748397,
+      "learning_rate": 1.102334025193924e-07,
+      "loss": 0.2456,
+      "step": 5629
+    },
+    {
+      "epoch": 1.8797996661101837,
+      "grad_norm": 0.4613747531584229,
+      "learning_rate": 1.0962577543612796e-07,
+      "loss": 0.2457,
+      "step": 5630
+    },
+    {
+      "epoch": 1.8801335559265442,
+      "grad_norm": 0.473260981521093,
+      "learning_rate": 1.0901980909866605e-07,
+      "loss": 0.2474,
+      "step": 5631
+    },
+    {
+      "epoch": 1.8804674457429047,
+      "grad_norm": 0.49737439819465923,
+      "learning_rate": 1.0841550371278875e-07,
+      "loss": 0.2461,
+      "step": 5632
+    },
+    {
+      "epoch": 1.8808013355592654,
+      "grad_norm": 0.4776875480793625,
+      "learning_rate": 1.0781285948371467e-07,
+      "loss": 0.2436,
+      "step": 5633
+    },
+    {
+      "epoch": 1.8811352253756262,
+      "grad_norm": 0.4725551821198631,
+      "learning_rate": 1.07211876616099e-07,
+      "loss": 0.2467,
+      "step": 5634
+    },
+    {
+      "epoch": 1.8814691151919867,
+      "grad_norm": 0.461011958779526,
+      "learning_rate": 1.0661255531403292e-07,
+      "loss": 0.2429,
+      "step": 5635
+    },
+    {
+      "epoch": 1.8818030050083472,
+      "grad_norm": 0.4674538205079181,
+      "learning_rate": 1.0601489578104252e-07,
+      "loss": 0.249,
+      "step": 5636
+    },
+    {
+      "epoch": 1.8821368948247077,
+      "grad_norm": 0.45530518714477974,
+      "learning_rate": 1.0541889822008932e-07,
+      "loss": 0.2361,
+      "step": 5637
+    },
+    {
+      "epoch": 1.8824707846410684,
+      "grad_norm": 0.4852733886274472,
+      "learning_rate": 1.0482456283357145e-07,
+      "loss": 0.2587,
+      "step": 5638
+    },
+    {
+      "epoch": 1.8828046744574292,
+      "grad_norm": 0.45956418890076806,
+      "learning_rate": 1.04231889823323e-07,
+      "loss": 0.2414,
+      "step": 5639
+    },
+    {
+      "epoch": 1.8831385642737897,
+      "grad_norm": 0.47268906474135824,
+      "learning_rate": 1.0364087939061129e-07,
+      "loss": 0.2464,
+      "step": 5640
+    },
+    {
+      "epoch": 1.8834724540901502,
+      "grad_norm": 0.45441502535203204,
+      "learning_rate": 1.0305153173614135e-07,
+      "loss": 0.2405,
+      "step": 5641
+    },
+    {
+      "epoch": 1.8838063439065107,
+      "grad_norm": 0.4591422534812823,
+      "learning_rate": 1.0246384706005142e-07,
+      "loss": 0.2494,
+      "step": 5642
+    },
+    {
+      "epoch": 1.8841402337228714,
+      "grad_norm": 0.44846035425160546,
+      "learning_rate": 1.0187782556191739e-07,
+      "loss": 0.2361,
+      "step": 5643
+    },
+    {
+      "epoch": 1.8844741235392322,
+      "grad_norm": 0.46969616696474403,
+      "learning_rate": 1.0129346744074897e-07,
+      "loss": 0.2417,
+      "step": 5644
+    },
+    {
+      "epoch": 1.8848080133555927,
+      "grad_norm": 0.4691657905296209,
+      "learning_rate": 1.0071077289499076e-07,
+      "loss": 0.2534,
+      "step": 5645
+    },
+    {
+      "epoch": 1.8851419031719532,
+      "grad_norm": 0.4537392122682773,
+      "learning_rate": 1.0012974212252224e-07,
+      "loss": 0.2389,
+      "step": 5646
+    },
+    {
+      "epoch": 1.8854757929883137,
+      "grad_norm": 0.46020938030077735,
+      "learning_rate": 9.955037532065947e-08,
+      "loss": 0.2477,
+      "step": 5647
+    },
+    {
+      "epoch": 1.8858096828046744,
+      "grad_norm": 0.4538943545302913,
+      "learning_rate": 9.897267268615285e-08,
+      "loss": 0.2434,
+      "step": 5648
+    },
+    {
+      "epoch": 1.8861435726210352,
+      "grad_norm": 0.46733103858789593,
+      "learning_rate": 9.839663441518598e-08,
+      "loss": 0.2369,
+      "step": 5649
+    },
+    {
+      "epoch": 1.8864774624373957,
+      "grad_norm": 0.4538663523834215,
+      "learning_rate": 9.782226070337908e-08,
+      "loss": 0.2447,
+      "step": 5650
+    },
+    {
+      "epoch": 1.8868113522537562,
+      "grad_norm": 0.4685294657082337,
+      "learning_rate": 9.724955174578609e-08,
+      "loss": 0.2431,
+      "step": 5651
+    },
+    {
+      "epoch": 1.887145242070117,
+      "grad_norm": 0.47001541201789065,
+      "learning_rate": 9.667850773689647e-08,
+      "loss": 0.2433,
+      "step": 5652
+    },
+    {
+      "epoch": 1.8874791318864774,
+      "grad_norm": 0.4672551551701302,
+      "learning_rate": 9.610912887063395e-08,
+      "loss": 0.2458,
+      "step": 5653
+    },
+    {
+      "epoch": 1.8878130217028382,
+      "grad_norm": 0.4725124512005744,
+      "learning_rate": 9.554141534035611e-08,
+      "loss": 0.2525,
+      "step": 5654
+    },
+    {
+      "epoch": 1.8881469115191987,
+      "grad_norm": 0.4675147659527351,
+      "learning_rate": 9.497536733885537e-08,
+      "loss": 0.2525,
+      "step": 5655
+    },
+    {
+      "epoch": 1.8884808013355592,
+      "grad_norm": 0.46742374791859076,
+      "learning_rate": 9.441098505835911e-08,
+      "loss": 0.2446,
+      "step": 5656
+    },
+    {
+      "epoch": 1.88881469115192,
+      "grad_norm": 0.459515510620899,
+      "learning_rate": 9.384826869052899e-08,
+      "loss": 0.2397,
+      "step": 5657
+    },
+    {
+      "epoch": 1.8891485809682804,
+      "grad_norm": 0.45302125039606983,
+      "learning_rate": 9.328721842646049e-08,
+      "loss": 0.2396,
+      "step": 5658
+    },
+    {
+      "epoch": 1.8894824707846412,
+      "grad_norm": 0.4777340760170382,
+      "learning_rate": 9.272783445668287e-08,
+      "loss": 0.2502,
+      "step": 5659
+    },
+    {
+      "epoch": 1.8898163606010017,
+      "grad_norm": 0.46914942012276045,
+      "learning_rate": 9.217011697115974e-08,
+      "loss": 0.2506,
+      "step": 5660
+    },
+    {
+      "epoch": 1.8901502504173622,
+      "grad_norm": 0.4520230631974771,
+      "learning_rate": 9.161406615929013e-08,
+      "loss": 0.2537,
+      "step": 5661
+    },
+    {
+      "epoch": 1.890484140233723,
+      "grad_norm": 0.47418078019087523,
+      "learning_rate": 9.105968220990524e-08,
+      "loss": 0.2504,
+      "step": 5662
+    },
+    {
+      "epoch": 1.8908180300500834,
+      "grad_norm": 0.463951367851662,
+      "learning_rate": 9.050696531127167e-08,
+      "loss": 0.2492,
+      "step": 5663
+    },
+    {
+      "epoch": 1.8911519198664442,
+      "grad_norm": 0.47478277912068856,
+      "learning_rate": 8.995591565108819e-08,
+      "loss": 0.2521,
+      "step": 5664
+    },
+    {
+      "epoch": 1.8914858096828047,
+      "grad_norm": 0.4639760668815153,
+      "learning_rate": 8.940653341648897e-08,
+      "loss": 0.232,
+      "step": 5665
+    },
+    {
+      "epoch": 1.8918196994991652,
+      "grad_norm": 0.4578221949246857,
+      "learning_rate": 8.885881879404201e-08,
+      "loss": 0.2306,
+      "step": 5666
+    },
+    {
+      "epoch": 1.892153589315526,
+      "grad_norm": 0.461417353258876,
+      "learning_rate": 8.831277196974741e-08,
+      "loss": 0.253,
+      "step": 5667
+    },
+    {
+      "epoch": 1.8924874791318866,
+      "grad_norm": 0.470775426919451,
+      "learning_rate": 8.776839312904074e-08,
+      "loss": 0.2373,
+      "step": 5668
+    },
+    {
+      "epoch": 1.8928213689482472,
+      "grad_norm": 0.4712861346585117,
+      "learning_rate": 8.722568245678908e-08,
+      "loss": 0.2469,
+      "step": 5669
+    },
+    {
+      "epoch": 1.8931552587646077,
+      "grad_norm": 0.46920635168357994,
+      "learning_rate": 8.668464013729506e-08,
+      "loss": 0.2507,
+      "step": 5670
+    },
+    {
+      "epoch": 1.8934891485809682,
+      "grad_norm": 0.4666713480952104,
+      "learning_rate": 8.614526635429388e-08,
+      "loss": 0.2408,
+      "step": 5671
+    },
+    {
+      "epoch": 1.893823038397329,
+      "grad_norm": 0.4580628400397286,
+      "learning_rate": 8.560756129095294e-08,
+      "loss": 0.2293,
+      "step": 5672
+    },
+    {
+      "epoch": 1.8941569282136896,
+      "grad_norm": 0.453784228800319,
+      "learning_rate": 8.50715251298756e-08,
+      "loss": 0.2419,
+      "step": 5673
+    },
+    {
+      "epoch": 1.8944908180300501,
+      "grad_norm": 0.4484309416690045,
+      "learning_rate": 8.45371580530957e-08,
+      "loss": 0.2443,
+      "step": 5674
+    },
+    {
+      "epoch": 1.8948247078464107,
+      "grad_norm": 0.47249156313891716,
+      "learning_rate": 8.400446024208309e-08,
+      "loss": 0.2516,
+      "step": 5675
+    },
+    {
+      "epoch": 1.8951585976627712,
+      "grad_norm": 0.48438301747078005,
+      "learning_rate": 8.347343187773805e-08,
+      "loss": 0.2502,
+      "step": 5676
+    },
+    {
+      "epoch": 1.895492487479132,
+      "grad_norm": 0.45805404043609727,
+      "learning_rate": 8.294407314039521e-08,
+      "loss": 0.2509,
+      "step": 5677
+    },
+    {
+      "epoch": 1.8958263772954926,
+      "grad_norm": 0.4587571534575136,
+      "learning_rate": 8.24163842098219e-08,
+      "loss": 0.243,
+      "step": 5678
+    },
+    {
+      "epoch": 1.8961602671118531,
+      "grad_norm": 0.4629701927528089,
+      "learning_rate": 8.189036526521921e-08,
+      "loss": 0.2381,
+      "step": 5679
+    },
+    {
+      "epoch": 1.8964941569282137,
+      "grad_norm": 0.4725292376231659,
+      "learning_rate": 8.13660164852198e-08,
+      "loss": 0.2422,
+      "step": 5680
+    },
+    {
+      "epoch": 1.8968280467445742,
+      "grad_norm": 0.46003609750049823,
+      "learning_rate": 8.084333804788958e-08,
+      "loss": 0.2385,
+      "step": 5681
+    },
+    {
+      "epoch": 1.897161936560935,
+      "grad_norm": 0.4580807859416563,
+      "learning_rate": 8.032233013072821e-08,
+      "loss": 0.251,
+      "step": 5682
+    },
+    {
+      "epoch": 1.8974958263772956,
+      "grad_norm": 0.47316808678989813,
+      "learning_rate": 7.980299291066696e-08,
+      "loss": 0.2524,
+      "step": 5683
+    },
+    {
+      "epoch": 1.8978297161936561,
+      "grad_norm": 0.4671273276719206,
+      "learning_rate": 7.928532656407029e-08,
+      "loss": 0.2479,
+      "step": 5684
+    },
+    {
+      "epoch": 1.8981636060100167,
+      "grad_norm": 0.4637397612753655,
+      "learning_rate": 7.876933126673481e-08,
+      "loss": 0.2384,
+      "step": 5685
+    },
+    {
+      "epoch": 1.8984974958263772,
+      "grad_norm": 0.4643078694042257,
+      "learning_rate": 7.825500719389035e-08,
+      "loss": 0.2415,
+      "step": 5686
+    },
+    {
+      "epoch": 1.898831385642738,
+      "grad_norm": 0.4725408244973861,
+      "learning_rate": 7.774235452019718e-08,
+      "loss": 0.2416,
+      "step": 5687
+    },
+    {
+      "epoch": 1.8991652754590986,
+      "grad_norm": 0.47489469389123945,
+      "learning_rate": 7.723137341975107e-08,
+      "loss": 0.255,
+      "step": 5688
+    },
+    {
+      "epoch": 1.8994991652754591,
+      "grad_norm": 0.45797703977019644,
+      "learning_rate": 7.672206406607763e-08,
+      "loss": 0.2341,
+      "step": 5689
+    },
+    {
+      "epoch": 1.8998330550918197,
+      "grad_norm": 0.4720998360218292,
+      "learning_rate": 7.62144266321363e-08,
+      "loss": 0.2427,
+      "step": 5690
+    },
+    {
+      "epoch": 1.9001669449081802,
+      "grad_norm": 0.47260925486698707,
+      "learning_rate": 7.570846129031861e-08,
+      "loss": 0.2465,
+      "step": 5691
+    },
+    {
+      "epoch": 1.900500834724541,
+      "grad_norm": 0.4790155439220563,
+      "learning_rate": 7.520416821244603e-08,
+      "loss": 0.2477,
+      "step": 5692
+    },
+    {
+      "epoch": 1.9008347245409016,
+      "grad_norm": 0.45992420115348936,
+      "learning_rate": 7.470154756977544e-08,
+      "loss": 0.2472,
+      "step": 5693
+    },
+    {
+      "epoch": 1.9011686143572621,
+      "grad_norm": 0.4443297422054657,
+      "learning_rate": 7.420059953299363e-08,
+      "loss": 0.2417,
+      "step": 5694
+    },
+    {
+      "epoch": 1.9015025041736227,
+      "grad_norm": 0.4850179903432914,
+      "learning_rate": 7.37013242722201e-08,
+      "loss": 0.2624,
+      "step": 5695
+    },
+    {
+      "epoch": 1.9018363939899832,
+      "grad_norm": 0.4830513562553159,
+      "learning_rate": 7.320372195700532e-08,
+      "loss": 0.2551,
+      "step": 5696
+    },
+    {
+      "epoch": 1.902170283806344,
+      "grad_norm": 0.4599207374896554,
+      "learning_rate": 7.270779275633411e-08,
+      "loss": 0.2424,
+      "step": 5697
+    },
+    {
+      "epoch": 1.9025041736227046,
+      "grad_norm": 0.470358676246888,
+      "learning_rate": 7.22135368386201e-08,
+      "loss": 0.2412,
+      "step": 5698
+    },
+    {
+      "epoch": 1.9028380634390651,
+      "grad_norm": 0.477649088351064,
+      "learning_rate": 7.172095437171067e-08,
+      "loss": 0.2558,
+      "step": 5699
+    },
+    {
+      "epoch": 1.9031719532554257,
+      "grad_norm": 0.47918812268119354,
+      "learning_rate": 7.123004552288427e-08,
+      "loss": 0.2415,
+      "step": 5700
+    },
+    {
+      "epoch": 1.9035058430717862,
+      "grad_norm": 0.47042241057611783,
+      "learning_rate": 7.074081045885028e-08,
+      "loss": 0.2397,
+      "step": 5701
+    },
+    {
+      "epoch": 1.903839732888147,
+      "grad_norm": 0.4633073289845294,
+      "learning_rate": 7.02532493457514e-08,
+      "loss": 0.2431,
+      "step": 5702
+    },
+    {
+      "epoch": 1.9041736227045076,
+      "grad_norm": 0.46783605788085275,
+      "learning_rate": 6.97673623491607e-08,
+      "loss": 0.2384,
+      "step": 5703
+    },
+    {
+      "epoch": 1.9045075125208681,
+      "grad_norm": 0.47575481696873795,
+      "learning_rate": 6.928314963408234e-08,
+      "loss": 0.2475,
+      "step": 5704
+    },
+    {
+      "epoch": 1.9048414023372287,
+      "grad_norm": 0.4730341054480879,
+      "learning_rate": 6.880061136495252e-08,
+      "loss": 0.246,
+      "step": 5705
+    },
+    {
+      "epoch": 1.9051752921535892,
+      "grad_norm": 0.46211000798885177,
+      "learning_rate": 6.831974770563909e-08,
+      "loss": 0.2557,
+      "step": 5706
+    },
+    {
+      "epoch": 1.90550918196995,
+      "grad_norm": 0.47869371884935824,
+      "learning_rate": 6.784055881944085e-08,
+      "loss": 0.2445,
+      "step": 5707
+    },
+    {
+      "epoch": 1.9058430717863106,
+      "grad_norm": 0.45263736508512736,
+      "learning_rate": 6.736304486908763e-08,
+      "loss": 0.2486,
+      "step": 5708
+    },
+    {
+      "epoch": 1.9061769616026711,
+      "grad_norm": 0.45563967555125096,
+      "learning_rate": 6.688720601674137e-08,
+      "loss": 0.2302,
+      "step": 5709
+    },
+    {
+      "epoch": 1.9065108514190316,
+      "grad_norm": 0.4741280049514062,
+      "learning_rate": 6.641304242399282e-08,
+      "loss": 0.2492,
+      "step": 5710
+    },
+    {
+      "epoch": 1.9068447412353922,
+      "grad_norm": 0.4827650749102926,
+      "learning_rate": 6.594055425186763e-08,
+      "loss": 0.2388,
+      "step": 5711
+    },
+    {
+      "epoch": 1.907178631051753,
+      "grad_norm": 0.45912888584349526,
+      "learning_rate": 6.54697416608191e-08,
+      "loss": 0.2445,
+      "step": 5712
+    },
+    {
+      "epoch": 1.9075125208681136,
+      "grad_norm": 0.4637809233940702,
+      "learning_rate": 6.500060481073267e-08,
+      "loss": 0.2469,
+      "step": 5713
+    },
+    {
+      "epoch": 1.9078464106844741,
+      "grad_norm": 0.48654119169518356,
+      "learning_rate": 6.453314386092535e-08,
+      "loss": 0.2407,
+      "step": 5714
+    },
+    {
+      "epoch": 1.9081803005008346,
+      "grad_norm": 0.4575166067949966,
+      "learning_rate": 6.406735897014348e-08,
+      "loss": 0.245,
+      "step": 5715
+    },
+    {
+      "epoch": 1.9085141903171954,
+      "grad_norm": 0.47179149587982716,
+      "learning_rate": 6.360325029656667e-08,
+      "loss": 0.2534,
+      "step": 5716
+    },
+    {
+      "epoch": 1.908848080133556,
+      "grad_norm": 0.47910556225608447,
+      "learning_rate": 6.314081799780269e-08,
+      "loss": 0.252,
+      "step": 5717
+    },
+    {
+      "epoch": 1.9091819699499166,
+      "grad_norm": 0.46540834043157003,
+      "learning_rate": 6.268006223089207e-08,
+      "loss": 0.2461,
+      "step": 5718
+    },
+    {
+      "epoch": 1.9095158597662771,
+      "grad_norm": 0.4794144711489349,
+      "learning_rate": 6.222098315230407e-08,
+      "loss": 0.2502,
+      "step": 5719
+    },
+    {
+      "epoch": 1.9098497495826376,
+      "grad_norm": 0.47459013715637244,
+      "learning_rate": 6.176358091794011e-08,
+      "loss": 0.2464,
+      "step": 5720
+    },
+    {
+      "epoch": 1.9101836393989984,
+      "grad_norm": 0.49596498931959765,
+      "learning_rate": 6.130785568313202e-08,
+      "loss": 0.2554,
+      "step": 5721
+    },
+    {
+      "epoch": 1.910517529215359,
+      "grad_norm": 0.4684389059606607,
+      "learning_rate": 6.085380760264103e-08,
+      "loss": 0.249,
+      "step": 5722
+    },
+    {
+      "epoch": 1.9108514190317196,
+      "grad_norm": 0.4532032290717607,
+      "learning_rate": 6.040143683065936e-08,
+      "loss": 0.2452,
+      "step": 5723
+    },
+    {
+      "epoch": 1.9111853088480801,
+      "grad_norm": 0.4922837093690479,
+      "learning_rate": 5.995074352081021e-08,
+      "loss": 0.2526,
+      "step": 5724
+    },
+    {
+      "epoch": 1.9115191986644406,
+      "grad_norm": 0.4551400383244318,
+      "learning_rate": 5.950172782614672e-08,
+      "loss": 0.2388,
+      "step": 5725
+    },
+    {
+      "epoch": 1.9118530884808014,
+      "grad_norm": 0.47612561451161906,
+      "learning_rate": 5.905438989915246e-08,
+      "loss": 0.2544,
+      "step": 5726
+    },
+    {
+      "epoch": 1.9121869782971619,
+      "grad_norm": 0.46774172316670987,
+      "learning_rate": 5.860872989174149e-08,
+      "loss": 0.2299,
+      "step": 5727
+    },
+    {
+      "epoch": 1.9125208681135226,
+      "grad_norm": 0.4601537452599328,
+      "learning_rate": 5.8164747955256064e-08,
+      "loss": 0.2484,
+      "step": 5728
+    },
+    {
+      "epoch": 1.9128547579298831,
+      "grad_norm": 0.44624412740128566,
+      "learning_rate": 5.772244424047169e-08,
+      "loss": 0.2488,
+      "step": 5729
+    },
+    {
+      "epoch": 1.9131886477462436,
+      "grad_norm": 0.4656825353488079,
+      "learning_rate": 5.728181889759155e-08,
+      "loss": 0.2475,
+      "step": 5730
+    },
+    {
+      "epoch": 1.9135225375626044,
+      "grad_norm": 0.45368690033768166,
+      "learning_rate": 5.684287207625039e-08,
+      "loss": 0.2489,
+      "step": 5731
+    },
+    {
+      "epoch": 1.9138564273789649,
+      "grad_norm": 0.46113993152878374,
+      "learning_rate": 5.6405603925511196e-08,
+      "loss": 0.2336,
+      "step": 5732
+    },
+    {
+      "epoch": 1.9141903171953256,
+      "grad_norm": 0.47225297518418397,
+      "learning_rate": 5.597001459386908e-08,
+      "loss": 0.2497,
+      "step": 5733
+    },
+    {
+      "epoch": 1.9145242070116861,
+      "grad_norm": 0.46691272285755664,
+      "learning_rate": 5.5536104229247934e-08,
+      "loss": 0.246,
+      "step": 5734
+    },
+    {
+      "epoch": 1.9148580968280466,
+      "grad_norm": 0.4692960086796795,
+      "learning_rate": 5.510387297900099e-08,
+      "loss": 0.2423,
+      "step": 5735
+    },
+    {
+      "epoch": 1.9151919866444074,
+      "grad_norm": 0.46441078482080905,
+      "learning_rate": 5.46733209899114e-08,
+      "loss": 0.234,
+      "step": 5736
+    },
+    {
+      "epoch": 1.915525876460768,
+      "grad_norm": 0.46894079631936514,
+      "learning_rate": 5.4244448408192764e-08,
+      "loss": 0.2457,
+      "step": 5737
+    },
+    {
+      "epoch": 1.9158597662771286,
+      "grad_norm": 0.48003694590451595,
+      "learning_rate": 5.3817255379488565e-08,
+      "loss": 0.2563,
+      "step": 5738
+    },
+    {
+      "epoch": 1.9161936560934891,
+      "grad_norm": 0.4544856126179062,
+      "learning_rate": 5.33917420488711e-08,
+      "loss": 0.2364,
+      "step": 5739
+    },
+    {
+      "epoch": 1.9165275459098496,
+      "grad_norm": 0.49640983857884236,
+      "learning_rate": 5.2967908560841444e-08,
+      "loss": 0.2539,
+      "step": 5740
+    },
+    {
+      "epoch": 1.9168614357262104,
+      "grad_norm": 0.44532826241306017,
+      "learning_rate": 5.254575505933279e-08,
+      "loss": 0.2389,
+      "step": 5741
+    },
+    {
+      "epoch": 1.917195325542571,
+      "grad_norm": 0.4826228024709936,
+      "learning_rate": 5.2125281687704924e-08,
+      "loss": 0.2493,
+      "step": 5742
+    },
+    {
+      "epoch": 1.9175292153589316,
+      "grad_norm": 0.4681459803119626,
+      "learning_rate": 5.170648858874916e-08,
+      "loss": 0.2474,
+      "step": 5743
+    },
+    {
+      "epoch": 1.9178631051752921,
+      "grad_norm": 0.4848282815713978,
+      "learning_rate": 5.1289375904685637e-08,
+      "loss": 0.2473,
+      "step": 5744
+    },
+    {
+      "epoch": 1.9181969949916526,
+      "grad_norm": 0.4730419049380932,
+      "learning_rate": 5.0873943777163814e-08,
+      "loss": 0.2438,
+      "step": 5745
+    },
+    {
+      "epoch": 1.9185308848080134,
+      "grad_norm": 0.46535789302970526,
+      "learning_rate": 5.046019234726085e-08,
+      "loss": 0.2408,
+      "step": 5746
+    },
+    {
+      "epoch": 1.918864774624374,
+      "grad_norm": 0.46865627551496225,
+      "learning_rate": 5.004812175548657e-08,
+      "loss": 0.2441,
+      "step": 5747
+    },
+    {
+      "epoch": 1.9191986644407346,
+      "grad_norm": 0.47394501071015877,
+      "learning_rate": 4.963773214177681e-08,
+      "loss": 0.2535,
+      "step": 5748
+    },
+    {
+      "epoch": 1.9195325542570951,
+      "grad_norm": 0.4880146661287826,
+      "learning_rate": 4.922902364549731e-08,
+      "loss": 0.2533,
+      "step": 5749
+    },
+    {
+      "epoch": 1.9198664440734556,
+      "grad_norm": 0.4679497384717906,
+      "learning_rate": 4.8821996405444825e-08,
+      "loss": 0.2571,
+      "step": 5750
+    },
+    {
+      "epoch": 1.9202003338898164,
+      "grad_norm": 0.45504140648970814,
+      "learning_rate": 4.8416650559842685e-08,
+      "loss": 0.2408,
+      "step": 5751
+    },
+    {
+      "epoch": 1.920534223706177,
+      "grad_norm": 0.45626333410266456,
+      "learning_rate": 4.8012986246344653e-08,
+      "loss": 0.2391,
+      "step": 5752
+    },
+    {
+      "epoch": 1.9208681135225376,
+      "grad_norm": 0.4641916299109691,
+      "learning_rate": 4.761100360203274e-08,
+      "loss": 0.2474,
+      "step": 5753
+    },
+    {
+      "epoch": 1.9212020033388981,
+      "grad_norm": 0.46497365900294474,
+      "learning_rate": 4.7210702763418324e-08,
+      "loss": 0.2468,
+      "step": 5754
+    },
+    {
+      "epoch": 1.9215358931552586,
+      "grad_norm": 0.46487598102852623,
+      "learning_rate": 4.6812083866441535e-08,
+      "loss": 0.2441,
+      "step": 5755
+    },
+    {
+      "epoch": 1.9218697829716194,
+      "grad_norm": 0.44301500982161884,
+      "learning_rate": 4.641514704647132e-08,
+      "loss": 0.2424,
+      "step": 5756
+    },
+    {
+      "epoch": 1.92220367278798,
+      "grad_norm": 0.4635616041526578,
+      "learning_rate": 4.6019892438304845e-08,
+      "loss": 0.2433,
+      "step": 5757
+    },
+    {
+      "epoch": 1.9225375626043406,
+      "grad_norm": 0.47151642464306914,
+      "learning_rate": 4.562632017616975e-08,
+      "loss": 0.2462,
+      "step": 5758
+    },
+    {
+      "epoch": 1.9228714524207011,
+      "grad_norm": 0.4787289154024756,
+      "learning_rate": 4.523443039372022e-08,
+      "loss": 0.2447,
+      "step": 5759
+    },
+    {
+      "epoch": 1.9232053422370616,
+      "grad_norm": 0.4661806597871375,
+      "learning_rate": 4.484422322403981e-08,
+      "loss": 0.2437,
+      "step": 5760
+    },
+    {
+      "epoch": 1.9235392320534224,
+      "grad_norm": 0.4647416867124425,
+      "learning_rate": 4.44556987996414e-08,
+      "loss": 0.2423,
+      "step": 5761
+    },
+    {
+      "epoch": 1.923873121869783,
+      "grad_norm": 0.45380856266495045,
+      "learning_rate": 4.406885725246557e-08,
+      "loss": 0.2403,
+      "step": 5762
+    },
+    {
+      "epoch": 1.9242070116861436,
+      "grad_norm": 0.49338386976220894,
+      "learning_rate": 4.3683698713882226e-08,
+      "loss": 0.2572,
+      "step": 5763
+    },
+    {
+      "epoch": 1.9245409015025041,
+      "grad_norm": 0.4569055827771995,
+      "learning_rate": 4.3300223314688416e-08,
+      "loss": 0.2444,
+      "step": 5764
+    },
+    {
+      "epoch": 1.9248747913188646,
+      "grad_norm": 0.4683041919017558,
+      "learning_rate": 4.2918431185110523e-08,
+      "loss": 0.2394,
+      "step": 5765
+    },
+    {
+      "epoch": 1.9252086811352254,
+      "grad_norm": 0.46844312067542115,
+      "learning_rate": 4.253832245480372e-08,
+      "loss": 0.2452,
+      "step": 5766
+    },
+    {
+      "epoch": 1.925542570951586,
+      "grad_norm": 0.46183827688079426,
+      "learning_rate": 4.2159897252850854e-08,
+      "loss": 0.2509,
+      "step": 5767
+    },
+    {
+      "epoch": 1.9258764607679466,
+      "grad_norm": 0.4455901457880818,
+      "learning_rate": 4.178315570776303e-08,
+      "loss": 0.2312,
+      "step": 5768
+    },
+    {
+      "epoch": 1.9262103505843071,
+      "grad_norm": 0.48526425462046946,
+      "learning_rate": 4.140809794747958e-08,
+      "loss": 0.2517,
+      "step": 5769
+    },
+    {
+      "epoch": 1.9265442404006676,
+      "grad_norm": 0.4714544076367465,
+      "learning_rate": 4.103472409936859e-08,
+      "loss": 0.2505,
+      "step": 5770
+    },
+    {
+      "epoch": 1.9268781302170284,
+      "grad_norm": 0.45999029314227685,
+      "learning_rate": 4.0663034290225333e-08,
+      "loss": 0.2456,
+      "step": 5771
+    },
+    {
+      "epoch": 1.927212020033389,
+      "grad_norm": 0.44246842409563886,
+      "learning_rate": 4.029302864627438e-08,
+      "loss": 0.2404,
+      "step": 5772
+    },
+    {
+      "epoch": 1.9275459098497496,
+      "grad_norm": 0.46307628664581135,
+      "learning_rate": 3.992470729316689e-08,
+      "loss": 0.2524,
+      "step": 5773
+    },
+    {
+      "epoch": 1.9278797996661101,
+      "grad_norm": 0.45021706008606244,
+      "learning_rate": 3.9558070355983356e-08,
+      "loss": 0.2371,
+      "step": 5774
+    },
+    {
+      "epoch": 1.9282136894824706,
+      "grad_norm": 0.48703385369063684,
+      "learning_rate": 3.919311795923253e-08,
+      "loss": 0.2432,
+      "step": 5775
+    },
+    {
+      "epoch": 1.9285475792988314,
+      "grad_norm": 0.46711089364219577,
+      "learning_rate": 3.882985022684915e-08,
+      "loss": 0.2404,
+      "step": 5776
+    },
+    {
+      "epoch": 1.928881469115192,
+      "grad_norm": 0.4845607037612298,
+      "learning_rate": 3.846826728219788e-08,
+      "loss": 0.2437,
+      "step": 5777
+    },
+    {
+      "epoch": 1.9292153589315526,
+      "grad_norm": 0.4741975985614504,
+      "learning_rate": 3.8108369248069376e-08,
+      "loss": 0.2554,
+      "step": 5778
+    },
+    {
+      "epoch": 1.9295492487479131,
+      "grad_norm": 0.4645735274248853,
+      "learning_rate": 3.775015624668477e-08,
+      "loss": 0.248,
+      "step": 5779
+    },
+    {
+      "epoch": 1.9298831385642736,
+      "grad_norm": 0.48701166696918946,
+      "learning_rate": 3.739362839968952e-08,
+      "loss": 0.2461,
+      "step": 5780
+    },
+    {
+      "epoch": 1.9302170283806344,
+      "grad_norm": 0.4864679831230735,
+      "learning_rate": 3.703878582816012e-08,
+      "loss": 0.2415,
+      "step": 5781
+    },
+    {
+      "epoch": 1.930550918196995,
+      "grad_norm": 0.4782679430633318,
+      "learning_rate": 3.66856286525985e-08,
+      "loss": 0.2421,
+      "step": 5782
+    },
+    {
+      "epoch": 1.9308848080133556,
+      "grad_norm": 0.47043986223878503,
+      "learning_rate": 3.633415699293541e-08,
+      "loss": 0.2505,
+      "step": 5783
+    },
+    {
+      "epoch": 1.9312186978297161,
+      "grad_norm": 0.44698184751148856,
+      "learning_rate": 3.598437096852869e-08,
+      "loss": 0.2381,
+      "step": 5784
+    },
+    {
+      "epoch": 1.9315525876460768,
+      "grad_norm": 0.4653187490857565,
+      "learning_rate": 3.5636270698163336e-08,
+      "loss": 0.2471,
+      "step": 5785
+    },
+    {
+      "epoch": 1.9318864774624374,
+      "grad_norm": 0.46052971047567665,
+      "learning_rate": 3.528985630005366e-08,
+      "loss": 0.2486,
+      "step": 5786
+    },
+    {
+      "epoch": 1.932220367278798,
+      "grad_norm": 0.46605954090489143,
+      "learning_rate": 3.494512789183835e-08,
+      "loss": 0.251,
+      "step": 5787
+    },
+    {
+      "epoch": 1.9325542570951586,
+      "grad_norm": 0.4756456526183567,
+      "learning_rate": 3.4602085590587084e-08,
+      "loss": 0.2494,
+      "step": 5788
+    },
+    {
+      "epoch": 1.9328881469115191,
+      "grad_norm": 0.4639923261937635,
+      "learning_rate": 3.426072951279502e-08,
+      "loss": 0.2461,
+      "step": 5789
+    },
+    {
+      "epoch": 1.9332220367278798,
+      "grad_norm": 0.450277058509999,
+      "learning_rate": 3.392105977438387e-08,
+      "loss": 0.2478,
+      "step": 5790
+    },
+    {
+      "epoch": 1.9335559265442404,
+      "grad_norm": 0.4811236541697796,
+      "learning_rate": 3.3583076490704134e-08,
+      "loss": 0.2539,
+      "step": 5791
+    },
+    {
+      "epoch": 1.933889816360601,
+      "grad_norm": 0.467804718428422,
+      "learning_rate": 3.324677977653401e-08,
+      "loss": 0.2445,
+      "step": 5792
+    },
+    {
+      "epoch": 1.9342237061769616,
+      "grad_norm": 0.4643317494643002,
+      "learning_rate": 3.2912169746077136e-08,
+      "loss": 0.2405,
+      "step": 5793
+    },
+    {
+      "epoch": 1.934557595993322,
+      "grad_norm": 0.4839921921110155,
+      "learning_rate": 3.2579246512965954e-08,
+      "loss": 0.2619,
+      "step": 5794
+    },
+    {
+      "epoch": 1.9348914858096828,
+      "grad_norm": 0.45994693547892773,
+      "learning_rate": 3.224801019025947e-08,
+      "loss": 0.2429,
+      "step": 5795
+    },
+    {
+      "epoch": 1.9352253756260434,
+      "grad_norm": 0.46983344422040474,
+      "learning_rate": 3.1918460890443256e-08,
+      "loss": 0.2471,
+      "step": 5796
+    },
+    {
+      "epoch": 1.935559265442404,
+      "grad_norm": 0.4498105658975242,
+      "learning_rate": 3.159059872543113e-08,
+      "loss": 0.2324,
+      "step": 5797
+    },
+    {
+      "epoch": 1.9358931552587646,
+      "grad_norm": 0.47416597427113855,
+      "learning_rate": 3.126442380656292e-08,
+      "loss": 0.2538,
+      "step": 5798
+    },
+    {
+      "epoch": 1.936227045075125,
+      "grad_norm": 0.46764739376443876,
+      "learning_rate": 3.093993624460612e-08,
+      "loss": 0.2521,
+      "step": 5799
+    },
+    {
+      "epoch": 1.9365609348914858,
+      "grad_norm": 0.471900840141941,
+      "learning_rate": 3.0617136149755365e-08,
+      "loss": 0.2516,
+      "step": 5800
+    },
+    {
+      "epoch": 1.9368948247078466,
+      "grad_norm": 0.48123408389234984,
+      "learning_rate": 3.0296023631631866e-08,
+      "loss": 0.2571,
+      "step": 5801
+    },
+    {
+      "epoch": 1.937228714524207,
+      "grad_norm": 0.4408271254718044,
+      "learning_rate": 2.997659879928283e-08,
+      "loss": 0.2304,
+      "step": 5802
+    },
+    {
+      "epoch": 1.9375626043405676,
+      "grad_norm": 0.45809380110974746,
+      "learning_rate": 2.965886176118482e-08,
+      "loss": 0.2541,
+      "step": 5803
+    },
+    {
+      "epoch": 1.937896494156928,
+      "grad_norm": 0.4829477256933657,
+      "learning_rate": 2.9342812625238194e-08,
+      "loss": 0.2461,
+      "step": 5804
+    },
+    {
+      "epoch": 1.9382303839732888,
+      "grad_norm": 0.4609632171125203,
+      "learning_rate": 2.9028451498772646e-08,
+      "loss": 0.2428,
+      "step": 5805
+    },
+    {
+      "epoch": 1.9385642737896496,
+      "grad_norm": 0.46870301237318607,
+      "learning_rate": 2.8715778488543344e-08,
+      "loss": 0.2477,
+      "step": 5806
+    },
+    {
+      "epoch": 1.93889816360601,
+      "grad_norm": 0.4675529801069862,
+      "learning_rate": 2.8404793700732012e-08,
+      "loss": 0.2502,
+      "step": 5807
+    },
+    {
+      "epoch": 1.9392320534223706,
+      "grad_norm": 0.4642483058927234,
+      "learning_rate": 2.8095497240947512e-08,
+      "loss": 0.2454,
+      "step": 5808
+    },
+    {
+      "epoch": 1.939565943238731,
+      "grad_norm": 0.46348951533500776,
+      "learning_rate": 2.7787889214225817e-08,
+      "loss": 0.2437,
+      "step": 5809
+    },
+    {
+      "epoch": 1.9398998330550918,
+      "grad_norm": 0.47204934834457385,
+      "learning_rate": 2.7481969725028923e-08,
+      "loss": 0.2519,
+      "step": 5810
+    },
+    {
+      "epoch": 1.9402337228714526,
+      "grad_norm": 0.4856371626102723,
+      "learning_rate": 2.7177738877244842e-08,
+      "loss": 0.2521,
+      "step": 5811
+    },
+    {
+      "epoch": 1.940567612687813,
+      "grad_norm": 0.4545944632901646,
+      "learning_rate": 2.6875196774189815e-08,
+      "loss": 0.248,
+      "step": 5812
+    },
+    {
+      "epoch": 1.9409015025041736,
+      "grad_norm": 0.45460495951954255,
+      "learning_rate": 2.6574343518604995e-08,
+      "loss": 0.2431,
+      "step": 5813
+    },
+    {
+      "epoch": 1.941235392320534,
+      "grad_norm": 0.482248913338906,
+      "learning_rate": 2.6275179212658652e-08,
+      "loss": 0.2519,
+      "step": 5814
+    },
+    {
+      "epoch": 1.9415692821368948,
+      "grad_norm": 0.4750404733597846,
+      "learning_rate": 2.5977703957945632e-08,
+      "loss": 0.2536,
+      "step": 5815
+    },
+    {
+      "epoch": 1.9419031719532556,
+      "grad_norm": 0.4446431203127775,
+      "learning_rate": 2.5681917855486238e-08,
+      "loss": 0.2395,
+      "step": 5816
+    },
+    {
+      "epoch": 1.942237061769616,
+      "grad_norm": 0.4651928413401164,
+      "learning_rate": 2.5387821005729008e-08,
+      "loss": 0.2433,
+      "step": 5817
+    },
+    {
+      "epoch": 1.9425709515859766,
+      "grad_norm": 0.4744881762034826,
+      "learning_rate": 2.5095413508547385e-08,
+      "loss": 0.2557,
+      "step": 5818
+    },
+    {
+      "epoch": 1.942904841402337,
+      "grad_norm": 0.4538411145554794,
+      "learning_rate": 2.4804695463240825e-08,
+      "loss": 0.2368,
+      "step": 5819
+    },
+    {
+      "epoch": 1.9432387312186978,
+      "grad_norm": 0.4626069748498256,
+      "learning_rate": 2.451566696853591e-08,
+      "loss": 0.2485,
+      "step": 5820
+    },
+    {
+      "epoch": 1.9435726210350586,
+      "grad_norm": 0.46670729503058056,
+      "learning_rate": 2.4228328122585797e-08,
+      "loss": 0.2545,
+      "step": 5821
+    },
+    {
+      "epoch": 1.943906510851419,
+      "grad_norm": 0.47334229233138725,
+      "learning_rate": 2.394267902296854e-08,
+      "loss": 0.2499,
+      "step": 5822
+    },
+    {
+      "epoch": 1.9442404006677796,
+      "grad_norm": 0.47675692852266216,
+      "learning_rate": 2.3658719766689874e-08,
+      "loss": 0.2524,
+      "step": 5823
+    },
+    {
+      "epoch": 1.94457429048414,
+      "grad_norm": 0.4742423258331298,
+      "learning_rate": 2.3376450450179888e-08,
+      "loss": 0.2476,
+      "step": 5824
+    },
+    {
+      "epoch": 1.9449081803005008,
+      "grad_norm": 0.4621841070305756,
+      "learning_rate": 2.30958711692969e-08,
+      "loss": 0.239,
+      "step": 5825
+    },
+    {
+      "epoch": 1.9452420701168616,
+      "grad_norm": 0.47155667461686496,
+      "learning_rate": 2.281698201932303e-08,
+      "loss": 0.247,
+      "step": 5826
+    },
+    {
+      "epoch": 1.945575959933222,
+      "grad_norm": 0.46932203013461415,
+      "learning_rate": 2.253978309496918e-08,
+      "loss": 0.2388,
+      "step": 5827
+    },
+    {
+      "epoch": 1.9459098497495826,
+      "grad_norm": 0.48921260406456113,
+      "learning_rate": 2.226427449036894e-08,
+      "loss": 0.2526,
+      "step": 5828
+    },
+    {
+      "epoch": 1.946243739565943,
+      "grad_norm": 0.4575627011754293,
+      "learning_rate": 2.199045629908414e-08,
+      "loss": 0.2413,
+      "step": 5829
+    },
+    {
+      "epoch": 1.9465776293823038,
+      "grad_norm": 0.4667650063968615,
+      "learning_rate": 2.1718328614102614e-08,
+      "loss": 0.2513,
+      "step": 5830
+    },
+    {
+      "epoch": 1.9469115191986646,
+      "grad_norm": 0.4576257197958439,
+      "learning_rate": 2.1447891527837107e-08,
+      "loss": 0.2486,
+      "step": 5831
+    },
+    {
+      "epoch": 1.947245409015025,
+      "grad_norm": 0.46588105164297555,
+      "learning_rate": 2.1179145132126933e-08,
+      "loss": 0.2413,
+      "step": 5832
+    },
+    {
+      "epoch": 1.9475792988313856,
+      "grad_norm": 0.4596660002723031,
+      "learning_rate": 2.091208951823631e-08,
+      "loss": 0.2468,
+      "step": 5833
+    },
+    {
+      "epoch": 1.947913188647746,
+      "grad_norm": 0.4724860810691081,
+      "learning_rate": 2.0646724776856586e-08,
+      "loss": 0.2398,
+      "step": 5834
+    },
+    {
+      "epoch": 1.9482470784641068,
+      "grad_norm": 0.45980286412607907,
+      "learning_rate": 2.0383050998104004e-08,
+      "loss": 0.2387,
+      "step": 5835
+    },
+    {
+      "epoch": 1.9485809682804676,
+      "grad_norm": 0.48201915214601765,
+      "learning_rate": 2.0121068271520826e-08,
+      "loss": 0.2571,
+      "step": 5836
+    },
+    {
+      "epoch": 1.948914858096828,
+      "grad_norm": 0.464055431386661,
+      "learning_rate": 1.9860776686075332e-08,
+      "loss": 0.2388,
+      "step": 5837
+    },
+    {
+      "epoch": 1.9492487479131886,
+      "grad_norm": 0.45911001922518163,
+      "learning_rate": 1.960217633016015e-08,
+      "loss": 0.2435,
+      "step": 5838
+    },
+    {
+      "epoch": 1.949582637729549,
+      "grad_norm": 0.48039706375565766,
+      "learning_rate": 1.9345267291596136e-08,
+      "loss": 0.2482,
+      "step": 5839
+    },
+    {
+      "epoch": 1.9499165275459098,
+      "grad_norm": 0.459188851072753,
+      "learning_rate": 1.9090049657627397e-08,
+      "loss": 0.2413,
+      "step": 5840
+    },
+    {
+      "epoch": 1.9502504173622706,
+      "grad_norm": 0.476444402537618,
+      "learning_rate": 1.88365235149246e-08,
+      "loss": 0.2466,
+      "step": 5841
+    },
+    {
+      "epoch": 1.950584307178631,
+      "grad_norm": 0.47060592907935367,
+      "learning_rate": 1.8584688949583872e-08,
+      "loss": 0.2486,
+      "step": 5842
+    },
+    {
+      "epoch": 1.9509181969949916,
+      "grad_norm": 0.46318748990214603,
+      "learning_rate": 1.833454604712681e-08,
+      "loss": 0.2429,
+      "step": 5843
+    },
+    {
+      "epoch": 1.951252086811352,
+      "grad_norm": 0.47336528393720634,
+      "learning_rate": 1.8086094892501015e-08,
+      "loss": 0.2507,
+      "step": 5844
+    },
+    {
+      "epoch": 1.9515859766277128,
+      "grad_norm": 0.46082951739717704,
+      "learning_rate": 1.7839335570079553e-08,
+      "loss": 0.2492,
+      "step": 5845
+    },
+    {
+      "epoch": 1.9519198664440736,
+      "grad_norm": 0.4604825669540938,
+      "learning_rate": 1.7594268163659277e-08,
+      "loss": 0.2411,
+      "step": 5846
+    },
+    {
+      "epoch": 1.952253756260434,
+      "grad_norm": 0.4659557038143726,
+      "learning_rate": 1.7350892756465286e-08,
+      "loss": 0.2522,
+      "step": 5847
+    },
+    {
+      "epoch": 1.9525876460767946,
+      "grad_norm": 0.48090630926264244,
+      "learning_rate": 1.7109209431145913e-08,
+      "loss": 0.2491,
+      "step": 5848
+    },
+    {
+      "epoch": 1.9529215358931553,
+      "grad_norm": 0.46561025619608765,
+      "learning_rate": 1.68692182697755e-08,
+      "loss": 0.2383,
+      "step": 5849
+    },
+    {
+      "epoch": 1.9532554257095158,
+      "grad_norm": 0.46447329356324873,
+      "learning_rate": 1.663091935385386e-08,
+      "loss": 0.2398,
+      "step": 5850
+    },
+    {
+      "epoch": 1.9535893155258766,
+      "grad_norm": 0.4810007162476151,
+      "learning_rate": 1.639431276430625e-08,
+      "loss": 0.2578,
+      "step": 5851
+    },
+    {
+      "epoch": 1.953923205342237,
+      "grad_norm": 0.4684760888571553,
+      "learning_rate": 1.6159398581482855e-08,
+      "loss": 0.2451,
+      "step": 5852
+    },
+    {
+      "epoch": 1.9542570951585976,
+      "grad_norm": 0.47460736916319207,
+      "learning_rate": 1.5926176885159296e-08,
+      "loss": 0.252,
+      "step": 5853
+    },
+    {
+      "epoch": 1.9545909849749583,
+      "grad_norm": 0.47064940786797027,
+      "learning_rate": 1.5694647754537217e-08,
+      "loss": 0.2462,
+      "step": 5854
+    },
+    {
+      "epoch": 1.9549248747913188,
+      "grad_norm": 0.4748887471676456,
+      "learning_rate": 1.546481126824151e-08,
+      "loss": 0.2551,
+      "step": 5855
+    },
+    {
+      "epoch": 1.9552587646076796,
+      "grad_norm": 0.4759878589803303,
+      "learning_rate": 1.523666750432362e-08,
+      "loss": 0.2601,
+      "step": 5856
+    },
+    {
+      "epoch": 1.95559265442404,
+      "grad_norm": 0.4526935429818948,
+      "learning_rate": 1.5010216540261026e-08,
+      "loss": 0.2334,
+      "step": 5857
+    },
+    {
+      "epoch": 1.9559265442404006,
+      "grad_norm": 0.459573698906253,
+      "learning_rate": 1.4785458452954427e-08,
+      "loss": 0.2345,
+      "step": 5858
+    },
+    {
+      "epoch": 1.9562604340567613,
+      "grad_norm": 0.46521069254256997,
+      "learning_rate": 1.4562393318731106e-08,
+      "loss": 0.239,
+      "step": 5859
+    },
+    {
+      "epoch": 1.9565943238731218,
+      "grad_norm": 0.4728596399221697,
+      "learning_rate": 1.4341021213342133e-08,
+      "loss": 0.2532,
+      "step": 5860
+    },
+    {
+      "epoch": 1.9569282136894826,
+      "grad_norm": 0.4811712041469633,
+      "learning_rate": 1.4121342211964595e-08,
+      "loss": 0.2502,
+      "step": 5861
+    },
+    {
+      "epoch": 1.957262103505843,
+      "grad_norm": 0.4712728954347851,
+      "learning_rate": 1.3903356389200484e-08,
+      "loss": 0.2452,
+      "step": 5862
+    },
+    {
+      "epoch": 1.9575959933222036,
+      "grad_norm": 0.47051624928599806,
+      "learning_rate": 1.3687063819076697e-08,
+      "loss": 0.2499,
+      "step": 5863
+    },
+    {
+      "epoch": 1.9579298831385643,
+      "grad_norm": 0.4698555055145159,
+      "learning_rate": 1.347246457504503e-08,
+      "loss": 0.2472,
+      "step": 5864
+    },
+    {
+      "epoch": 1.9582637729549248,
+      "grad_norm": 0.46672855873502,
+      "learning_rate": 1.3259558729981637e-08,
+      "loss": 0.2383,
+      "step": 5865
+    },
+    {
+      "epoch": 1.9585976627712856,
+      "grad_norm": 0.4713325636088405,
+      "learning_rate": 1.3048346356189234e-08,
+      "loss": 0.2544,
+      "step": 5866
+    },
+    {
+      "epoch": 1.958931552587646,
+      "grad_norm": 0.4740520023161019,
+      "learning_rate": 1.2838827525394337e-08,
+      "loss": 0.2501,
+      "step": 5867
+    },
+    {
+      "epoch": 1.9592654424040066,
+      "grad_norm": 0.4728085108067724,
+      "learning_rate": 1.2631002308747809e-08,
+      "loss": 0.2378,
+      "step": 5868
+    },
+    {
+      "epoch": 1.9595993322203673,
+      "grad_norm": 0.4680335588661896,
+      "learning_rate": 1.2424870776826526e-08,
+      "loss": 0.2523,
+      "step": 5869
+    },
+    {
+      "epoch": 1.959933222036728,
+      "grad_norm": 0.462898729219732,
+      "learning_rate": 1.222043299963116e-08,
+      "loss": 0.2452,
+      "step": 5870
+    },
+    {
+      "epoch": 1.9602671118530886,
+      "grad_norm": 0.46809461106958683,
+      "learning_rate": 1.20176890465884e-08,
+      "loss": 0.2532,
+      "step": 5871
+    },
+    {
+      "epoch": 1.960601001669449,
+      "grad_norm": 0.4711307886246223,
+      "learning_rate": 1.1816638986548723e-08,
+      "loss": 0.2554,
+      "step": 5872
+    },
+    {
+      "epoch": 1.9609348914858096,
+      "grad_norm": 0.454683844864658,
+      "learning_rate": 1.1617282887787517e-08,
+      "loss": 0.2385,
+      "step": 5873
+    },
+    {
+      "epoch": 1.9612687813021703,
+      "grad_norm": 0.4680286091318821,
+      "learning_rate": 1.1419620818005072e-08,
+      "loss": 0.2487,
+      "step": 5874
+    },
+    {
+      "epoch": 1.961602671118531,
+      "grad_norm": 0.4630261119396645,
+      "learning_rate": 1.1223652844327137e-08,
+      "loss": 0.2397,
+      "step": 5875
+    },
+    {
+      "epoch": 1.9619365609348915,
+      "grad_norm": 0.4670271390530878,
+      "learning_rate": 1.1029379033302145e-08,
+      "loss": 0.235,
+      "step": 5876
+    },
+    {
+      "epoch": 1.962270450751252,
+      "grad_norm": 0.46010105814581365,
+      "learning_rate": 1.083679945090621e-08,
+      "loss": 0.2466,
+      "step": 5877
+    },
+    {
+      "epoch": 1.9626043405676126,
+      "grad_norm": 0.47173356660016,
+      "learning_rate": 1.0645914162536463e-08,
+      "loss": 0.2334,
+      "step": 5878
+    },
+    {
+      "epoch": 1.9629382303839733,
+      "grad_norm": 0.4635312760237499,
+      "learning_rate": 1.0456723233017718e-08,
+      "loss": 0.2439,
+      "step": 5879
+    },
+    {
+      "epoch": 1.963272120200334,
+      "grad_norm": 0.4750678490663704,
+      "learning_rate": 1.026922672659858e-08,
+      "loss": 0.2511,
+      "step": 5880
+    },
+    {
+      "epoch": 1.9636060100166945,
+      "grad_norm": 0.48695231094345914,
+      "learning_rate": 1.0083424706950894e-08,
+      "loss": 0.2605,
+      "step": 5881
+    },
+    {
+      "epoch": 1.963939899833055,
+      "grad_norm": 0.4664795344032291,
+      "learning_rate": 9.899317237172524e-09,
+      "loss": 0.248,
+      "step": 5882
+    },
+    {
+      "epoch": 1.9642737896494156,
+      "grad_norm": 0.4814379238232628,
+      "learning_rate": 9.71690437978512e-09,
+      "loss": 0.2556,
+      "step": 5883
+    },
+    {
+      "epoch": 1.9646076794657763,
+      "grad_norm": 0.4708202659383417,
+      "learning_rate": 9.536186196735242e-09,
+      "loss": 0.2563,
+      "step": 5884
+    },
+    {
+      "epoch": 1.964941569282137,
+      "grad_norm": 0.4510449821528043,
+      "learning_rate": 9.357162749394355e-09,
+      "loss": 0.2335,
+      "step": 5885
+    },
+    {
+      "epoch": 1.9652754590984975,
+      "grad_norm": 0.456420284512248,
+      "learning_rate": 9.17983409855716e-09,
+      "loss": 0.2412,
+      "step": 5886
+    },
+    {
+      "epoch": 1.965609348914858,
+      "grad_norm": 0.4629106103173813,
+      "learning_rate": 9.004200304444377e-09,
+      "loss": 0.238,
+      "step": 5887
+    },
+    {
+      "epoch": 1.9659432387312186,
+      "grad_norm": 0.46631154185642226,
+      "learning_rate": 8.83026142669996e-09,
+      "loss": 0.2415,
+      "step": 5888
+    },
+    {
+      "epoch": 1.9662771285475793,
+      "grad_norm": 0.44740727706048317,
+      "learning_rate": 8.658017524392216e-09,
+      "loss": 0.2436,
+      "step": 5889
+    },
+    {
+      "epoch": 1.96661101836394,
+      "grad_norm": 0.45453155277727325,
+      "learning_rate": 8.487468656014908e-09,
+      "loss": 0.2319,
+      "step": 5890
+    },
+    {
+      "epoch": 1.9669449081803005,
+      "grad_norm": 0.4714740842532806,
+      "learning_rate": 8.318614879485044e-09,
+      "loss": 0.2456,
+      "step": 5891
+    },
+    {
+      "epoch": 1.967278797996661,
+      "grad_norm": 0.47645285744714033,
+      "learning_rate": 8.151456252144529e-09,
+      "loss": 0.2522,
+      "step": 5892
+    },
+    {
+      "epoch": 1.9676126878130216,
+      "grad_norm": 0.459575721228396,
+      "learning_rate": 7.98599283076018e-09,
+      "loss": 0.2456,
+      "step": 5893
+    },
+    {
+      "epoch": 1.9679465776293823,
+      "grad_norm": 0.4548357697809658,
+      "learning_rate": 7.82222467152205e-09,
+      "loss": 0.2509,
+      "step": 5894
+    },
+    {
+      "epoch": 1.968280467445743,
+      "grad_norm": 0.47676020374443195,
+      "learning_rate": 7.660151830045093e-09,
+      "loss": 0.2462,
+      "step": 5895
+    },
+    {
+      "epoch": 1.9686143572621035,
+      "grad_norm": 0.45559828921355217,
+      "learning_rate": 7.499774361368062e-09,
+      "loss": 0.2313,
+      "step": 5896
+    },
+    {
+      "epoch": 1.968948247078464,
+      "grad_norm": 0.4796866933014068,
+      "learning_rate": 7.341092319954612e-09,
+      "loss": 0.2407,
+      "step": 5897
+    },
+    {
+      "epoch": 1.9692821368948246,
+      "grad_norm": 0.47442844068031526,
+      "learning_rate": 7.184105759692195e-09,
+      "loss": 0.242,
+      "step": 5898
+    },
+    {
+      "epoch": 1.9696160267111853,
+      "grad_norm": 0.4773658841646696,
+      "learning_rate": 7.028814733893163e-09,
+      "loss": 0.2497,
+      "step": 5899
+    },
+    {
+      "epoch": 1.969949916527546,
+      "grad_norm": 0.47160445176500027,
+      "learning_rate": 6.8752192952931115e-09,
+      "loss": 0.2521,
+      "step": 5900
+    },
+    {
+      "epoch": 1.9702838063439065,
+      "grad_norm": 0.4702541140209925,
+      "learning_rate": 6.723319496051983e-09,
+      "loss": 0.243,
+      "step": 5901
+    },
+    {
+      "epoch": 1.970617696160267,
+      "grad_norm": 0.46859087280078965,
+      "learning_rate": 6.573115387754625e-09,
+      "loss": 0.2395,
+      "step": 5902
+    },
+    {
+      "epoch": 1.9709515859766276,
+      "grad_norm": 0.4765179398353073,
+      "learning_rate": 6.424607021409679e-09,
+      "loss": 0.2489,
+      "step": 5903
+    },
+    {
+      "epoch": 1.9712854757929883,
+      "grad_norm": 0.46049312618398724,
+      "learning_rate": 6.277794447449026e-09,
+      "loss": 0.243,
+      "step": 5904
+    },
+    {
+      "epoch": 1.971619365609349,
+      "grad_norm": 0.4740796900565124,
+      "learning_rate": 6.132677715730562e-09,
+      "loss": 0.2553,
+      "step": 5905
+    },
+    {
+      "epoch": 1.9719532554257095,
+      "grad_norm": 0.46698037624022926,
+      "learning_rate": 5.98925687553431e-09,
+      "loss": 0.2495,
+      "step": 5906
+    },
+    {
+      "epoch": 1.97228714524207,
+      "grad_norm": 0.4839894831549759,
+      "learning_rate": 5.8475319755657566e-09,
+      "loss": 0.2454,
+      "step": 5907
+    },
+    {
+      "epoch": 1.9726210350584306,
+      "grad_norm": 0.46389883913103885,
+      "learning_rate": 5.707503063953623e-09,
+      "loss": 0.2536,
+      "step": 5908
+    },
+    {
+      "epoch": 1.9729549248747913,
+      "grad_norm": 0.4577660047463464,
+      "learning_rate": 5.569170188250983e-09,
+      "loss": 0.2518,
+      "step": 5909
+    },
+    {
+      "epoch": 1.973288814691152,
+      "grad_norm": 0.4860605741891927,
+      "learning_rate": 5.4325333954347024e-09,
+      "loss": 0.2474,
+      "step": 5910
+    },
+    {
+      "epoch": 1.9736227045075125,
+      "grad_norm": 0.46904939562316234,
+      "learning_rate": 5.297592731907109e-09,
+      "loss": 0.2402,
+      "step": 5911
+    },
+    {
+      "epoch": 1.973956594323873,
+      "grad_norm": 0.4637116023923323,
+      "learning_rate": 5.164348243492101e-09,
+      "loss": 0.2414,
+      "step": 5912
+    },
+    {
+      "epoch": 1.9742904841402336,
+      "grad_norm": 0.46871543177019054,
+      "learning_rate": 5.032799975439595e-09,
+      "loss": 0.2477,
+      "step": 5913
+    },
+    {
+      "epoch": 1.9746243739565943,
+      "grad_norm": 0.4766324962890092,
+      "learning_rate": 4.90294797242219e-09,
+      "loss": 0.2465,
+      "step": 5914
+    },
+    {
+      "epoch": 1.974958263772955,
+      "grad_norm": 0.4705039720367972,
+      "learning_rate": 4.7747922785373875e-09,
+      "loss": 0.2574,
+      "step": 5915
+    },
+    {
+      "epoch": 1.9752921535893155,
+      "grad_norm": 0.4647163665669312,
+      "learning_rate": 4.648332937305377e-09,
+      "loss": 0.2345,
+      "step": 5916
+    },
+    {
+      "epoch": 1.975626043405676,
+      "grad_norm": 0.4805515378877553,
+      "learning_rate": 4.523569991672361e-09,
+      "loss": 0.2594,
+      "step": 5917
+    },
+    {
+      "epoch": 1.9759599332220368,
+      "grad_norm": 0.46152983263187797,
+      "learning_rate": 4.4005034840061135e-09,
+      "loss": 0.2393,
+      "step": 5918
+    },
+    {
+      "epoch": 1.9762938230383973,
+      "grad_norm": 0.48561429140016044,
+      "learning_rate": 4.279133456099871e-09,
+      "loss": 0.2438,
+      "step": 5919
+    },
+    {
+      "epoch": 1.976627712854758,
+      "grad_norm": 0.44616705587307215,
+      "learning_rate": 4.159459949170108e-09,
+      "loss": 0.2432,
+      "step": 5920
+    },
+    {
+      "epoch": 1.9769616026711185,
+      "grad_norm": 0.4485539507814143,
+      "learning_rate": 4.041483003857649e-09,
+      "loss": 0.2346,
+      "step": 5921
+    },
+    {
+      "epoch": 1.977295492487479,
+      "grad_norm": 0.46473763603258283,
+      "learning_rate": 3.925202660226557e-09,
+      "loss": 0.2427,
+      "step": 5922
+    },
+    {
+      "epoch": 1.9776293823038398,
+      "grad_norm": 0.45023420134400616,
+      "learning_rate": 3.810618957765244e-09,
+      "loss": 0.2443,
+      "step": 5923
+    },
+    {
+      "epoch": 1.9779632721202003,
+      "grad_norm": 0.4617133744158428,
+      "learning_rate": 3.6977319353853626e-09,
+      "loss": 0.2431,
+      "step": 5924
+    },
+    {
+      "epoch": 1.978297161936561,
+      "grad_norm": 0.47325136273968227,
+      "learning_rate": 3.5865416314229127e-09,
+      "loss": 0.2387,
+      "step": 5925
+    },
+    {
+      "epoch": 1.9786310517529215,
+      "grad_norm": 0.4530758583325198,
+      "learning_rate": 3.4770480836376908e-09,
+      "loss": 0.2434,
+      "step": 5926
+    },
+    {
+      "epoch": 1.978964941569282,
+      "grad_norm": 0.48769116251670047,
+      "learning_rate": 3.3692513292132855e-09,
+      "loss": 0.2578,
+      "step": 5927
+    },
+    {
+      "epoch": 1.9792988313856428,
+      "grad_norm": 0.4863250878630164,
+      "learning_rate": 3.2631514047565262e-09,
+      "loss": 0.2512,
+      "step": 5928
+    },
+    {
+      "epoch": 1.9796327212020033,
+      "grad_norm": 0.46755630846424906,
+      "learning_rate": 3.15874834629859e-09,
+      "loss": 0.2449,
+      "step": 5929
+    },
+    {
+      "epoch": 1.979966611018364,
+      "grad_norm": 0.4854283579321504,
+      "learning_rate": 3.0560421892938952e-09,
+      "loss": 0.2535,
+      "step": 5930
+    },
+    {
+      "epoch": 1.9803005008347245,
+      "grad_norm": 0.45982848895427364,
+      "learning_rate": 2.9550329686217625e-09,
+      "loss": 0.2352,
+      "step": 5931
+    },
+    {
+      "epoch": 1.980634390651085,
+      "grad_norm": 0.4663419600474562,
+      "learning_rate": 2.8557207185836435e-09,
+      "loss": 0.2423,
+      "step": 5932
+    },
+    {
+      "epoch": 1.9809682804674458,
+      "grad_norm": 0.4573944576561235,
+      "learning_rate": 2.758105472905337e-09,
+      "loss": 0.2511,
+      "step": 5933
+    },
+    {
+      "epoch": 1.9813021702838065,
+      "grad_norm": 0.4778188121976251,
+      "learning_rate": 2.662187264737548e-09,
+      "loss": 0.2444,
+      "step": 5934
+    },
+    {
+      "epoch": 1.981636060100167,
+      "grad_norm": 0.4762262173041292,
+      "learning_rate": 2.5679661266525544e-09,
+      "loss": 0.2597,
+      "step": 5935
+    },
+    {
+      "epoch": 1.9819699499165275,
+      "grad_norm": 0.46490902872125367,
+      "learning_rate": 2.4754420906475396e-09,
+      "loss": 0.2579,
+      "step": 5936
+    },
+    {
+      "epoch": 1.982303839732888,
+      "grad_norm": 0.4937159944348199,
+      "learning_rate": 2.38461518814348e-09,
+      "loss": 0.2466,
+      "step": 5937
+    },
+    {
+      "epoch": 1.9826377295492488,
+      "grad_norm": 0.4592026282246009,
+      "learning_rate": 2.2954854499845914e-09,
+      "loss": 0.2441,
+      "step": 5938
+    },
+    {
+      "epoch": 1.9829716193656095,
+      "grad_norm": 0.4748220325431462,
+      "learning_rate": 2.2080529064388845e-09,
+      "loss": 0.2466,
+      "step": 5939
+    },
+    {
+      "epoch": 1.98330550918197,
+      "grad_norm": 0.46803355545115266,
+      "learning_rate": 2.122317587197609e-09,
+      "loss": 0.2461,
+      "step": 5940
+    },
+    {
+      "epoch": 1.9836393989983305,
+      "grad_norm": 0.4665182336739775,
+      "learning_rate": 2.038279521376918e-09,
+      "loss": 0.2502,
+      "step": 5941
+    },
+    {
+      "epoch": 1.983973288814691,
+      "grad_norm": 0.47684638153220843,
+      "learning_rate": 1.955938737515095e-09,
+      "loss": 0.251,
+      "step": 5942
+    },
+    {
+      "epoch": 1.9843071786310518,
+      "grad_norm": 0.4791603478209716,
+      "learning_rate": 1.875295263574772e-09,
+      "loss": 0.2568,
+      "step": 5943
+    },
+    {
+      "epoch": 1.9846410684474125,
+      "grad_norm": 0.46916047041926007,
+      "learning_rate": 1.796349126941821e-09,
+      "loss": 0.2503,
+      "step": 5944
+    },
+    {
+      "epoch": 1.984974958263773,
+      "grad_norm": 0.45275907423740264,
+      "learning_rate": 1.7191003544259067e-09,
+      "loss": 0.2417,
+      "step": 5945
+    },
+    {
+      "epoch": 1.9853088480801335,
+      "grad_norm": 0.45343953527669956,
+      "learning_rate": 1.6435489722610442e-09,
+      "loss": 0.2416,
+      "step": 5946
+    },
+    {
+      "epoch": 1.985642737896494,
+      "grad_norm": 0.4613701175053969,
+      "learning_rate": 1.5696950061033777e-09,
+      "loss": 0.2475,
+      "step": 5947
+    },
+    {
+      "epoch": 1.9859766277128548,
+      "grad_norm": 0.46373879893156145,
+      "learning_rate": 1.4975384810339555e-09,
+      "loss": 0.2481,
+      "step": 5948
+    },
+    {
+      "epoch": 1.9863105175292155,
+      "grad_norm": 0.47748664831746324,
+      "learning_rate": 1.4270794215559548e-09,
+      "loss": 0.2535,
+      "step": 5949
+    },
+    {
+      "epoch": 1.986644407345576,
+      "grad_norm": 0.45819173020684906,
+      "learning_rate": 1.358317851597457e-09,
+      "loss": 0.2452,
+      "step": 5950
+    },
+    {
+      "epoch": 1.9869782971619365,
+      "grad_norm": 0.4618580579049733,
+      "learning_rate": 1.2912537945092285e-09,
+      "loss": 0.2467,
+      "step": 5951
+    },
+    {
+      "epoch": 1.987312186978297,
+      "grad_norm": 0.4733452543393261,
+      "learning_rate": 1.225887273066384e-09,
+      "loss": 0.2475,
+      "step": 5952
+    },
+    {
+      "epoch": 1.9876460767946578,
+      "grad_norm": 0.481620817040103,
+      "learning_rate": 1.1622183094661677e-09,
+      "loss": 0.2464,
+      "step": 5953
+    },
+    {
+      "epoch": 1.9879799666110185,
+      "grad_norm": 0.4503779449165332,
+      "learning_rate": 1.100246925331283e-09,
+      "loss": 0.2457,
+      "step": 5954
+    },
+    {
+      "epoch": 1.988313856427379,
+      "grad_norm": 0.4694231252470685,
+      "learning_rate": 1.0399731417060077e-09,
+      "loss": 0.2465,
+      "step": 5955
+    },
+    {
+      "epoch": 1.9886477462437395,
+      "grad_norm": 0.47431963467994076,
+      "learning_rate": 9.813969790595235e-10,
+      "loss": 0.2515,
+      "step": 5956
+    },
+    {
+      "epoch": 1.9889816360601,
+      "grad_norm": 0.4508686644236103,
+      "learning_rate": 9.245184572842514e-10,
+      "loss": 0.2391,
+      "step": 5957
+    },
+    {
+      "epoch": 1.9893155258764608,
+      "grad_norm": 0.4810667441454546,
+      "learning_rate": 8.693375956947414e-10,
+      "loss": 0.2538,
+      "step": 5958
+    },
+    {
+      "epoch": 1.9896494156928215,
+      "grad_norm": 0.47922426706985793,
+      "learning_rate": 8.158544130315582e-10,
+      "loss": 0.2495,
+      "step": 5959
+    },
+    {
+      "epoch": 1.989983305509182,
+      "grad_norm": 0.471036572971478,
+      "learning_rate": 7.640689274562851e-10,
+      "loss": 0.2387,
+      "step": 5960
+    },
+    {
+      "epoch": 1.9903171953255425,
+      "grad_norm": 0.469185737508836,
+      "learning_rate": 7.139811565548549e-10,
+      "loss": 0.2438,
+      "step": 5961
+    },
+    {
+      "epoch": 1.990651085141903,
+      "grad_norm": 0.4589105906596277,
+      "learning_rate": 6.655911173381047e-10,
+      "loss": 0.2387,
+      "step": 5962
+    },
+    {
+      "epoch": 1.9909849749582638,
+      "grad_norm": 0.47891363548418103,
+      "learning_rate": 6.188988262373353e-10,
+      "loss": 0.2384,
+      "step": 5963
+    },
+    {
+      "epoch": 1.9913188647746245,
+      "grad_norm": 0.4732529692784966,
+      "learning_rate": 5.739042991104171e-10,
+      "loss": 0.2463,
+      "step": 5964
+    },
+    {
+      "epoch": 1.991652754590985,
+      "grad_norm": 0.4613619206229335,
+      "learning_rate": 5.306075512367947e-10,
+      "loss": 0.2469,
+      "step": 5965
+    },
+    {
+      "epoch": 1.9919866444073455,
+      "grad_norm": 0.46530427370492566,
+      "learning_rate": 4.890085973197067e-10,
+      "loss": 0.2382,
+      "step": 5966
+    },
+    {
+      "epoch": 1.992320534223706,
+      "grad_norm": 0.4534804615429222,
+      "learning_rate": 4.4910745148563083e-10,
+      "loss": 0.2495,
+      "step": 5967
+    },
+    {
+      "epoch": 1.9926544240400668,
+      "grad_norm": 0.46849274329145263,
+      "learning_rate": 4.1090412728539417e-10,
+      "loss": 0.2506,
+      "step": 5968
+    },
+    {
+      "epoch": 1.9929883138564275,
+      "grad_norm": 0.46631326910776516,
+      "learning_rate": 3.7439863769250793e-10,
+      "loss": 0.2521,
+      "step": 5969
+    },
+    {
+      "epoch": 1.993322203672788,
+      "grad_norm": 0.47240374559018067,
+      "learning_rate": 3.395909951037224e-10,
+      "loss": 0.2413,
+      "step": 5970
+    },
+    {
+      "epoch": 1.9936560934891485,
+      "grad_norm": 0.4721385470098938,
+      "learning_rate": 3.0648121133958206e-10,
+      "loss": 0.2471,
+      "step": 5971
+    },
+    {
+      "epoch": 1.993989983305509,
+      "grad_norm": 0.4783646518956141,
+      "learning_rate": 2.750692976444258e-10,
+      "loss": 0.242,
+      "step": 5972
+    },
+    {
+      "epoch": 1.9943238731218698,
+      "grad_norm": 0.4749067869116955,
+      "learning_rate": 2.4535526468527636e-10,
+      "loss": 0.2549,
+      "step": 5973
+    },
+    {
+      "epoch": 1.9946577629382305,
+      "grad_norm": 0.4674152460108366,
+      "learning_rate": 2.1733912255295087e-10,
+      "loss": 0.25,
+      "step": 5974
+    },
+    {
+      "epoch": 1.994991652754591,
+      "grad_norm": 0.44269039615707423,
+      "learning_rate": 1.910208807615055e-10,
+      "loss": 0.2462,
+      "step": 5975
+    },
+    {
+      "epoch": 1.9953255425709515,
+      "grad_norm": 0.47695410626998014,
+      "learning_rate": 1.664005482487907e-10,
+      "loss": 0.239,
+      "step": 5976
+    },
+    {
+      "epoch": 1.995659432387312,
+      "grad_norm": 0.44914098903590083,
+      "learning_rate": 1.43478133375341e-10,
+      "loss": 0.2412,
+      "step": 5977
+    },
+    {
+      "epoch": 1.9959933222036728,
+      "grad_norm": 0.4569057792956168,
+      "learning_rate": 1.2225364392548512e-10,
+      "loss": 0.2426,
+      "step": 5978
+    },
+    {
+      "epoch": 1.9963272120200335,
+      "grad_norm": 0.46068038008790746,
+      "learning_rate": 1.0272708710679091e-10,
+      "loss": 0.2462,
+      "step": 5979
+    },
+    {
+      "epoch": 1.996661101836394,
+      "grad_norm": 0.4777783691179534,
+      "learning_rate": 8.489846955117565e-11,
+      "loss": 0.2492,
+      "step": 5980
+    },
+    {
+      "epoch": 1.9969949916527545,
+      "grad_norm": 0.4658715119864357,
+      "learning_rate": 6.876779731213035e-11,
+      "loss": 0.2456,
+      "step": 5981
+    },
+    {
+      "epoch": 1.9973288814691152,
+      "grad_norm": 0.455124420968284,
+      "learning_rate": 5.4335075868605647e-11,
+      "loss": 0.2349,
+      "step": 5982
+    },
+    {
+      "epoch": 1.9976627712854758,
+      "grad_norm": 0.4642470660169114,
+      "learning_rate": 4.160031012057086e-11,
+      "loss": 0.258,
+      "step": 5983
+    },
+    {
+      "epoch": 1.9979966611018365,
+      "grad_norm": 0.46775612978558667,
+      "learning_rate": 3.056350439401001e-11,
+      "loss": 0.2457,
+      "step": 5984
+    },
+    {
+      "epoch": 1.998330550918197,
+      "grad_norm": 0.4808136049911246,
+      "learning_rate": 2.1224662436480913e-11,
+      "loss": 0.2612,
+      "step": 5985
+    },
+    {
+      "epoch": 1.9986644407345575,
+      "grad_norm": 0.4665559361338028,
+      "learning_rate": 1.3583787418780525e-11,
+      "loss": 0.2452,
+      "step": 5986
+    },
+    {
+      "epoch": 1.9989983305509182,
+      "grad_norm": 0.4696834373918854,
+      "learning_rate": 7.640881936610278e-12,
+      "loss": 0.2462,
+      "step": 5987
+    },
+    {
+      "epoch": 1.9993322203672788,
+      "grad_norm": 0.4714498960877185,
+      "learning_rate": 3.3959480078005203e-12,
+      "loss": 0.2533,
+      "step": 5988
+    },
+    {
+      "epoch": 1.9996661101836395,
+      "grad_norm": 0.47501665180593355,
+      "learning_rate": 8.489870739758487e-13,
+      "loss": 0.2598,
+      "step": 5989
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4676204565642214,
+      "learning_rate": 0.0,
+      "loss": 0.239,
+      "step": 5990
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.2709648013114929,
+      "eval_runtime": 405.584,
+      "eval_samples_per_second": 49.75,
+      "eval_steps_per_second": 0.39,
+      "step": 5990
+    },
+    {
+      "epoch": 2.0,
+      "step": 5990,
+      "total_flos": 2071185792860160.0,
+      "train_loss": 0.28316346206181436,
+      "train_runtime": 60100.1098,
+      "train_samples_per_second": 12.757,
+      "train_steps_per_second": 0.1
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5990,
+  "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": 2071185792860160.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}