diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24534 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6762631629794222,
+  "eval_steps": 500,
+  "global_step": 3500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019321804656554923,
+      "grad_norm": 26.163360595703125,
+      "learning_rate": 0.0,
+      "loss": 0.9633,
+      "step": 1
+    },
+    {
+      "epoch": 0.00038643609313109847,
+      "grad_norm": 29.663562774658203,
+      "learning_rate": 1.9305019305019306e-08,
+      "loss": 1.0219,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005796541396966477,
+      "grad_norm": 28.01646614074707,
+      "learning_rate": 3.861003861003861e-08,
+      "loss": 1.0184,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007728721862621969,
+      "grad_norm": 31.074716567993164,
+      "learning_rate": 5.791505791505792e-08,
+      "loss": 1.1334,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009660902328277461,
+      "grad_norm": 23.155662536621094,
+      "learning_rate": 7.722007722007723e-08,
+      "loss": 1.092,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011593082793932953,
+      "grad_norm": 26.564712524414062,
+      "learning_rate": 9.652509652509653e-08,
+      "loss": 1.1165,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013525263259588445,
+      "grad_norm": 28.848468780517578,
+      "learning_rate": 1.1583011583011584e-07,
+      "loss": 1.0429,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015457443725243939,
+      "grad_norm": 26.71477508544922,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 0.9717,
+      "step": 8
+    },
+    {
+      "epoch": 0.001738962419089943,
+      "grad_norm": 25.45704460144043,
+      "learning_rate": 1.5444015444015445e-07,
+      "loss": 0.9446,
+      "step": 9
+    },
+    {
+      "epoch": 0.0019321804656554922,
+      "grad_norm": 25.267148971557617,
+      "learning_rate": 1.7374517374517376e-07,
+      "loss": 1.0087,
+      "step": 10
+    },
+    {
+      "epoch": 0.0021253985122210413,
+      "grad_norm": 29.721664428710938,
+      "learning_rate": 1.9305019305019306e-07,
+      "loss": 1.01,
+      "step": 11
+    },
+    {
+      "epoch": 0.0023186165587865907,
+      "grad_norm": 31.159250259399414,
+      "learning_rate": 2.1235521235521237e-07,
+      "loss": 0.9193,
+      "step": 12
+    },
+    {
+      "epoch": 0.00251183460535214,
+      "grad_norm": 31.521038055419922,
+      "learning_rate": 2.3166023166023168e-07,
+      "loss": 1.0528,
+      "step": 13
+    },
+    {
+      "epoch": 0.002705052651917689,
+      "grad_norm": 28.082170486450195,
+      "learning_rate": 2.5096525096525096e-07,
+      "loss": 1.0206,
+      "step": 14
+    },
+    {
+      "epoch": 0.0028982706984832384,
+      "grad_norm": 30.327272415161133,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 0.9312,
+      "step": 15
+    },
+    {
+      "epoch": 0.0030914887450487877,
+      "grad_norm": 24.757890701293945,
+      "learning_rate": 2.895752895752896e-07,
+      "loss": 0.9209,
+      "step": 16
+    },
+    {
+      "epoch": 0.0032847067916143367,
+      "grad_norm": 25.048492431640625,
+      "learning_rate": 3.088803088803089e-07,
+      "loss": 0.8759,
+      "step": 17
+    },
+    {
+      "epoch": 0.003477924838179886,
+      "grad_norm": 25.817691802978516,
+      "learning_rate": 3.281853281853282e-07,
+      "loss": 0.9557,
+      "step": 18
+    },
+    {
+      "epoch": 0.0036711428847454354,
+      "grad_norm": 22.10857582092285,
+      "learning_rate": 3.474903474903475e-07,
+      "loss": 0.7857,
+      "step": 19
+    },
+    {
+      "epoch": 0.0038643609313109843,
+      "grad_norm": 24.98918914794922,
+      "learning_rate": 3.6679536679536685e-07,
+      "loss": 0.903,
+      "step": 20
+    },
+    {
+      "epoch": 0.004057578977876533,
+      "grad_norm": 21.72258186340332,
+      "learning_rate": 3.8610038610038613e-07,
+      "loss": 0.8668,
+      "step": 21
+    },
+    {
+      "epoch": 0.004250797024442083,
+      "grad_norm": 14.98244857788086,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 0.796,
+      "step": 22
+    },
+    {
+      "epoch": 0.004444015071007632,
+      "grad_norm": 12.133209228515625,
+      "learning_rate": 4.2471042471042474e-07,
+      "loss": 0.7185,
+      "step": 23
+    },
+    {
+      "epoch": 0.004637233117573181,
+      "grad_norm": 4.995375156402588,
+      "learning_rate": 4.440154440154441e-07,
+      "loss": 0.726,
+      "step": 24
+    },
+    {
+      "epoch": 0.004830451164138731,
+      "grad_norm": 5.531341075897217,
+      "learning_rate": 4.6332046332046336e-07,
+      "loss": 0.818,
+      "step": 25
+    },
+    {
+      "epoch": 0.00502366921070428,
+      "grad_norm": 5.521533966064453,
+      "learning_rate": 4.826254826254826e-07,
+      "loss": 0.8053,
+      "step": 26
+    },
+    {
+      "epoch": 0.005216887257269829,
+      "grad_norm": 4.053627014160156,
+      "learning_rate": 5.019305019305019e-07,
+      "loss": 0.7399,
+      "step": 27
+    },
+    {
+      "epoch": 0.005410105303835378,
+      "grad_norm": 5.711602210998535,
+      "learning_rate": 5.212355212355213e-07,
+      "loss": 0.7507,
+      "step": 28
+    },
+    {
+      "epoch": 0.005603323350400927,
+      "grad_norm": 3.0087835788726807,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.7232,
+      "step": 29
+    },
+    {
+      "epoch": 0.005796541396966477,
+      "grad_norm": 1.4082857370376587,
+      "learning_rate": 5.598455598455599e-07,
+      "loss": 0.6894,
+      "step": 30
+    },
+    {
+      "epoch": 0.005989759443532026,
+      "grad_norm": 1.487677812576294,
+      "learning_rate": 5.791505791505792e-07,
+      "loss": 0.7559,
+      "step": 31
+    },
+    {
+      "epoch": 0.0061829774900975755,
+      "grad_norm": 1.2217342853546143,
+      "learning_rate": 5.984555984555985e-07,
+      "loss": 0.7427,
+      "step": 32
+    },
+    {
+      "epoch": 0.006376195536663124,
+      "grad_norm": 1.3117166757583618,
+      "learning_rate": 6.177606177606178e-07,
+      "loss": 0.7047,
+      "step": 33
+    },
+    {
+      "epoch": 0.006569413583228673,
+      "grad_norm": 1.2920383214950562,
+      "learning_rate": 6.370656370656371e-07,
+      "loss": 0.7732,
+      "step": 34
+    },
+    {
+      "epoch": 0.006762631629794223,
+      "grad_norm": 1.2037408351898193,
+      "learning_rate": 6.563706563706564e-07,
+      "loss": 0.6421,
+      "step": 35
+    },
+    {
+      "epoch": 0.006955849676359772,
+      "grad_norm": 1.1959179639816284,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 0.7019,
+      "step": 36
+    },
+    {
+      "epoch": 0.0071490677229253214,
+      "grad_norm": 1.0364854335784912,
+      "learning_rate": 6.94980694980695e-07,
+      "loss": 0.7732,
+      "step": 37
+    },
+    {
+      "epoch": 0.007342285769490871,
+      "grad_norm": 1.1884748935699463,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.7032,
+      "step": 38
+    },
+    {
+      "epoch": 0.007535503816056419,
+      "grad_norm": 1.0902557373046875,
+      "learning_rate": 7.335907335907337e-07,
+      "loss": 0.6941,
+      "step": 39
+    },
+    {
+      "epoch": 0.007728721862621969,
+      "grad_norm": 1.1322697401046753,
+      "learning_rate": 7.52895752895753e-07,
+      "loss": 0.686,
+      "step": 40
+    },
+    {
+      "epoch": 0.007921939909187518,
+      "grad_norm": 1.0249947309494019,
+      "learning_rate": 7.722007722007723e-07,
+      "loss": 0.7382,
+      "step": 41
+    },
+    {
+      "epoch": 0.008115157955753067,
+      "grad_norm": 0.8608828186988831,
+      "learning_rate": 7.915057915057915e-07,
+      "loss": 0.6251,
+      "step": 42
+    },
+    {
+      "epoch": 0.008308376002318617,
+      "grad_norm": 0.9509369730949402,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.7057,
+      "step": 43
+    },
+    {
+      "epoch": 0.008501594048884165,
+      "grad_norm": 0.8577322959899902,
+      "learning_rate": 8.301158301158302e-07,
+      "loss": 0.6911,
+      "step": 44
+    },
+    {
+      "epoch": 0.008694812095449716,
+      "grad_norm": 0.8552514314651489,
+      "learning_rate": 8.494208494208495e-07,
+      "loss": 0.6617,
+      "step": 45
+    },
+    {
+      "epoch": 0.008888030142015264,
+      "grad_norm": 0.8920540809631348,
+      "learning_rate": 8.687258687258689e-07,
+      "loss": 0.6863,
+      "step": 46
+    },
+    {
+      "epoch": 0.009081248188580814,
+      "grad_norm": 0.8744629621505737,
+      "learning_rate": 8.880308880308882e-07,
+      "loss": 0.6728,
+      "step": 47
+    },
+    {
+      "epoch": 0.009274466235146363,
+      "grad_norm": 0.797126293182373,
+      "learning_rate": 9.073359073359074e-07,
+      "loss": 0.6683,
+      "step": 48
+    },
+    {
+      "epoch": 0.009467684281711911,
+      "grad_norm": 0.9033235907554626,
+      "learning_rate": 9.266409266409267e-07,
+      "loss": 0.7079,
+      "step": 49
+    },
+    {
+      "epoch": 0.009660902328277461,
+      "grad_norm": 0.8397656083106995,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 0.6752,
+      "step": 50
+    },
+    {
+      "epoch": 0.00985412037484301,
+      "grad_norm": 0.9273586273193359,
+      "learning_rate": 9.652509652509653e-07,
+      "loss": 0.7305,
+      "step": 51
+    },
+    {
+      "epoch": 0.01004733842140856,
+      "grad_norm": 0.868870198726654,
+      "learning_rate": 9.845559845559847e-07,
+      "loss": 0.7118,
+      "step": 52
+    },
+    {
+      "epoch": 0.010240556467974109,
+      "grad_norm": 0.8337823152542114,
+      "learning_rate": 1.0038610038610038e-06,
+      "loss": 0.6904,
+      "step": 53
+    },
+    {
+      "epoch": 0.010433774514539657,
+      "grad_norm": 0.7683207988739014,
+      "learning_rate": 1.0231660231660232e-06,
+      "loss": 0.7228,
+      "step": 54
+    },
+    {
+      "epoch": 0.010626992561105207,
+      "grad_norm": 0.8864586353302002,
+      "learning_rate": 1.0424710424710426e-06,
+      "loss": 0.64,
+      "step": 55
+    },
+    {
+      "epoch": 0.010820210607670756,
+      "grad_norm": 0.7800060510635376,
+      "learning_rate": 1.0617760617760618e-06,
+      "loss": 0.6449,
+      "step": 56
+    },
+    {
+      "epoch": 0.011013428654236306,
+      "grad_norm": 0.8159433007240295,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.6899,
+      "step": 57
+    },
+    {
+      "epoch": 0.011206646700801855,
+      "grad_norm": 0.8030359148979187,
+      "learning_rate": 1.1003861003861005e-06,
+      "loss": 0.6566,
+      "step": 58
+    },
+    {
+      "epoch": 0.011399864747367405,
+      "grad_norm": 0.8253495097160339,
+      "learning_rate": 1.1196911196911197e-06,
+      "loss": 0.6994,
+      "step": 59
+    },
+    {
+      "epoch": 0.011593082793932953,
+      "grad_norm": 0.8272079825401306,
+      "learning_rate": 1.1389961389961391e-06,
+      "loss": 0.6586,
+      "step": 60
+    },
+    {
+      "epoch": 0.011786300840498502,
+      "grad_norm": 0.8360865116119385,
+      "learning_rate": 1.1583011583011585e-06,
+      "loss": 0.6525,
+      "step": 61
+    },
+    {
+      "epoch": 0.011979518887064052,
+      "grad_norm": 0.746075451374054,
+      "learning_rate": 1.1776061776061777e-06,
+      "loss": 0.6641,
+      "step": 62
+    },
+    {
+      "epoch": 0.0121727369336296,
+      "grad_norm": 0.8202755451202393,
+      "learning_rate": 1.196911196911197e-06,
+      "loss": 0.7259,
+      "step": 63
+    },
+    {
+      "epoch": 0.012365954980195151,
+      "grad_norm": 0.8042411804199219,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 0.6618,
+      "step": 64
+    },
+    {
+      "epoch": 0.0125591730267607,
+      "grad_norm": 0.7820135354995728,
+      "learning_rate": 1.2355212355212356e-06,
+      "loss": 0.7134,
+      "step": 65
+    },
+    {
+      "epoch": 0.012752391073326248,
+      "grad_norm": 0.743039608001709,
+      "learning_rate": 1.254826254826255e-06,
+      "loss": 0.6557,
+      "step": 66
+    },
+    {
+      "epoch": 0.012945609119891798,
+      "grad_norm": 0.7439308762550354,
+      "learning_rate": 1.2741312741312742e-06,
+      "loss": 0.6222,
+      "step": 67
+    },
+    {
+      "epoch": 0.013138827166457347,
+      "grad_norm": 0.7946411371231079,
+      "learning_rate": 1.2934362934362933e-06,
+      "loss": 0.6498,
+      "step": 68
+    },
+    {
+      "epoch": 0.013332045213022897,
+      "grad_norm": 0.8667552471160889,
+      "learning_rate": 1.3127413127413127e-06,
+      "loss": 0.7377,
+      "step": 69
+    },
+    {
+      "epoch": 0.013525263259588445,
+      "grad_norm": 0.8318952322006226,
+      "learning_rate": 1.3320463320463323e-06,
+      "loss": 0.742,
+      "step": 70
+    },
+    {
+      "epoch": 0.013718481306153996,
+      "grad_norm": 0.7701277136802673,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.6799,
+      "step": 71
+    },
+    {
+      "epoch": 0.013911699352719544,
+      "grad_norm": 0.7692044973373413,
+      "learning_rate": 1.3706563706563709e-06,
+      "loss": 0.6675,
+      "step": 72
+    },
+    {
+      "epoch": 0.014104917399285093,
+      "grad_norm": 0.7828958630561829,
+      "learning_rate": 1.38996138996139e-06,
+      "loss": 0.6818,
+      "step": 73
+    },
+    {
+      "epoch": 0.014298135445850643,
+      "grad_norm": 0.8129189014434814,
+      "learning_rate": 1.4092664092664092e-06,
+      "loss": 0.6802,
+      "step": 74
+    },
+    {
+      "epoch": 0.014491353492416191,
+      "grad_norm": 0.7519441246986389,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.5872,
+      "step": 75
+    },
+    {
+      "epoch": 0.014684571538981742,
+      "grad_norm": 0.768528163433075,
+      "learning_rate": 1.4478764478764478e-06,
+      "loss": 0.6383,
+      "step": 76
+    },
+    {
+      "epoch": 0.01487778958554729,
+      "grad_norm": 0.7637743949890137,
+      "learning_rate": 1.4671814671814674e-06,
+      "loss": 0.6191,
+      "step": 77
+    },
+    {
+      "epoch": 0.015071007632112839,
+      "grad_norm": 0.798071563243866,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 0.7289,
+      "step": 78
+    },
+    {
+      "epoch": 0.015264225678678389,
+      "grad_norm": 0.7936199903488159,
+      "learning_rate": 1.505791505791506e-06,
+      "loss": 0.6684,
+      "step": 79
+    },
+    {
+      "epoch": 0.015457443725243937,
+      "grad_norm": 0.8443831205368042,
+      "learning_rate": 1.5250965250965251e-06,
+      "loss": 0.7054,
+      "step": 80
+    },
+    {
+      "epoch": 0.015650661771809488,
+      "grad_norm": 0.7631838321685791,
+      "learning_rate": 1.5444015444015445e-06,
+      "loss": 0.6723,
+      "step": 81
+    },
+    {
+      "epoch": 0.015843879818375036,
+      "grad_norm": 0.7617886066436768,
+      "learning_rate": 1.5637065637065637e-06,
+      "loss": 0.67,
+      "step": 82
+    },
+    {
+      "epoch": 0.016037097864940585,
+      "grad_norm": 0.8348072171211243,
+      "learning_rate": 1.583011583011583e-06,
+      "loss": 0.7122,
+      "step": 83
+    },
+    {
+      "epoch": 0.016230315911506133,
+      "grad_norm": 0.7492917776107788,
+      "learning_rate": 1.6023166023166027e-06,
+      "loss": 0.6724,
+      "step": 84
+    },
+    {
+      "epoch": 0.016423533958071685,
+      "grad_norm": 0.7187228798866272,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.6602,
+      "step": 85
+    },
+    {
+      "epoch": 0.016616752004637234,
+      "grad_norm": 0.7930755615234375,
+      "learning_rate": 1.640926640926641e-06,
+      "loss": 0.683,
+      "step": 86
+    },
+    {
+      "epoch": 0.016809970051202782,
+      "grad_norm": 0.7337602972984314,
+      "learning_rate": 1.6602316602316604e-06,
+      "loss": 0.6585,
+      "step": 87
+    },
+    {
+      "epoch": 0.01700318809776833,
+      "grad_norm": 0.6961308121681213,
+      "learning_rate": 1.6795366795366796e-06,
+      "loss": 0.684,
+      "step": 88
+    },
+    {
+      "epoch": 0.01719640614433388,
+      "grad_norm": 0.744563639163971,
+      "learning_rate": 1.698841698841699e-06,
+      "loss": 0.5808,
+      "step": 89
+    },
+    {
+      "epoch": 0.01738962419089943,
+      "grad_norm": 0.7431725263595581,
+      "learning_rate": 1.7181467181467181e-06,
+      "loss": 0.6767,
+      "step": 90
+    },
+    {
+      "epoch": 0.01758284223746498,
+      "grad_norm": 0.7845579981803894,
+      "learning_rate": 1.7374517374517377e-06,
+      "loss": 0.6403,
+      "step": 91
+    },
+    {
+      "epoch": 0.017776060284030528,
+      "grad_norm": 0.7574836015701294,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 0.6502,
+      "step": 92
+    },
+    {
+      "epoch": 0.017969278330596077,
+      "grad_norm": 0.7749143838882446,
+      "learning_rate": 1.7760617760617763e-06,
+      "loss": 0.6884,
+      "step": 93
+    },
+    {
+      "epoch": 0.01816249637716163,
+      "grad_norm": 0.8078687191009521,
+      "learning_rate": 1.7953667953667955e-06,
+      "loss": 0.5869,
+      "step": 94
+    },
+    {
+      "epoch": 0.018355714423727177,
+      "grad_norm": 0.7935137748718262,
+      "learning_rate": 1.8146718146718149e-06,
+      "loss": 0.6864,
+      "step": 95
+    },
+    {
+      "epoch": 0.018548932470292726,
+      "grad_norm": 0.7307217121124268,
+      "learning_rate": 1.833976833976834e-06,
+      "loss": 0.6798,
+      "step": 96
+    },
+    {
+      "epoch": 0.018742150516858274,
+      "grad_norm": 0.681380569934845,
+      "learning_rate": 1.8532818532818534e-06,
+      "loss": 0.7073,
+      "step": 97
+    },
+    {
+      "epoch": 0.018935368563423823,
+      "grad_norm": 0.7885120511054993,
+      "learning_rate": 1.8725868725868726e-06,
+      "loss": 0.6973,
+      "step": 98
+    },
+    {
+      "epoch": 0.019128586609989374,
+      "grad_norm": 0.7846948504447937,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.6716,
+      "step": 99
+    },
+    {
+      "epoch": 0.019321804656554923,
+      "grad_norm": 0.6699268221855164,
+      "learning_rate": 1.9111969111969114e-06,
+      "loss": 0.6496,
+      "step": 100
+    },
+    {
+      "epoch": 0.01951502270312047,
+      "grad_norm": 0.7176694869995117,
+      "learning_rate": 1.9305019305019305e-06,
+      "loss": 0.6462,
+      "step": 101
+    },
+    {
+      "epoch": 0.01970824074968602,
+      "grad_norm": 0.800673246383667,
+      "learning_rate": 1.94980694980695e-06,
+      "loss": 0.7165,
+      "step": 102
+    },
+    {
+      "epoch": 0.01990145879625157,
+      "grad_norm": 0.8949143886566162,
+      "learning_rate": 1.9691119691119693e-06,
+      "loss": 0.6973,
+      "step": 103
+    },
+    {
+      "epoch": 0.02009467684281712,
+      "grad_norm": 0.7779495120048523,
+      "learning_rate": 1.9884169884169885e-06,
+      "loss": 0.6584,
+      "step": 104
+    },
+    {
+      "epoch": 0.02028789488938267,
+      "grad_norm": 0.7875708937644958,
+      "learning_rate": 2.0077220077220077e-06,
+      "loss": 0.6456,
+      "step": 105
+    },
+    {
+      "epoch": 0.020481112935948217,
+      "grad_norm": 0.8520517349243164,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 0.7009,
+      "step": 106
+    },
+    {
+      "epoch": 0.020674330982513766,
+      "grad_norm": 0.7782955765724182,
+      "learning_rate": 2.0463320463320464e-06,
+      "loss": 0.7399,
+      "step": 107
+    },
+    {
+      "epoch": 0.020867549029079314,
+      "grad_norm": 0.7730876207351685,
+      "learning_rate": 2.065637065637066e-06,
+      "loss": 0.664,
+      "step": 108
+    },
+    {
+      "epoch": 0.021060767075644866,
+      "grad_norm": 0.7593525052070618,
+      "learning_rate": 2.084942084942085e-06,
+      "loss": 0.6507,
+      "step": 109
+    },
+    {
+      "epoch": 0.021253985122210415,
+      "grad_norm": 0.8029621243476868,
+      "learning_rate": 2.1042471042471044e-06,
+      "loss": 0.7209,
+      "step": 110
+    },
+    {
+      "epoch": 0.021447203168775963,
+      "grad_norm": 0.8133061528205872,
+      "learning_rate": 2.1235521235521236e-06,
+      "loss": 0.7219,
+      "step": 111
+    },
+    {
+      "epoch": 0.021640421215341512,
+      "grad_norm": 0.7595868706703186,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 0.6665,
+      "step": 112
+    },
+    {
+      "epoch": 0.02183363926190706,
+      "grad_norm": 0.7699278593063354,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.6771,
+      "step": 113
+    },
+    {
+      "epoch": 0.022026857308472612,
+      "grad_norm": 0.7549155354499817,
+      "learning_rate": 2.181467181467182e-06,
+      "loss": 0.7265,
+      "step": 114
+    },
+    {
+      "epoch": 0.02222007535503816,
+      "grad_norm": 0.7687058448791504,
+      "learning_rate": 2.200772200772201e-06,
+      "loss": 0.6549,
+      "step": 115
+    },
+    {
+      "epoch": 0.02241329340160371,
+      "grad_norm": 0.7540101408958435,
+      "learning_rate": 2.2200772200772203e-06,
+      "loss": 0.6287,
+      "step": 116
+    },
+    {
+      "epoch": 0.022606511448169258,
+      "grad_norm": 0.735726535320282,
+      "learning_rate": 2.2393822393822394e-06,
+      "loss": 0.5984,
+      "step": 117
+    },
+    {
+      "epoch": 0.02279972949473481,
+      "grad_norm": 0.713107705116272,
+      "learning_rate": 2.2586872586872586e-06,
+      "loss": 0.6068,
+      "step": 118
+    },
+    {
+      "epoch": 0.02299294754130036,
+      "grad_norm": 0.8987348079681396,
+      "learning_rate": 2.2779922779922782e-06,
+      "loss": 0.6887,
+      "step": 119
+    },
+    {
+      "epoch": 0.023186165587865907,
+      "grad_norm": 0.756963849067688,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 0.6838,
+      "step": 120
+    },
+    {
+      "epoch": 0.023379383634431455,
+      "grad_norm": 0.7681220769882202,
+      "learning_rate": 2.316602316602317e-06,
+      "loss": 0.6571,
+      "step": 121
+    },
+    {
+      "epoch": 0.023572601680997004,
+      "grad_norm": 0.7811074256896973,
+      "learning_rate": 2.335907335907336e-06,
+      "loss": 0.6164,
+      "step": 122
+    },
+    {
+      "epoch": 0.023765819727562556,
+      "grad_norm": 0.794479489326477,
+      "learning_rate": 2.3552123552123553e-06,
+      "loss": 0.6519,
+      "step": 123
+    },
+    {
+      "epoch": 0.023959037774128104,
+      "grad_norm": 0.7203382253646851,
+      "learning_rate": 2.3745173745173745e-06,
+      "loss": 0.6149,
+      "step": 124
+    },
+    {
+      "epoch": 0.024152255820693653,
+      "grad_norm": 0.8292742967605591,
+      "learning_rate": 2.393822393822394e-06,
+      "loss": 0.6846,
+      "step": 125
+    },
+    {
+      "epoch": 0.0243454738672592,
+      "grad_norm": 0.7453843355178833,
+      "learning_rate": 2.4131274131274133e-06,
+      "loss": 0.6574,
+      "step": 126
+    },
+    {
+      "epoch": 0.02453869191382475,
+      "grad_norm": 0.8354947566986084,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.7754,
+      "step": 127
+    },
+    {
+      "epoch": 0.024731909960390302,
+      "grad_norm": 0.8018094301223755,
+      "learning_rate": 2.451737451737452e-06,
+      "loss": 0.6658,
+      "step": 128
+    },
+    {
+      "epoch": 0.02492512800695585,
+      "grad_norm": 0.8007246851921082,
+      "learning_rate": 2.4710424710424712e-06,
+      "loss": 0.6393,
+      "step": 129
+    },
+    {
+      "epoch": 0.0251183460535214,
+      "grad_norm": 0.8323417901992798,
+      "learning_rate": 2.4903474903474904e-06,
+      "loss": 0.694,
+      "step": 130
+    },
+    {
+      "epoch": 0.025311564100086947,
+      "grad_norm": 0.743604302406311,
+      "learning_rate": 2.50965250965251e-06,
+      "loss": 0.6417,
+      "step": 131
+    },
+    {
+      "epoch": 0.025504782146652496,
+      "grad_norm": 0.7219024300575256,
+      "learning_rate": 2.528957528957529e-06,
+      "loss": 0.5877,
+      "step": 132
+    },
+    {
+      "epoch": 0.025698000193218048,
+      "grad_norm": 0.7326737642288208,
+      "learning_rate": 2.5482625482625484e-06,
+      "loss": 0.6197,
+      "step": 133
+    },
+    {
+      "epoch": 0.025891218239783596,
+      "grad_norm": 0.781621515750885,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 0.6547,
+      "step": 134
+    },
+    {
+      "epoch": 0.026084436286349145,
+      "grad_norm": 0.7975575923919678,
+      "learning_rate": 2.5868725868725867e-06,
+      "loss": 0.5903,
+      "step": 135
+    },
+    {
+      "epoch": 0.026277654332914693,
+      "grad_norm": 0.744665265083313,
+      "learning_rate": 2.6061776061776063e-06,
+      "loss": 0.6857,
+      "step": 136
+    },
+    {
+      "epoch": 0.026470872379480242,
+      "grad_norm": 0.7605339288711548,
+      "learning_rate": 2.6254826254826255e-06,
+      "loss": 0.7075,
+      "step": 137
+    },
+    {
+      "epoch": 0.026664090426045794,
+      "grad_norm": 0.7432572245597839,
+      "learning_rate": 2.644787644787645e-06,
+      "loss": 0.6389,
+      "step": 138
+    },
+    {
+      "epoch": 0.026857308472611342,
+      "grad_norm": 0.733087956905365,
+      "learning_rate": 2.6640926640926647e-06,
+      "loss": 0.5688,
+      "step": 139
+    },
+    {
+      "epoch": 0.02705052651917689,
+      "grad_norm": 0.8768149614334106,
+      "learning_rate": 2.683397683397684e-06,
+      "loss": 0.7636,
+      "step": 140
+    },
+    {
+      "epoch": 0.02724374456574244,
+      "grad_norm": 0.6562630534172058,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.609,
+      "step": 141
+    },
+    {
+      "epoch": 0.02743696261230799,
+      "grad_norm": 0.8046941757202148,
+      "learning_rate": 2.722007722007722e-06,
+      "loss": 0.6839,
+      "step": 142
+    },
+    {
+      "epoch": 0.02763018065887354,
+      "grad_norm": 0.7078561782836914,
+      "learning_rate": 2.7413127413127418e-06,
+      "loss": 0.6011,
+      "step": 143
+    },
+    {
+      "epoch": 0.02782339870543909,
+      "grad_norm": 0.736460268497467,
+      "learning_rate": 2.760617760617761e-06,
+      "loss": 0.5955,
+      "step": 144
+    },
+    {
+      "epoch": 0.028016616752004637,
+      "grad_norm": 0.8229767084121704,
+      "learning_rate": 2.77992277992278e-06,
+      "loss": 0.7007,
+      "step": 145
+    },
+    {
+      "epoch": 0.028209834798570185,
+      "grad_norm": 0.7205531597137451,
+      "learning_rate": 2.7992277992277993e-06,
+      "loss": 0.6266,
+      "step": 146
+    },
+    {
+      "epoch": 0.028403052845135737,
+      "grad_norm": 0.7434218525886536,
+      "learning_rate": 2.8185328185328185e-06,
+      "loss": 0.654,
+      "step": 147
+    },
+    {
+      "epoch": 0.028596270891701286,
+      "grad_norm": 0.7654426693916321,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 0.6991,
+      "step": 148
+    },
+    {
+      "epoch": 0.028789488938266834,
+      "grad_norm": 0.764680027961731,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.7306,
+      "step": 149
+    },
+    {
+      "epoch": 0.028982706984832383,
+      "grad_norm": 0.8201958537101746,
+      "learning_rate": 2.8764478764478764e-06,
+      "loss": 0.6982,
+      "step": 150
+    },
+    {
+      "epoch": 0.02917592503139793,
+      "grad_norm": 0.759795069694519,
+      "learning_rate": 2.8957528957528956e-06,
+      "loss": 0.6749,
+      "step": 151
+    },
+    {
+      "epoch": 0.029369143077963483,
+      "grad_norm": 0.724219560623169,
+      "learning_rate": 2.915057915057915e-06,
+      "loss": 0.6494,
+      "step": 152
+    },
+    {
+      "epoch": 0.02956236112452903,
+      "grad_norm": 0.6902620792388916,
+      "learning_rate": 2.934362934362935e-06,
+      "loss": 0.5698,
+      "step": 153
+    },
+    {
+      "epoch": 0.02975557917109458,
+      "grad_norm": 0.8445480465888977,
+      "learning_rate": 2.953667953667954e-06,
+      "loss": 0.6649,
+      "step": 154
+    },
+    {
+      "epoch": 0.02994879721766013,
+      "grad_norm": 0.7947871088981628,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.6308,
+      "step": 155
+    },
+    {
+      "epoch": 0.030142015264225677,
+      "grad_norm": 0.7539578676223755,
+      "learning_rate": 2.9922779922779927e-06,
+      "loss": 0.6729,
+      "step": 156
+    },
+    {
+      "epoch": 0.03033523331079123,
+      "grad_norm": 0.796784520149231,
+      "learning_rate": 3.011583011583012e-06,
+      "loss": 0.672,
+      "step": 157
+    },
+    {
+      "epoch": 0.030528451357356778,
+      "grad_norm": 0.81133633852005,
+      "learning_rate": 3.030888030888031e-06,
+      "loss": 0.6719,
+      "step": 158
+    },
+    {
+      "epoch": 0.030721669403922326,
+      "grad_norm": 0.8171602487564087,
+      "learning_rate": 3.0501930501930503e-06,
+      "loss": 0.7088,
+      "step": 159
+    },
+    {
+      "epoch": 0.030914887450487875,
+      "grad_norm": 0.8611062169075012,
+      "learning_rate": 3.06949806949807e-06,
+      "loss": 0.7287,
+      "step": 160
+    },
+    {
+      "epoch": 0.031108105497053423,
+      "grad_norm": 0.7351294159889221,
+      "learning_rate": 3.088803088803089e-06,
+      "loss": 0.6615,
+      "step": 161
+    },
+    {
+      "epoch": 0.031301323543618975,
+      "grad_norm": 0.7424891591072083,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 0.5974,
+      "step": 162
+    },
+    {
+      "epoch": 0.03149454159018452,
+      "grad_norm": 0.7051703929901123,
+      "learning_rate": 3.1274131274131274e-06,
+      "loss": 0.6111,
+      "step": 163
+    },
+    {
+      "epoch": 0.03168775963675007,
+      "grad_norm": 0.7657168507575989,
+      "learning_rate": 3.146718146718147e-06,
+      "loss": 0.6466,
+      "step": 164
+    },
+    {
+      "epoch": 0.031880977683315624,
+      "grad_norm": 0.7239560484886169,
+      "learning_rate": 3.166023166023166e-06,
+      "loss": 0.6313,
+      "step": 165
+    },
+    {
+      "epoch": 0.03207419572988117,
+      "grad_norm": 0.7457195520401001,
+      "learning_rate": 3.1853281853281853e-06,
+      "loss": 0.6253,
+      "step": 166
+    },
+    {
+      "epoch": 0.03226741377644672,
+      "grad_norm": 0.7441765666007996,
+      "learning_rate": 3.2046332046332054e-06,
+      "loss": 0.6189,
+      "step": 167
+    },
+    {
+      "epoch": 0.032460631823012266,
+      "grad_norm": 0.8655751943588257,
+      "learning_rate": 3.2239382239382245e-06,
+      "loss": 0.5932,
+      "step": 168
+    },
+    {
+      "epoch": 0.03265384986957782,
+      "grad_norm": 0.6850665807723999,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.6107,
+      "step": 169
+    },
+    {
+      "epoch": 0.03284706791614337,
+      "grad_norm": 0.7339829802513123,
+      "learning_rate": 3.262548262548263e-06,
+      "loss": 0.7331,
+      "step": 170
+    },
+    {
+      "epoch": 0.033040285962708915,
+      "grad_norm": 0.8805243968963623,
+      "learning_rate": 3.281853281853282e-06,
+      "loss": 0.6656,
+      "step": 171
+    },
+    {
+      "epoch": 0.03323350400927447,
+      "grad_norm": 0.8892763257026672,
+      "learning_rate": 3.3011583011583016e-06,
+      "loss": 0.7129,
+      "step": 172
+    },
+    {
+      "epoch": 0.03342672205584001,
+      "grad_norm": 0.7906018495559692,
+      "learning_rate": 3.320463320463321e-06,
+      "loss": 0.6572,
+      "step": 173
+    },
+    {
+      "epoch": 0.033619940102405564,
+      "grad_norm": 0.7841282486915588,
+      "learning_rate": 3.33976833976834e-06,
+      "loss": 0.6759,
+      "step": 174
+    },
+    {
+      "epoch": 0.033813158148971116,
+      "grad_norm": 0.821782112121582,
+      "learning_rate": 3.359073359073359e-06,
+      "loss": 0.6879,
+      "step": 175
+    },
+    {
+      "epoch": 0.03400637619553666,
+      "grad_norm": 0.7901400923728943,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 0.6065,
+      "step": 176
+    },
+    {
+      "epoch": 0.03419959424210221,
+      "grad_norm": 0.8000864386558533,
+      "learning_rate": 3.397683397683398e-06,
+      "loss": 0.6467,
+      "step": 177
+    },
+    {
+      "epoch": 0.03439281228866776,
+      "grad_norm": 0.7840276956558228,
+      "learning_rate": 3.416988416988417e-06,
+      "loss": 0.6793,
+      "step": 178
+    },
+    {
+      "epoch": 0.03458603033523331,
+      "grad_norm": 0.7378012537956238,
+      "learning_rate": 3.4362934362934363e-06,
+      "loss": 0.6619,
+      "step": 179
+    },
+    {
+      "epoch": 0.03477924838179886,
+      "grad_norm": 0.7361359596252441,
+      "learning_rate": 3.4555984555984555e-06,
+      "loss": 0.623,
+      "step": 180
+    },
+    {
+      "epoch": 0.03497246642836441,
+      "grad_norm": 0.7999025583267212,
+      "learning_rate": 3.4749034749034755e-06,
+      "loss": 0.6035,
+      "step": 181
+    },
+    {
+      "epoch": 0.03516568447492996,
+      "grad_norm": 0.744756817817688,
+      "learning_rate": 3.4942084942084947e-06,
+      "loss": 0.7065,
+      "step": 182
+    },
+    {
+      "epoch": 0.03535890252149551,
+      "grad_norm": 0.8944364786148071,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.7177,
+      "step": 183
+    },
+    {
+      "epoch": 0.035552120568061056,
+      "grad_norm": 0.7549863457679749,
+      "learning_rate": 3.5328185328185334e-06,
+      "loss": 0.6599,
+      "step": 184
+    },
+    {
+      "epoch": 0.03574533861462661,
+      "grad_norm": 0.7908114790916443,
+      "learning_rate": 3.5521235521235526e-06,
+      "loss": 0.5985,
+      "step": 185
+    },
+    {
+      "epoch": 0.03593855666119215,
+      "grad_norm": 0.7791903018951416,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 0.654,
+      "step": 186
+    },
+    {
+      "epoch": 0.036131774707757705,
+      "grad_norm": 0.7113376259803772,
+      "learning_rate": 3.590733590733591e-06,
+      "loss": 0.6191,
+      "step": 187
+    },
+    {
+      "epoch": 0.03632499275432326,
+      "grad_norm": 0.7906246185302734,
+      "learning_rate": 3.61003861003861e-06,
+      "loss": 0.6354,
+      "step": 188
+    },
+    {
+      "epoch": 0.0365182108008888,
+      "grad_norm": 0.7284888625144958,
+      "learning_rate": 3.6293436293436297e-06,
+      "loss": 0.6882,
+      "step": 189
+    },
+    {
+      "epoch": 0.036711428847454354,
+      "grad_norm": 0.7666733860969543,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 0.6582,
+      "step": 190
+    },
+    {
+      "epoch": 0.0369046468940199,
+      "grad_norm": 0.850034773349762,
+      "learning_rate": 3.667953667953668e-06,
+      "loss": 0.7483,
+      "step": 191
+    },
+    {
+      "epoch": 0.03709786494058545,
+      "grad_norm": 0.8628738522529602,
+      "learning_rate": 3.6872586872586872e-06,
+      "loss": 0.634,
+      "step": 192
+    },
+    {
+      "epoch": 0.037291082987151,
+      "grad_norm": 0.7830010652542114,
+      "learning_rate": 3.706563706563707e-06,
+      "loss": 0.6703,
+      "step": 193
+    },
+    {
+      "epoch": 0.03748430103371655,
+      "grad_norm": 0.7513682842254639,
+      "learning_rate": 3.725868725868726e-06,
+      "loss": 0.6949,
+      "step": 194
+    },
+    {
+      "epoch": 0.0376775190802821,
+      "grad_norm": 0.8186044096946716,
+      "learning_rate": 3.745173745173745e-06,
+      "loss": 0.6467,
+      "step": 195
+    },
+    {
+      "epoch": 0.037870737126847645,
+      "grad_norm": 0.7451923489570618,
+      "learning_rate": 3.7644787644787652e-06,
+      "loss": 0.699,
+      "step": 196
+    },
+    {
+      "epoch": 0.0380639551734132,
+      "grad_norm": 0.7992118000984192,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.6623,
+      "step": 197
+    },
+    {
+      "epoch": 0.03825717321997875,
+      "grad_norm": 0.7421767115592957,
+      "learning_rate": 3.8030888030888036e-06,
+      "loss": 0.596,
+      "step": 198
+    },
+    {
+      "epoch": 0.038450391266544294,
+      "grad_norm": 0.7688970565795898,
+      "learning_rate": 3.822393822393823e-06,
+      "loss": 0.6625,
+      "step": 199
+    },
+    {
+      "epoch": 0.038643609313109846,
+      "grad_norm": 0.7276183366775513,
+      "learning_rate": 3.841698841698842e-06,
+      "loss": 0.6789,
+      "step": 200
+    },
+    {
+      "epoch": 0.03883682735967539,
+      "grad_norm": 0.8299877643585205,
+      "learning_rate": 3.861003861003861e-06,
+      "loss": 0.708,
+      "step": 201
+    },
+    {
+      "epoch": 0.03903004540624094,
+      "grad_norm": 0.8517568111419678,
+      "learning_rate": 3.880308880308881e-06,
+      "loss": 0.7024,
+      "step": 202
+    },
+    {
+      "epoch": 0.039223263452806495,
+      "grad_norm": 0.7293545007705688,
+      "learning_rate": 3.8996138996139e-06,
+      "loss": 0.6743,
+      "step": 203
+    },
+    {
+      "epoch": 0.03941648149937204,
+      "grad_norm": 0.7845068573951721,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 0.6421,
+      "step": 204
+    },
+    {
+      "epoch": 0.03960969954593759,
+      "grad_norm": 0.9616631865501404,
+      "learning_rate": 3.938223938223939e-06,
+      "loss": 0.6642,
+      "step": 205
+    },
+    {
+      "epoch": 0.03980291759250314,
+      "grad_norm": 0.7878134846687317,
+      "learning_rate": 3.957528957528957e-06,
+      "loss": 0.6414,
+      "step": 206
+    },
+    {
+      "epoch": 0.03999613563906869,
+      "grad_norm": 0.8062455654144287,
+      "learning_rate": 3.976833976833977e-06,
+      "loss": 0.6636,
+      "step": 207
+    },
+    {
+      "epoch": 0.04018935368563424,
+      "grad_norm": 0.8306580185890198,
+      "learning_rate": 3.996138996138997e-06,
+      "loss": 0.7168,
+      "step": 208
+    },
+    {
+      "epoch": 0.040382571732199786,
+      "grad_norm": 0.8037571907043457,
+      "learning_rate": 4.015444015444015e-06,
+      "loss": 0.679,
+      "step": 209
+    },
+    {
+      "epoch": 0.04057578977876534,
+      "grad_norm": 0.8206264972686768,
+      "learning_rate": 4.034749034749035e-06,
+      "loss": 0.6976,
+      "step": 210
+    },
+    {
+      "epoch": 0.04076900782533088,
+      "grad_norm": 0.7435989379882812,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.6867,
+      "step": 211
+    },
+    {
+      "epoch": 0.040962225871896435,
+      "grad_norm": 0.7743417620658875,
+      "learning_rate": 4.073359073359074e-06,
+      "loss": 0.6215,
+      "step": 212
+    },
+    {
+      "epoch": 0.04115544391846199,
+      "grad_norm": 0.822509765625,
+      "learning_rate": 4.092664092664093e-06,
+      "loss": 0.7236,
+      "step": 213
+    },
+    {
+      "epoch": 0.04134866196502753,
+      "grad_norm": 0.8466773629188538,
+      "learning_rate": 4.1119691119691125e-06,
+      "loss": 0.68,
+      "step": 214
+    },
+    {
+      "epoch": 0.041541880011593084,
+      "grad_norm": 0.760277271270752,
+      "learning_rate": 4.131274131274132e-06,
+      "loss": 0.6758,
+      "step": 215
+    },
+    {
+      "epoch": 0.04173509805815863,
+      "grad_norm": 0.7525784969329834,
+      "learning_rate": 4.150579150579151e-06,
+      "loss": 0.6587,
+      "step": 216
+    },
+    {
+      "epoch": 0.04192831610472418,
+      "grad_norm": 0.7359393239021301,
+      "learning_rate": 4.16988416988417e-06,
+      "loss": 0.6621,
+      "step": 217
+    },
+    {
+      "epoch": 0.04212153415128973,
+      "grad_norm": 0.7852592468261719,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 0.6753,
+      "step": 218
+    },
+    {
+      "epoch": 0.04231475219785528,
+      "grad_norm": 0.7648798227310181,
+      "learning_rate": 4.208494208494209e-06,
+      "loss": 0.6736,
+      "step": 219
+    },
+    {
+      "epoch": 0.04250797024442083,
+      "grad_norm": 0.783355712890625,
+      "learning_rate": 4.227799227799228e-06,
+      "loss": 0.6047,
+      "step": 220
+    },
+    {
+      "epoch": 0.042701188290986375,
+      "grad_norm": 0.7008970379829407,
+      "learning_rate": 4.247104247104247e-06,
+      "loss": 0.6644,
+      "step": 221
+    },
+    {
+      "epoch": 0.04289440633755193,
+      "grad_norm": 0.7515130639076233,
+      "learning_rate": 4.266409266409267e-06,
+      "loss": 0.7163,
+      "step": 222
+    },
+    {
+      "epoch": 0.04308762438411748,
+      "grad_norm": 0.8187398910522461,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.6315,
+      "step": 223
+    },
+    {
+      "epoch": 0.043280842430683024,
+      "grad_norm": 0.8671156167984009,
+      "learning_rate": 4.305019305019305e-06,
+      "loss": 0.7136,
+      "step": 224
+    },
+    {
+      "epoch": 0.043474060477248576,
+      "grad_norm": 0.8156034350395203,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.6163,
+      "step": 225
+    },
+    {
+      "epoch": 0.04366727852381412,
+      "grad_norm": 0.8157216310501099,
+      "learning_rate": 4.343629343629344e-06,
+      "loss": 0.6575,
+      "step": 226
+    },
+    {
+      "epoch": 0.04386049657037967,
+      "grad_norm": 0.8146572709083557,
+      "learning_rate": 4.362934362934364e-06,
+      "loss": 0.6245,
+      "step": 227
+    },
+    {
+      "epoch": 0.044053714616945225,
+      "grad_norm": 0.7275128364562988,
+      "learning_rate": 4.382239382239383e-06,
+      "loss": 0.6258,
+      "step": 228
+    },
+    {
+      "epoch": 0.04424693266351077,
+      "grad_norm": 0.7390708327293396,
+      "learning_rate": 4.401544401544402e-06,
+      "loss": 0.6133,
+      "step": 229
+    },
+    {
+      "epoch": 0.04444015071007632,
+      "grad_norm": 0.7703117728233337,
+      "learning_rate": 4.420849420849421e-06,
+      "loss": 0.7237,
+      "step": 230
+    },
+    {
+      "epoch": 0.044633368756641874,
+      "grad_norm": 0.7367171049118042,
+      "learning_rate": 4.4401544401544405e-06,
+      "loss": 0.5658,
+      "step": 231
+    },
+    {
+      "epoch": 0.04482658680320742,
+      "grad_norm": 0.8489328026771545,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.7311,
+      "step": 232
+    },
+    {
+      "epoch": 0.04501980484977297,
+      "grad_norm": 0.7662409543991089,
+      "learning_rate": 4.478764478764479e-06,
+      "loss": 0.5798,
+      "step": 233
+    },
+    {
+      "epoch": 0.045213022896338516,
+      "grad_norm": 0.7435179352760315,
+      "learning_rate": 4.4980694980694985e-06,
+      "loss": 0.6543,
+      "step": 234
+    },
+    {
+      "epoch": 0.04540624094290407,
+      "grad_norm": 0.8153126239776611,
+      "learning_rate": 4.517374517374517e-06,
+      "loss": 0.6979,
+      "step": 235
+    },
+    {
+      "epoch": 0.04559945898946962,
+      "grad_norm": 0.7994595170021057,
+      "learning_rate": 4.536679536679537e-06,
+      "loss": 0.6591,
+      "step": 236
+    },
+    {
+      "epoch": 0.045792677036035165,
+      "grad_norm": 0.8926814794540405,
+      "learning_rate": 4.5559845559845564e-06,
+      "loss": 0.678,
+      "step": 237
+    },
+    {
+      "epoch": 0.04598589508260072,
+      "grad_norm": 0.7888785600662231,
+      "learning_rate": 4.575289575289575e-06,
+      "loss": 0.63,
+      "step": 238
+    },
+    {
+      "epoch": 0.04617911312916626,
+      "grad_norm": 0.6954684257507324,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.6309,
+      "step": 239
+    },
+    {
+      "epoch": 0.046372331175731814,
+      "grad_norm": 0.8547117114067078,
+      "learning_rate": 4.613899613899614e-06,
+      "loss": 0.6971,
+      "step": 240
+    },
+    {
+      "epoch": 0.046565549222297366,
+      "grad_norm": 0.7050371766090393,
+      "learning_rate": 4.633204633204634e-06,
+      "loss": 0.6043,
+      "step": 241
+    },
+    {
+      "epoch": 0.04675876726886291,
+      "grad_norm": 0.8056876063346863,
+      "learning_rate": 4.652509652509653e-06,
+      "loss": 0.6682,
+      "step": 242
+    },
+    {
+      "epoch": 0.04695198531542846,
+      "grad_norm": 0.8728021383285522,
+      "learning_rate": 4.671814671814672e-06,
+      "loss": 0.7004,
+      "step": 243
+    },
+    {
+      "epoch": 0.04714520336199401,
+      "grad_norm": 0.8158904314041138,
+      "learning_rate": 4.691119691119692e-06,
+      "loss": 0.597,
+      "step": 244
+    },
+    {
+      "epoch": 0.04733842140855956,
+      "grad_norm": 0.7607523202896118,
+      "learning_rate": 4.710424710424711e-06,
+      "loss": 0.6659,
+      "step": 245
+    },
+    {
+      "epoch": 0.04753163945512511,
+      "grad_norm": 0.7238503694534302,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 0.6017,
+      "step": 246
+    },
+    {
+      "epoch": 0.04772485750169066,
+      "grad_norm": 0.8198790550231934,
+      "learning_rate": 4.749034749034749e-06,
+      "loss": 0.5974,
+      "step": 247
+    },
+    {
+      "epoch": 0.04791807554825621,
+      "grad_norm": 0.7328284978866577,
+      "learning_rate": 4.768339768339769e-06,
+      "loss": 0.6649,
+      "step": 248
+    },
+    {
+      "epoch": 0.048111293594821754,
+      "grad_norm": 0.7986383438110352,
+      "learning_rate": 4.787644787644788e-06,
+      "loss": 0.6628,
+      "step": 249
+    },
+    {
+      "epoch": 0.048304511641387306,
+      "grad_norm": 0.8126029372215271,
+      "learning_rate": 4.806949806949807e-06,
+      "loss": 0.6284,
+      "step": 250
+    },
+    {
+      "epoch": 0.04849772968795286,
+      "grad_norm": 0.7029486894607544,
+      "learning_rate": 4.8262548262548266e-06,
+      "loss": 0.6472,
+      "step": 251
+    },
+    {
+      "epoch": 0.0486909477345184,
+      "grad_norm": 0.7995674014091492,
+      "learning_rate": 4.845559845559845e-06,
+      "loss": 0.6602,
+      "step": 252
+    },
+    {
+      "epoch": 0.048884165781083955,
+      "grad_norm": 0.79188072681427,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.6118,
+      "step": 253
+    },
+    {
+      "epoch": 0.0490773838276495,
+      "grad_norm": 0.7684262990951538,
+      "learning_rate": 4.8841698841698845e-06,
+      "loss": 0.6757,
+      "step": 254
+    },
+    {
+      "epoch": 0.04927060187421505,
+      "grad_norm": 0.7866804003715515,
+      "learning_rate": 4.903474903474904e-06,
+      "loss": 0.6552,
+      "step": 255
+    },
+    {
+      "epoch": 0.049463819920780604,
+      "grad_norm": 0.7230879068374634,
+      "learning_rate": 4.922779922779924e-06,
+      "loss": 0.62,
+      "step": 256
+    },
+    {
+      "epoch": 0.04965703796734615,
+      "grad_norm": 0.7624667286872864,
+      "learning_rate": 4.9420849420849425e-06,
+      "loss": 0.5521,
+      "step": 257
+    },
+    {
+      "epoch": 0.0498502560139117,
+      "grad_norm": 0.7553189992904663,
+      "learning_rate": 4.961389961389962e-06,
+      "loss": 0.6282,
+      "step": 258
+    },
+    {
+      "epoch": 0.050043474060477246,
+      "grad_norm": 0.7877834439277649,
+      "learning_rate": 4.980694980694981e-06,
+      "loss": 0.6442,
+      "step": 259
+    },
+    {
+      "epoch": 0.0502366921070428,
+      "grad_norm": 0.83204585313797,
+      "learning_rate": 5e-06,
+      "loss": 0.6616,
+      "step": 260
+    },
+    {
+      "epoch": 0.05042991015360835,
+      "grad_norm": 0.7907417416572571,
+      "learning_rate": 5.01930501930502e-06,
+      "loss": 0.63,
+      "step": 261
+    },
+    {
+      "epoch": 0.050623128200173895,
+      "grad_norm": 0.7460136413574219,
+      "learning_rate": 5.038610038610039e-06,
+      "loss": 0.6362,
+      "step": 262
+    },
+    {
+      "epoch": 0.05081634624673945,
+      "grad_norm": 0.7727457284927368,
+      "learning_rate": 5.057915057915058e-06,
+      "loss": 0.6068,
+      "step": 263
+    },
+    {
+      "epoch": 0.05100956429330499,
+      "grad_norm": 0.8016122579574585,
+      "learning_rate": 5.077220077220077e-06,
+      "loss": 0.6769,
+      "step": 264
+    },
+    {
+      "epoch": 0.051202782339870544,
+      "grad_norm": 0.7778483629226685,
+      "learning_rate": 5.096525096525097e-06,
+      "loss": 0.6612,
+      "step": 265
+    },
+    {
+      "epoch": 0.051396000386436096,
+      "grad_norm": 0.7092305421829224,
+      "learning_rate": 5.115830115830116e-06,
+      "loss": 0.624,
+      "step": 266
+    },
+    {
+      "epoch": 0.05158921843300164,
+      "grad_norm": 0.7606619000434875,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.6015,
+      "step": 267
+    },
+    {
+      "epoch": 0.05178243647956719,
+      "grad_norm": 0.7978340983390808,
+      "learning_rate": 5.154440154440155e-06,
+      "loss": 0.6891,
+      "step": 268
+    },
+    {
+      "epoch": 0.05197565452613274,
+      "grad_norm": 0.8385316133499146,
+      "learning_rate": 5.173745173745173e-06,
+      "loss": 0.7065,
+      "step": 269
+    },
+    {
+      "epoch": 0.05216887257269829,
+      "grad_norm": 0.7413119077682495,
+      "learning_rate": 5.193050193050193e-06,
+      "loss": 0.6456,
+      "step": 270
+    },
+    {
+      "epoch": 0.05236209061926384,
+      "grad_norm": 0.7719514966011047,
+      "learning_rate": 5.212355212355213e-06,
+      "loss": 0.6495,
+      "step": 271
+    },
+    {
+      "epoch": 0.05255530866582939,
+      "grad_norm": 0.8367528319358826,
+      "learning_rate": 5.231660231660231e-06,
+      "loss": 0.6957,
+      "step": 272
+    },
+    {
+      "epoch": 0.05274852671239494,
+      "grad_norm": 0.6857628226280212,
+      "learning_rate": 5.250965250965251e-06,
+      "loss": 0.5729,
+      "step": 273
+    },
+    {
+      "epoch": 0.052941744758960484,
+      "grad_norm": 0.7913002371788025,
+      "learning_rate": 5.2702702702702705e-06,
+      "loss": 0.6576,
+      "step": 274
+    },
+    {
+      "epoch": 0.053134962805526036,
+      "grad_norm": 0.7501757144927979,
+      "learning_rate": 5.28957528957529e-06,
+      "loss": 0.6287,
+      "step": 275
+    },
+    {
+      "epoch": 0.05332818085209159,
+      "grad_norm": 0.7898593544960022,
+      "learning_rate": 5.30888030888031e-06,
+      "loss": 0.6006,
+      "step": 276
+    },
+    {
+      "epoch": 0.05352139889865713,
+      "grad_norm": 0.7110061049461365,
+      "learning_rate": 5.328185328185329e-06,
+      "loss": 0.6419,
+      "step": 277
+    },
+    {
+      "epoch": 0.053714616945222685,
+      "grad_norm": 0.8224559426307678,
+      "learning_rate": 5.347490347490348e-06,
+      "loss": 0.6452,
+      "step": 278
+    },
+    {
+      "epoch": 0.05390783499178823,
+      "grad_norm": 0.7718434929847717,
+      "learning_rate": 5.366795366795368e-06,
+      "loss": 0.6451,
+      "step": 279
+    },
+    {
+      "epoch": 0.05410105303835378,
+      "grad_norm": 0.7362533211708069,
+      "learning_rate": 5.386100386100387e-06,
+      "loss": 0.6104,
+      "step": 280
+    },
+    {
+      "epoch": 0.054294271084919334,
+      "grad_norm": 0.7580875158309937,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.6462,
+      "step": 281
+    },
+    {
+      "epoch": 0.05448748913148488,
+      "grad_norm": 0.74604332447052,
+      "learning_rate": 5.424710424710426e-06,
+      "loss": 0.6742,
+      "step": 282
+    },
+    {
+      "epoch": 0.05468070717805043,
+      "grad_norm": 0.7118567228317261,
+      "learning_rate": 5.444015444015444e-06,
+      "loss": 0.6323,
+      "step": 283
+    },
+    {
+      "epoch": 0.05487392522461598,
+      "grad_norm": 0.7307426929473877,
+      "learning_rate": 5.463320463320464e-06,
+      "loss": 0.5919,
+      "step": 284
+    },
+    {
+      "epoch": 0.05506714327118153,
+      "grad_norm": 0.8564929962158203,
+      "learning_rate": 5.4826254826254836e-06,
+      "loss": 0.6635,
+      "step": 285
+    },
+    {
+      "epoch": 0.05526036131774708,
+      "grad_norm": 0.8071416616439819,
+      "learning_rate": 5.501930501930502e-06,
+      "loss": 0.668,
+      "step": 286
+    },
+    {
+      "epoch": 0.055453579364312625,
+      "grad_norm": 0.7841041088104248,
+      "learning_rate": 5.521235521235522e-06,
+      "loss": 0.5946,
+      "step": 287
+    },
+    {
+      "epoch": 0.05564679741087818,
+      "grad_norm": 0.7199358344078064,
+      "learning_rate": 5.540540540540541e-06,
+      "loss": 0.6169,
+      "step": 288
+    },
+    {
+      "epoch": 0.05584001545744373,
+      "grad_norm": 0.7220997214317322,
+      "learning_rate": 5.55984555984556e-06,
+      "loss": 0.6749,
+      "step": 289
+    },
+    {
+      "epoch": 0.056033233504009274,
+      "grad_norm": 0.807196319103241,
+      "learning_rate": 5.57915057915058e-06,
+      "loss": 0.6607,
+      "step": 290
+    },
+    {
+      "epoch": 0.056226451550574826,
+      "grad_norm": 0.6639469861984253,
+      "learning_rate": 5.598455598455599e-06,
+      "loss": 0.6314,
+      "step": 291
+    },
+    {
+      "epoch": 0.05641966959714037,
+      "grad_norm": 0.734620988368988,
+      "learning_rate": 5.617760617760618e-06,
+      "loss": 0.6027,
+      "step": 292
+    },
+    {
+      "epoch": 0.05661288764370592,
+      "grad_norm": 0.7420213222503662,
+      "learning_rate": 5.637065637065637e-06,
+      "loss": 0.6507,
+      "step": 293
+    },
+    {
+      "epoch": 0.056806105690271474,
+      "grad_norm": 0.756880521774292,
+      "learning_rate": 5.6563706563706566e-06,
+      "loss": 0.7113,
+      "step": 294
+    },
+    {
+      "epoch": 0.05699932373683702,
+      "grad_norm": 0.7640995979309082,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.6717,
+      "step": 295
+    },
+    {
+      "epoch": 0.05719254178340257,
+      "grad_norm": 0.80096435546875,
+      "learning_rate": 5.694980694980695e-06,
+      "loss": 0.6555,
+      "step": 296
+    },
+    {
+      "epoch": 0.057385759829968117,
+      "grad_norm": 0.7423076629638672,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.6187,
+      "step": 297
+    },
+    {
+      "epoch": 0.05757897787653367,
+      "grad_norm": 0.8070868849754333,
+      "learning_rate": 5.733590733590734e-06,
+      "loss": 0.6366,
+      "step": 298
+    },
+    {
+      "epoch": 0.05777219592309922,
+      "grad_norm": 0.7190064191818237,
+      "learning_rate": 5.752895752895753e-06,
+      "loss": 0.5748,
+      "step": 299
+    },
+    {
+      "epoch": 0.057965413969664766,
+      "grad_norm": 0.7650955319404602,
+      "learning_rate": 5.7722007722007725e-06,
+      "loss": 0.6303,
+      "step": 300
+    },
+    {
+      "epoch": 0.05815863201623032,
+      "grad_norm": 0.7590053081512451,
+      "learning_rate": 5.791505791505791e-06,
+      "loss": 0.6232,
+      "step": 301
+    },
+    {
+      "epoch": 0.05835185006279586,
+      "grad_norm": 0.7856601476669312,
+      "learning_rate": 5.810810810810811e-06,
+      "loss": 0.7095,
+      "step": 302
+    },
+    {
+      "epoch": 0.058545068109361414,
+      "grad_norm": 0.7115820646286011,
+      "learning_rate": 5.83011583011583e-06,
+      "loss": 0.6785,
+      "step": 303
+    },
+    {
+      "epoch": 0.058738286155926966,
+      "grad_norm": 0.7574310302734375,
+      "learning_rate": 5.84942084942085e-06,
+      "loss": 0.5915,
+      "step": 304
+    },
+    {
+      "epoch": 0.05893150420249251,
+      "grad_norm": 0.812767744064331,
+      "learning_rate": 5.86872586872587e-06,
+      "loss": 0.6068,
+      "step": 305
+    },
+    {
+      "epoch": 0.05912472224905806,
+      "grad_norm": 0.8499731421470642,
+      "learning_rate": 5.888030888030889e-06,
+      "loss": 0.6221,
+      "step": 306
+    },
+    {
+      "epoch": 0.05931794029562361,
+      "grad_norm": 0.6972988843917847,
+      "learning_rate": 5.907335907335908e-06,
+      "loss": 0.5928,
+      "step": 307
+    },
+    {
+      "epoch": 0.05951115834218916,
+      "grad_norm": 0.795735239982605,
+      "learning_rate": 5.9266409266409275e-06,
+      "loss": 0.6333,
+      "step": 308
+    },
+    {
+      "epoch": 0.05970437638875471,
+      "grad_norm": 0.7444106936454773,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.6068,
+      "step": 309
+    },
+    {
+      "epoch": 0.05989759443532026,
+      "grad_norm": 0.7811155319213867,
+      "learning_rate": 5.965250965250966e-06,
+      "loss": 0.5966,
+      "step": 310
+    },
+    {
+      "epoch": 0.06009081248188581,
+      "grad_norm": 0.7589752674102783,
+      "learning_rate": 5.9845559845559855e-06,
+      "loss": 0.6222,
+      "step": 311
+    },
+    {
+      "epoch": 0.060284030528451354,
+      "grad_norm": 0.7095033526420593,
+      "learning_rate": 6.003861003861004e-06,
+      "loss": 0.6558,
+      "step": 312
+    },
+    {
+      "epoch": 0.060477248575016906,
+      "grad_norm": 0.8051417469978333,
+      "learning_rate": 6.023166023166024e-06,
+      "loss": 0.6228,
+      "step": 313
+    },
+    {
+      "epoch": 0.06067046662158246,
+      "grad_norm": 0.7451211214065552,
+      "learning_rate": 6.0424710424710434e-06,
+      "loss": 0.5816,
+      "step": 314
+    },
+    {
+      "epoch": 0.060863684668148,
+      "grad_norm": 0.7650222182273865,
+      "learning_rate": 6.061776061776062e-06,
+      "loss": 0.6791,
+      "step": 315
+    },
+    {
+      "epoch": 0.061056902714713555,
+      "grad_norm": 0.7695963978767395,
+      "learning_rate": 6.081081081081082e-06,
+      "loss": 0.6803,
+      "step": 316
+    },
+    {
+      "epoch": 0.0612501207612791,
+      "grad_norm": 0.7684622406959534,
+      "learning_rate": 6.1003861003861005e-06,
+      "loss": 0.6424,
+      "step": 317
+    },
+    {
+      "epoch": 0.06144333880784465,
+      "grad_norm": 0.8167831301689148,
+      "learning_rate": 6.11969111969112e-06,
+      "loss": 0.6833,
+      "step": 318
+    },
+    {
+      "epoch": 0.061636556854410204,
+      "grad_norm": 0.861356794834137,
+      "learning_rate": 6.13899613899614e-06,
+      "loss": 0.6211,
+      "step": 319
+    },
+    {
+      "epoch": 0.06182977490097575,
+      "grad_norm": 0.7292737364768982,
+      "learning_rate": 6.1583011583011585e-06,
+      "loss": 0.657,
+      "step": 320
+    },
+    {
+      "epoch": 0.0620229929475413,
+      "grad_norm": 0.8020837306976318,
+      "learning_rate": 6.177606177606178e-06,
+      "loss": 0.6148,
+      "step": 321
+    },
+    {
+      "epoch": 0.062216210994106846,
+      "grad_norm": 0.7826812863349915,
+      "learning_rate": 6.196911196911197e-06,
+      "loss": 0.6276,
+      "step": 322
+    },
+    {
+      "epoch": 0.0624094290406724,
+      "grad_norm": 0.7368956208229065,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.6028,
+      "step": 323
+    },
+    {
+      "epoch": 0.06260264708723795,
+      "grad_norm": 0.6791717410087585,
+      "learning_rate": 6.235521235521236e-06,
+      "loss": 0.582,
+      "step": 324
+    },
+    {
+      "epoch": 0.0627958651338035,
+      "grad_norm": 0.7549542784690857,
+      "learning_rate": 6.254826254826255e-06,
+      "loss": 0.6698,
+      "step": 325
+    },
+    {
+      "epoch": 0.06298908318036904,
+      "grad_norm": 0.743270754814148,
+      "learning_rate": 6.274131274131274e-06,
+      "loss": 0.6336,
+      "step": 326
+    },
+    {
+      "epoch": 0.06318230122693459,
+      "grad_norm": 0.7452760934829712,
+      "learning_rate": 6.293436293436294e-06,
+      "loss": 0.6298,
+      "step": 327
+    },
+    {
+      "epoch": 0.06337551927350014,
+      "grad_norm": 0.7866285443305969,
+      "learning_rate": 6.312741312741313e-06,
+      "loss": 0.6837,
+      "step": 328
+    },
+    {
+      "epoch": 0.0635687373200657,
+      "grad_norm": 0.792092502117157,
+      "learning_rate": 6.332046332046332e-06,
+      "loss": 0.6642,
+      "step": 329
+    },
+    {
+      "epoch": 0.06376195536663125,
+      "grad_norm": 0.8661297559738159,
+      "learning_rate": 6.351351351351351e-06,
+      "loss": 0.6318,
+      "step": 330
+    },
+    {
+      "epoch": 0.06395517341319679,
+      "grad_norm": 0.7904163599014282,
+      "learning_rate": 6.370656370656371e-06,
+      "loss": 0.724,
+      "step": 331
+    },
+    {
+      "epoch": 0.06414839145976234,
+      "grad_norm": 0.7684714794158936,
+      "learning_rate": 6.389961389961391e-06,
+      "loss": 0.6344,
+      "step": 332
+    },
+    {
+      "epoch": 0.06434160950632789,
+      "grad_norm": 0.7642500996589661,
+      "learning_rate": 6.409266409266411e-06,
+      "loss": 0.6868,
+      "step": 333
+    },
+    {
+      "epoch": 0.06453482755289344,
+      "grad_norm": 0.8003482818603516,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 0.6791,
+      "step": 334
+    },
+    {
+      "epoch": 0.064728045599459,
+      "grad_norm": 0.7069498300552368,
+      "learning_rate": 6.447876447876449e-06,
+      "loss": 0.6168,
+      "step": 335
+    },
+    {
+      "epoch": 0.06492126364602453,
+      "grad_norm": 0.7803301811218262,
+      "learning_rate": 6.467181467181468e-06,
+      "loss": 0.668,
+      "step": 336
+    },
+    {
+      "epoch": 0.06511448169259008,
+      "grad_norm": 0.7732304334640503,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.6453,
+      "step": 337
+    },
+    {
+      "epoch": 0.06530769973915564,
+      "grad_norm": 0.7273716330528259,
+      "learning_rate": 6.505791505791507e-06,
+      "loss": 0.5712,
+      "step": 338
+    },
+    {
+      "epoch": 0.06550091778572119,
+      "grad_norm": 0.753078043460846,
+      "learning_rate": 6.525096525096526e-06,
+      "loss": 0.6736,
+      "step": 339
+    },
+    {
+      "epoch": 0.06569413583228674,
+      "grad_norm": 0.7467560172080994,
+      "learning_rate": 6.544401544401545e-06,
+      "loss": 0.578,
+      "step": 340
+    },
+    {
+      "epoch": 0.06588735387885228,
+      "grad_norm": 0.8697112202644348,
+      "learning_rate": 6.563706563706564e-06,
+      "loss": 0.6598,
+      "step": 341
+    },
+    {
+      "epoch": 0.06608057192541783,
+      "grad_norm": 0.7350066304206848,
+      "learning_rate": 6.583011583011584e-06,
+      "loss": 0.5897,
+      "step": 342
+    },
+    {
+      "epoch": 0.06627378997198338,
+      "grad_norm": 0.7733203768730164,
+      "learning_rate": 6.602316602316603e-06,
+      "loss": 0.6749,
+      "step": 343
+    },
+    {
+      "epoch": 0.06646700801854893,
+      "grad_norm": 0.772265613079071,
+      "learning_rate": 6.621621621621622e-06,
+      "loss": 0.6599,
+      "step": 344
+    },
+    {
+      "epoch": 0.06666022606511449,
+      "grad_norm": 0.7699328064918518,
+      "learning_rate": 6.640926640926642e-06,
+      "loss": 0.7044,
+      "step": 345
+    },
+    {
+      "epoch": 0.06685344411168002,
+      "grad_norm": 0.7921711206436157,
+      "learning_rate": 6.66023166023166e-06,
+      "loss": 0.6507,
+      "step": 346
+    },
+    {
+      "epoch": 0.06704666215824558,
+      "grad_norm": 0.6677191853523254,
+      "learning_rate": 6.67953667953668e-06,
+      "loss": 0.6098,
+      "step": 347
+    },
+    {
+      "epoch": 0.06723988020481113,
+      "grad_norm": 0.7781476974487305,
+      "learning_rate": 6.6988416988417e-06,
+      "loss": 0.6455,
+      "step": 348
+    },
+    {
+      "epoch": 0.06743309825137668,
+      "grad_norm": 0.7759572267532349,
+      "learning_rate": 6.718146718146718e-06,
+      "loss": 0.6165,
+      "step": 349
+    },
+    {
+      "epoch": 0.06762631629794223,
+      "grad_norm": 0.6900157332420349,
+      "learning_rate": 6.737451737451738e-06,
+      "loss": 0.6427,
+      "step": 350
+    },
+    {
+      "epoch": 0.06781953434450777,
+      "grad_norm": 0.7721596956253052,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.621,
+      "step": 351
+    },
+    {
+      "epoch": 0.06801275239107332,
+      "grad_norm": 0.7535867094993591,
+      "learning_rate": 6.776061776061776e-06,
+      "loss": 0.5996,
+      "step": 352
+    },
+    {
+      "epoch": 0.06820597043763887,
+      "grad_norm": 0.741678774356842,
+      "learning_rate": 6.795366795366796e-06,
+      "loss": 0.5928,
+      "step": 353
+    },
+    {
+      "epoch": 0.06839918848420443,
+      "grad_norm": 0.8253540396690369,
+      "learning_rate": 6.814671814671815e-06,
+      "loss": 0.6637,
+      "step": 354
+    },
+    {
+      "epoch": 0.06859240653076998,
+      "grad_norm": 0.8151888251304626,
+      "learning_rate": 6.833976833976834e-06,
+      "loss": 0.7079,
+      "step": 355
+    },
+    {
+      "epoch": 0.06878562457733552,
+      "grad_norm": 0.7356176376342773,
+      "learning_rate": 6.853281853281854e-06,
+      "loss": 0.6298,
+      "step": 356
+    },
+    {
+      "epoch": 0.06897884262390107,
+      "grad_norm": 0.6837357878684998,
+      "learning_rate": 6.872586872586873e-06,
+      "loss": 0.6978,
+      "step": 357
+    },
+    {
+      "epoch": 0.06917206067046662,
+      "grad_norm": 0.7189440131187439,
+      "learning_rate": 6.891891891891892e-06,
+      "loss": 0.5754,
+      "step": 358
+    },
+    {
+      "epoch": 0.06936527871703217,
+      "grad_norm": 0.7992641925811768,
+      "learning_rate": 6.911196911196911e-06,
+      "loss": 0.6165,
+      "step": 359
+    },
+    {
+      "epoch": 0.06955849676359772,
+      "grad_norm": 0.712891697883606,
+      "learning_rate": 6.9305019305019305e-06,
+      "loss": 0.606,
+      "step": 360
+    },
+    {
+      "epoch": 0.06975171481016328,
+      "grad_norm": 0.7928743958473206,
+      "learning_rate": 6.949806949806951e-06,
+      "loss": 0.6726,
+      "step": 361
+    },
+    {
+      "epoch": 0.06994493285672881,
+      "grad_norm": 0.8366293907165527,
+      "learning_rate": 6.9691119691119706e-06,
+      "loss": 0.6237,
+      "step": 362
+    },
+    {
+      "epoch": 0.07013815090329437,
+      "grad_norm": 0.8304416537284851,
+      "learning_rate": 6.988416988416989e-06,
+      "loss": 0.6898,
+      "step": 363
+    },
+    {
+      "epoch": 0.07033136894985992,
+      "grad_norm": 0.738990068435669,
+      "learning_rate": 7.007722007722009e-06,
+      "loss": 0.5993,
+      "step": 364
+    },
+    {
+      "epoch": 0.07052458699642547,
+      "grad_norm": 0.71754390001297,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.6121,
+      "step": 365
+    },
+    {
+      "epoch": 0.07071780504299102,
+      "grad_norm": 0.781915009021759,
+      "learning_rate": 7.046332046332047e-06,
+      "loss": 0.6327,
+      "step": 366
+    },
+    {
+      "epoch": 0.07091102308955656,
+      "grad_norm": 0.7772074341773987,
+      "learning_rate": 7.065637065637067e-06,
+      "loss": 0.6438,
+      "step": 367
+    },
+    {
+      "epoch": 0.07110424113612211,
+      "grad_norm": 0.7790648937225342,
+      "learning_rate": 7.084942084942086e-06,
+      "loss": 0.614,
+      "step": 368
+    },
+    {
+      "epoch": 0.07129745918268766,
+      "grad_norm": 0.8127062320709229,
+      "learning_rate": 7.104247104247105e-06,
+      "loss": 0.6155,
+      "step": 369
+    },
+    {
+      "epoch": 0.07149067722925322,
+      "grad_norm": 0.7364320158958435,
+      "learning_rate": 7.123552123552124e-06,
+      "loss": 0.6324,
+      "step": 370
+    },
+    {
+      "epoch": 0.07168389527581877,
+      "grad_norm": 0.8030374646186829,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.6694,
+      "step": 371
+    },
+    {
+      "epoch": 0.0718771133223843,
+      "grad_norm": 0.8097108006477356,
+      "learning_rate": 7.162162162162163e-06,
+      "loss": 0.6522,
+      "step": 372
+    },
+    {
+      "epoch": 0.07207033136894986,
+      "grad_norm": 0.7512496113777161,
+      "learning_rate": 7.181467181467182e-06,
+      "loss": 0.6874,
+      "step": 373
+    },
+    {
+      "epoch": 0.07226354941551541,
+      "grad_norm": 0.7259913682937622,
+      "learning_rate": 7.2007722007722015e-06,
+      "loss": 0.6026,
+      "step": 374
+    },
+    {
+      "epoch": 0.07245676746208096,
+      "grad_norm": 0.7729734182357788,
+      "learning_rate": 7.22007722007722e-06,
+      "loss": 0.6789,
+      "step": 375
+    },
+    {
+      "epoch": 0.07264998550864651,
+      "grad_norm": 0.7457165122032166,
+      "learning_rate": 7.23938223938224e-06,
+      "loss": 0.6829,
+      "step": 376
+    },
+    {
+      "epoch": 0.07284320355521205,
+      "grad_norm": 0.7708082795143127,
+      "learning_rate": 7.2586872586872595e-06,
+      "loss": 0.6688,
+      "step": 377
+    },
+    {
+      "epoch": 0.0730364216017776,
+      "grad_norm": 0.6661557555198669,
+      "learning_rate": 7.277992277992278e-06,
+      "loss": 0.5558,
+      "step": 378
+    },
+    {
+      "epoch": 0.07322963964834316,
+      "grad_norm": 0.7317824959754944,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.6087,
+      "step": 379
+    },
+    {
+      "epoch": 0.07342285769490871,
+      "grad_norm": 0.7614196538925171,
+      "learning_rate": 7.316602316602317e-06,
+      "loss": 0.6381,
+      "step": 380
+    },
+    {
+      "epoch": 0.07361607574147426,
+      "grad_norm": 0.7392153143882751,
+      "learning_rate": 7.335907335907336e-06,
+      "loss": 0.6761,
+      "step": 381
+    },
+    {
+      "epoch": 0.0738092937880398,
+      "grad_norm": 0.7957664728164673,
+      "learning_rate": 7.355212355212356e-06,
+      "loss": 0.5894,
+      "step": 382
+    },
+    {
+      "epoch": 0.07400251183460535,
+      "grad_norm": 0.7977986931800842,
+      "learning_rate": 7.3745173745173745e-06,
+      "loss": 0.6787,
+      "step": 383
+    },
+    {
+      "epoch": 0.0741957298811709,
+      "grad_norm": 0.7388816475868225,
+      "learning_rate": 7.393822393822394e-06,
+      "loss": 0.7445,
+      "step": 384
+    },
+    {
+      "epoch": 0.07438894792773645,
+      "grad_norm": 0.8354382514953613,
+      "learning_rate": 7.413127413127414e-06,
+      "loss": 0.6931,
+      "step": 385
+    },
+    {
+      "epoch": 0.074582165974302,
+      "grad_norm": 0.7825782895088196,
+      "learning_rate": 7.4324324324324324e-06,
+      "loss": 0.6037,
+      "step": 386
+    },
+    {
+      "epoch": 0.07477538402086754,
+      "grad_norm": 0.7611075043678284,
+      "learning_rate": 7.451737451737452e-06,
+      "loss": 0.6738,
+      "step": 387
+    },
+    {
+      "epoch": 0.0749686020674331,
+      "grad_norm": 0.7871357798576355,
+      "learning_rate": 7.471042471042471e-06,
+      "loss": 0.719,
+      "step": 388
+    },
+    {
+      "epoch": 0.07516182011399865,
+      "grad_norm": 0.7931118011474609,
+      "learning_rate": 7.49034749034749e-06,
+      "loss": 0.694,
+      "step": 389
+    },
+    {
+      "epoch": 0.0753550381605642,
+      "grad_norm": 0.7573150992393494,
+      "learning_rate": 7.509652509652511e-06,
+      "loss": 0.6128,
+      "step": 390
+    },
+    {
+      "epoch": 0.07554825620712975,
+      "grad_norm": 0.7644912004470825,
+      "learning_rate": 7.5289575289575304e-06,
+      "loss": 0.686,
+      "step": 391
+    },
+    {
+      "epoch": 0.07574147425369529,
+      "grad_norm": 0.7740148305892944,
+      "learning_rate": 7.548262548262549e-06,
+      "loss": 0.7004,
+      "step": 392
+    },
+    {
+      "epoch": 0.07593469230026084,
+      "grad_norm": 0.8279855251312256,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.7209,
+      "step": 393
+    },
+    {
+      "epoch": 0.0761279103468264,
+      "grad_norm": 0.7397010326385498,
+      "learning_rate": 7.5868725868725875e-06,
+      "loss": 0.6002,
+      "step": 394
+    },
+    {
+      "epoch": 0.07632112839339195,
+      "grad_norm": 0.7014784216880798,
+      "learning_rate": 7.606177606177607e-06,
+      "loss": 0.6706,
+      "step": 395
+    },
+    {
+      "epoch": 0.0765143464399575,
+      "grad_norm": 0.7208130955696106,
+      "learning_rate": 7.625482625482627e-06,
+      "loss": 0.6089,
+      "step": 396
+    },
+    {
+      "epoch": 0.07670756448652304,
+      "grad_norm": 0.7918097972869873,
+      "learning_rate": 7.644787644787645e-06,
+      "loss": 0.6429,
+      "step": 397
+    },
+    {
+      "epoch": 0.07690078253308859,
+      "grad_norm": 0.8130318522453308,
+      "learning_rate": 7.664092664092665e-06,
+      "loss": 0.651,
+      "step": 398
+    },
+    {
+      "epoch": 0.07709400057965414,
+      "grad_norm": 0.7192891240119934,
+      "learning_rate": 7.683397683397685e-06,
+      "loss": 0.5935,
+      "step": 399
+    },
+    {
+      "epoch": 0.07728721862621969,
+      "grad_norm": 0.7489885091781616,
+      "learning_rate": 7.702702702702704e-06,
+      "loss": 0.6326,
+      "step": 400
+    },
+    {
+      "epoch": 0.07748043667278524,
+      "grad_norm": 0.7583082914352417,
+      "learning_rate": 7.722007722007722e-06,
+      "loss": 0.6552,
+      "step": 401
+    },
+    {
+      "epoch": 0.07767365471935078,
+      "grad_norm": 0.8101308345794678,
+      "learning_rate": 7.741312741312742e-06,
+      "loss": 0.6805,
+      "step": 402
+    },
+    {
+      "epoch": 0.07786687276591633,
+      "grad_norm": 0.7590944766998291,
+      "learning_rate": 7.760617760617761e-06,
+      "loss": 0.6089,
+      "step": 403
+    },
+    {
+      "epoch": 0.07806009081248189,
+      "grad_norm": 0.7292355298995972,
+      "learning_rate": 7.779922779922781e-06,
+      "loss": 0.6671,
+      "step": 404
+    },
+    {
+      "epoch": 0.07825330885904744,
+      "grad_norm": 0.7393438220024109,
+      "learning_rate": 7.7992277992278e-06,
+      "loss": 0.6313,
+      "step": 405
+    },
+    {
+      "epoch": 0.07844652690561299,
+      "grad_norm": 0.7392045855522156,
+      "learning_rate": 7.818532818532818e-06,
+      "loss": 0.6222,
+      "step": 406
+    },
+    {
+      "epoch": 0.07863974495217853,
+      "grad_norm": 0.8252953290939331,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 0.6333,
+      "step": 407
+    },
+    {
+      "epoch": 0.07883296299874408,
+      "grad_norm": 0.7763678431510925,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.6928,
+      "step": 408
+    },
+    {
+      "epoch": 0.07902618104530963,
+      "grad_norm": 0.8488125205039978,
+      "learning_rate": 7.876447876447877e-06,
+      "loss": 0.6459,
+      "step": 409
+    },
+    {
+      "epoch": 0.07921939909187518,
+      "grad_norm": 0.6837076544761658,
+      "learning_rate": 7.895752895752897e-06,
+      "loss": 0.571,
+      "step": 410
+    },
+    {
+      "epoch": 0.07941261713844074,
+      "grad_norm": 0.6891146302223206,
+      "learning_rate": 7.915057915057915e-06,
+      "loss": 0.5585,
+      "step": 411
+    },
+    {
+      "epoch": 0.07960583518500627,
+      "grad_norm": 0.7227412462234497,
+      "learning_rate": 7.934362934362934e-06,
+      "loss": 0.5427,
+      "step": 412
+    },
+    {
+      "epoch": 0.07979905323157183,
+      "grad_norm": 0.7727156281471252,
+      "learning_rate": 7.953667953667954e-06,
+      "loss": 0.6652,
+      "step": 413
+    },
+    {
+      "epoch": 0.07999227127813738,
+      "grad_norm": 0.7511888742446899,
+      "learning_rate": 7.972972972972974e-06,
+      "loss": 0.6586,
+      "step": 414
+    },
+    {
+      "epoch": 0.08018548932470293,
+      "grad_norm": 0.7671670913696289,
+      "learning_rate": 7.992277992277993e-06,
+      "loss": 0.6018,
+      "step": 415
+    },
+    {
+      "epoch": 0.08037870737126848,
+      "grad_norm": 0.7388387322425842,
+      "learning_rate": 8.011583011583011e-06,
+      "loss": 0.5884,
+      "step": 416
+    },
+    {
+      "epoch": 0.08057192541783402,
+      "grad_norm": 0.7498449087142944,
+      "learning_rate": 8.03088803088803e-06,
+      "loss": 0.6689,
+      "step": 417
+    },
+    {
+      "epoch": 0.08076514346439957,
+      "grad_norm": 0.7684760093688965,
+      "learning_rate": 8.05019305019305e-06,
+      "loss": 0.6346,
+      "step": 418
+    },
+    {
+      "epoch": 0.08095836151096512,
+      "grad_norm": 0.726571261882782,
+      "learning_rate": 8.06949806949807e-06,
+      "loss": 0.6108,
+      "step": 419
+    },
+    {
+      "epoch": 0.08115157955753068,
+      "grad_norm": 0.7431203722953796,
+      "learning_rate": 8.08880308880309e-06,
+      "loss": 0.648,
+      "step": 420
+    },
+    {
+      "epoch": 0.08134479760409623,
+      "grad_norm": 0.7155138850212097,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.5751,
+      "step": 421
+    },
+    {
+      "epoch": 0.08153801565066177,
+      "grad_norm": 0.8111143112182617,
+      "learning_rate": 8.127413127413129e-06,
+      "loss": 0.6697,
+      "step": 422
+    },
+    {
+      "epoch": 0.08173123369722732,
+      "grad_norm": 0.8521466851234436,
+      "learning_rate": 8.146718146718148e-06,
+      "loss": 0.6944,
+      "step": 423
+    },
+    {
+      "epoch": 0.08192445174379287,
+      "grad_norm": 0.8040545582771301,
+      "learning_rate": 8.166023166023168e-06,
+      "loss": 0.7023,
+      "step": 424
+    },
+    {
+      "epoch": 0.08211766979035842,
+      "grad_norm": 0.7952727675437927,
+      "learning_rate": 8.185328185328186e-06,
+      "loss": 0.5846,
+      "step": 425
+    },
+    {
+      "epoch": 0.08231088783692397,
+      "grad_norm": 0.8531398177146912,
+      "learning_rate": 8.204633204633205e-06,
+      "loss": 0.6676,
+      "step": 426
+    },
+    {
+      "epoch": 0.08250410588348951,
+      "grad_norm": 0.7578323483467102,
+      "learning_rate": 8.223938223938225e-06,
+      "loss": 0.6389,
+      "step": 427
+    },
+    {
+      "epoch": 0.08269732393005506,
+      "grad_norm": 0.7510884404182434,
+      "learning_rate": 8.243243243243245e-06,
+      "loss": 0.7203,
+      "step": 428
+    },
+    {
+      "epoch": 0.08289054197662062,
+      "grad_norm": 0.7194547057151794,
+      "learning_rate": 8.262548262548264e-06,
+      "loss": 0.5113,
+      "step": 429
+    },
+    {
+      "epoch": 0.08308376002318617,
+      "grad_norm": 0.7199673056602478,
+      "learning_rate": 8.281853281853282e-06,
+      "loss": 0.6118,
+      "step": 430
+    },
+    {
+      "epoch": 0.08327697806975172,
+      "grad_norm": 0.7789958715438843,
+      "learning_rate": 8.301158301158302e-06,
+      "loss": 0.6682,
+      "step": 431
+    },
+    {
+      "epoch": 0.08347019611631726,
+      "grad_norm": 0.7987354397773743,
+      "learning_rate": 8.320463320463321e-06,
+      "loss": 0.6892,
+      "step": 432
+    },
+    {
+      "epoch": 0.08366341416288281,
+      "grad_norm": 0.7253822684288025,
+      "learning_rate": 8.33976833976834e-06,
+      "loss": 0.6507,
+      "step": 433
+    },
+    {
+      "epoch": 0.08385663220944836,
+      "grad_norm": 0.6939993500709534,
+      "learning_rate": 8.35907335907336e-06,
+      "loss": 0.6217,
+      "step": 434
+    },
+    {
+      "epoch": 0.08404985025601391,
+      "grad_norm": 0.8271985650062561,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.6617,
+      "step": 435
+    },
+    {
+      "epoch": 0.08424306830257947,
+      "grad_norm": 0.7391234040260315,
+      "learning_rate": 8.397683397683398e-06,
+      "loss": 0.685,
+      "step": 436
+    },
+    {
+      "epoch": 0.084436286349145,
+      "grad_norm": 0.7405696511268616,
+      "learning_rate": 8.416988416988418e-06,
+      "loss": 0.7177,
+      "step": 437
+    },
+    {
+      "epoch": 0.08462950439571056,
+      "grad_norm": 0.8010821342468262,
+      "learning_rate": 8.436293436293437e-06,
+      "loss": 0.6423,
+      "step": 438
+    },
+    {
+      "epoch": 0.08482272244227611,
+      "grad_norm": 0.7726134061813354,
+      "learning_rate": 8.455598455598457e-06,
+      "loss": 0.6392,
+      "step": 439
+    },
+    {
+      "epoch": 0.08501594048884166,
+      "grad_norm": 0.7739795446395874,
+      "learning_rate": 8.474903474903475e-06,
+      "loss": 0.629,
+      "step": 440
+    },
+    {
+      "epoch": 0.08520915853540721,
+      "grad_norm": 0.6706112623214722,
+      "learning_rate": 8.494208494208494e-06,
+      "loss": 0.5835,
+      "step": 441
+    },
+    {
+      "epoch": 0.08540237658197275,
+      "grad_norm": 0.7952768802642822,
+      "learning_rate": 8.513513513513514e-06,
+      "loss": 0.6409,
+      "step": 442
+    },
+    {
+      "epoch": 0.0855955946285383,
+      "grad_norm": 0.7198907136917114,
+      "learning_rate": 8.532818532818533e-06,
+      "loss": 0.6308,
+      "step": 443
+    },
+    {
+      "epoch": 0.08578881267510385,
+      "grad_norm": 0.6810705661773682,
+      "learning_rate": 8.552123552123553e-06,
+      "loss": 0.6457,
+      "step": 444
+    },
+    {
+      "epoch": 0.0859820307216694,
+      "grad_norm": 0.7231850624084473,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.5989,
+      "step": 445
+    },
+    {
+      "epoch": 0.08617524876823496,
+      "grad_norm": 0.7254806160926819,
+      "learning_rate": 8.59073359073359e-06,
+      "loss": 0.6692,
+      "step": 446
+    },
+    {
+      "epoch": 0.0863684668148005,
+      "grad_norm": 0.8079766631126404,
+      "learning_rate": 8.61003861003861e-06,
+      "loss": 0.6398,
+      "step": 447
+    },
+    {
+      "epoch": 0.08656168486136605,
+      "grad_norm": 0.743241012096405,
+      "learning_rate": 8.629343629343631e-06,
+      "loss": 0.616,
+      "step": 448
+    },
+    {
+      "epoch": 0.0867549029079316,
+      "grad_norm": 0.7463287711143494,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.6456,
+      "step": 449
+    },
+    {
+      "epoch": 0.08694812095449715,
+      "grad_norm": 0.7957670092582703,
+      "learning_rate": 8.667953667953669e-06,
+      "loss": 0.7197,
+      "step": 450
+    },
+    {
+      "epoch": 0.0871413390010627,
+      "grad_norm": 0.7901537418365479,
+      "learning_rate": 8.687258687258689e-06,
+      "loss": 0.5558,
+      "step": 451
+    },
+    {
+      "epoch": 0.08733455704762824,
+      "grad_norm": 0.7044151425361633,
+      "learning_rate": 8.706563706563708e-06,
+      "loss": 0.6327,
+      "step": 452
+    },
+    {
+      "epoch": 0.0875277750941938,
+      "grad_norm": 0.7957638502120972,
+      "learning_rate": 8.725868725868728e-06,
+      "loss": 0.6138,
+      "step": 453
+    },
+    {
+      "epoch": 0.08772099314075935,
+      "grad_norm": 0.7256485819816589,
+      "learning_rate": 8.745173745173746e-06,
+      "loss": 0.6114,
+      "step": 454
+    },
+    {
+      "epoch": 0.0879142111873249,
+      "grad_norm": 0.7707110643386841,
+      "learning_rate": 8.764478764478765e-06,
+      "loss": 0.6928,
+      "step": 455
+    },
+    {
+      "epoch": 0.08810742923389045,
+      "grad_norm": 0.8261554837226868,
+      "learning_rate": 8.783783783783785e-06,
+      "loss": 0.6277,
+      "step": 456
+    },
+    {
+      "epoch": 0.08830064728045599,
+      "grad_norm": 0.7276001572608948,
+      "learning_rate": 8.803088803088804e-06,
+      "loss": 0.6542,
+      "step": 457
+    },
+    {
+      "epoch": 0.08849386532702154,
+      "grad_norm": 0.7561659216880798,
+      "learning_rate": 8.822393822393824e-06,
+      "loss": 0.6314,
+      "step": 458
+    },
+    {
+      "epoch": 0.08868708337358709,
+      "grad_norm": 0.8290273547172546,
+      "learning_rate": 8.841698841698842e-06,
+      "loss": 0.6197,
+      "step": 459
+    },
+    {
+      "epoch": 0.08888030142015264,
+      "grad_norm": 0.7558900713920593,
+      "learning_rate": 8.861003861003861e-06,
+      "loss": 0.6608,
+      "step": 460
+    },
+    {
+      "epoch": 0.0890735194667182,
+      "grad_norm": 0.7640666961669922,
+      "learning_rate": 8.880308880308881e-06,
+      "loss": 0.6224,
+      "step": 461
+    },
+    {
+      "epoch": 0.08926673751328375,
+      "grad_norm": 0.755632758140564,
+      "learning_rate": 8.8996138996139e-06,
+      "loss": 0.6604,
+      "step": 462
+    },
+    {
+      "epoch": 0.08945995555984929,
+      "grad_norm": 0.7849288582801819,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.6654,
+      "step": 463
+    },
+    {
+      "epoch": 0.08965317360641484,
+      "grad_norm": 0.7703944444656372,
+      "learning_rate": 8.938223938223938e-06,
+      "loss": 0.6333,
+      "step": 464
+    },
+    {
+      "epoch": 0.08984639165298039,
+      "grad_norm": 0.8416056632995605,
+      "learning_rate": 8.957528957528958e-06,
+      "loss": 0.6928,
+      "step": 465
+    },
+    {
+      "epoch": 0.09003960969954594,
+      "grad_norm": 0.8207576274871826,
+      "learning_rate": 8.976833976833977e-06,
+      "loss": 0.6233,
+      "step": 466
+    },
+    {
+      "epoch": 0.0902328277461115,
+      "grad_norm": 0.6913374066352844,
+      "learning_rate": 8.996138996138997e-06,
+      "loss": 0.5847,
+      "step": 467
+    },
+    {
+      "epoch": 0.09042604579267703,
+      "grad_norm": 0.7499940395355225,
+      "learning_rate": 9.015444015444017e-06,
+      "loss": 0.6906,
+      "step": 468
+    },
+    {
+      "epoch": 0.09061926383924258,
+      "grad_norm": 0.7583680748939514,
+      "learning_rate": 9.034749034749034e-06,
+      "loss": 0.7371,
+      "step": 469
+    },
+    {
+      "epoch": 0.09081248188580814,
+      "grad_norm": 0.7475177645683289,
+      "learning_rate": 9.054054054054054e-06,
+      "loss": 0.5465,
+      "step": 470
+    },
+    {
+      "epoch": 0.09100569993237369,
+      "grad_norm": 0.6993744373321533,
+      "learning_rate": 9.073359073359074e-06,
+      "loss": 0.5743,
+      "step": 471
+    },
+    {
+      "epoch": 0.09119891797893924,
+      "grad_norm": 0.7570707201957703,
+      "learning_rate": 9.092664092664093e-06,
+      "loss": 0.6563,
+      "step": 472
+    },
+    {
+      "epoch": 0.09139213602550478,
+      "grad_norm": 0.854119598865509,
+      "learning_rate": 9.111969111969113e-06,
+      "loss": 0.6546,
+      "step": 473
+    },
+    {
+      "epoch": 0.09158535407207033,
+      "grad_norm": 0.7280111312866211,
+      "learning_rate": 9.13127413127413e-06,
+      "loss": 0.6626,
+      "step": 474
+    },
+    {
+      "epoch": 0.09177857211863588,
+      "grad_norm": 0.764812171459198,
+      "learning_rate": 9.15057915057915e-06,
+      "loss": 0.6793,
+      "step": 475
+    },
+    {
+      "epoch": 0.09197179016520143,
+      "grad_norm": 0.7746539115905762,
+      "learning_rate": 9.169884169884172e-06,
+      "loss": 0.6942,
+      "step": 476
+    },
+    {
+      "epoch": 0.09216500821176699,
+      "grad_norm": 0.7495526075363159,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.6342,
+      "step": 477
+    },
+    {
+      "epoch": 0.09235822625833252,
+      "grad_norm": 0.8682729601860046,
+      "learning_rate": 9.20849420849421e-06,
+      "loss": 0.6865,
+      "step": 478
+    },
+    {
+      "epoch": 0.09255144430489808,
+      "grad_norm": 0.7776036858558655,
+      "learning_rate": 9.227799227799229e-06,
+      "loss": 0.6094,
+      "step": 479
+    },
+    {
+      "epoch": 0.09274466235146363,
+      "grad_norm": 0.7742635607719421,
+      "learning_rate": 9.247104247104248e-06,
+      "loss": 0.636,
+      "step": 480
+    },
+    {
+      "epoch": 0.09293788039802918,
+      "grad_norm": 0.8219203352928162,
+      "learning_rate": 9.266409266409268e-06,
+      "loss": 0.6922,
+      "step": 481
+    },
+    {
+      "epoch": 0.09313109844459473,
+      "grad_norm": 0.728735625743866,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 0.5862,
+      "step": 482
+    },
+    {
+      "epoch": 0.09332431649116027,
+      "grad_norm": 0.8185111284255981,
+      "learning_rate": 9.305019305019305e-06,
+      "loss": 0.6511,
+      "step": 483
+    },
+    {
+      "epoch": 0.09351753453772582,
+      "grad_norm": 0.762100100517273,
+      "learning_rate": 9.324324324324325e-06,
+      "loss": 0.6655,
+      "step": 484
+    },
+    {
+      "epoch": 0.09371075258429137,
+      "grad_norm": 0.7781388163566589,
+      "learning_rate": 9.343629343629345e-06,
+      "loss": 0.6357,
+      "step": 485
+    },
+    {
+      "epoch": 0.09390397063085693,
+      "grad_norm": 0.704109251499176,
+      "learning_rate": 9.362934362934364e-06,
+      "loss": 0.6145,
+      "step": 486
+    },
+    {
+      "epoch": 0.09409718867742248,
+      "grad_norm": 0.7168439626693726,
+      "learning_rate": 9.382239382239384e-06,
+      "loss": 0.6356,
+      "step": 487
+    },
+    {
+      "epoch": 0.09429040672398802,
+      "grad_norm": 0.777740478515625,
+      "learning_rate": 9.401544401544402e-06,
+      "loss": 0.6714,
+      "step": 488
+    },
+    {
+      "epoch": 0.09448362477055357,
+      "grad_norm": 0.7185985445976257,
+      "learning_rate": 9.420849420849421e-06,
+      "loss": 0.66,
+      "step": 489
+    },
+    {
+      "epoch": 0.09467684281711912,
+      "grad_norm": 0.6486513614654541,
+      "learning_rate": 9.440154440154441e-06,
+      "loss": 0.6213,
+      "step": 490
+    },
+    {
+      "epoch": 0.09487006086368467,
+      "grad_norm": 0.6498697400093079,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.6238,
+      "step": 491
+    },
+    {
+      "epoch": 0.09506327891025022,
+      "grad_norm": 0.6611917018890381,
+      "learning_rate": 9.47876447876448e-06,
+      "loss": 0.6167,
+      "step": 492
+    },
+    {
+      "epoch": 0.09525649695681576,
+      "grad_norm": 0.7942330837249756,
+      "learning_rate": 9.498069498069498e-06,
+      "loss": 0.644,
+      "step": 493
+    },
+    {
+      "epoch": 0.09544971500338131,
+      "grad_norm": 0.7540901303291321,
+      "learning_rate": 9.517374517374518e-06,
+      "loss": 0.6224,
+      "step": 494
+    },
+    {
+      "epoch": 0.09564293304994687,
+      "grad_norm": 0.7334563732147217,
+      "learning_rate": 9.536679536679537e-06,
+      "loss": 0.6635,
+      "step": 495
+    },
+    {
+      "epoch": 0.09583615109651242,
+      "grad_norm": 0.7598876953125,
+      "learning_rate": 9.555984555984557e-06,
+      "loss": 0.6464,
+      "step": 496
+    },
+    {
+      "epoch": 0.09602936914307797,
+      "grad_norm": 0.8426600694656372,
+      "learning_rate": 9.575289575289576e-06,
+      "loss": 0.6606,
+      "step": 497
+    },
+    {
+      "epoch": 0.09622258718964351,
+      "grad_norm": 0.7364344596862793,
+      "learning_rate": 9.594594594594594e-06,
+      "loss": 0.5588,
+      "step": 498
+    },
+    {
+      "epoch": 0.09641580523620906,
+      "grad_norm": 0.8193737268447876,
+      "learning_rate": 9.613899613899614e-06,
+      "loss": 0.6676,
+      "step": 499
+    },
+    {
+      "epoch": 0.09660902328277461,
+      "grad_norm": 0.7770460844039917,
+      "learning_rate": 9.633204633204634e-06,
+      "loss": 0.6144,
+      "step": 500
+    },
+    {
+      "epoch": 0.09680224132934016,
+      "grad_norm": 0.7298727035522461,
+      "learning_rate": 9.652509652509653e-06,
+      "loss": 0.7085,
+      "step": 501
+    },
+    {
+      "epoch": 0.09699545937590572,
+      "grad_norm": 0.7573222517967224,
+      "learning_rate": 9.671814671814673e-06,
+      "loss": 0.7089,
+      "step": 502
+    },
+    {
+      "epoch": 0.09718867742247125,
+      "grad_norm": 0.8758789896965027,
+      "learning_rate": 9.69111969111969e-06,
+      "loss": 0.7028,
+      "step": 503
+    },
+    {
+      "epoch": 0.0973818954690368,
+      "grad_norm": 0.8487024903297424,
+      "learning_rate": 9.71042471042471e-06,
+      "loss": 0.6986,
+      "step": 504
+    },
+    {
+      "epoch": 0.09757511351560236,
+      "grad_norm": 0.8770574331283569,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.7298,
+      "step": 505
+    },
+    {
+      "epoch": 0.09776833156216791,
+      "grad_norm": 0.6735337376594543,
+      "learning_rate": 9.749034749034751e-06,
+      "loss": 0.588,
+      "step": 506
+    },
+    {
+      "epoch": 0.09796154960873346,
+      "grad_norm": 0.7934773564338684,
+      "learning_rate": 9.768339768339769e-06,
+      "loss": 0.6192,
+      "step": 507
+    },
+    {
+      "epoch": 0.098154767655299,
+      "grad_norm": 0.8215080499649048,
+      "learning_rate": 9.787644787644789e-06,
+      "loss": 0.6965,
+      "step": 508
+    },
+    {
+      "epoch": 0.09834798570186455,
+      "grad_norm": 0.7771565318107605,
+      "learning_rate": 9.806949806949808e-06,
+      "loss": 0.6304,
+      "step": 509
+    },
+    {
+      "epoch": 0.0985412037484301,
+      "grad_norm": 0.6850326061248779,
+      "learning_rate": 9.826254826254828e-06,
+      "loss": 0.6822,
+      "step": 510
+    },
+    {
+      "epoch": 0.09873442179499566,
+      "grad_norm": 0.7376495599746704,
+      "learning_rate": 9.845559845559847e-06,
+      "loss": 0.5352,
+      "step": 511
+    },
+    {
+      "epoch": 0.09892763984156121,
+      "grad_norm": 0.7366458177566528,
+      "learning_rate": 9.864864864864865e-06,
+      "loss": 0.621,
+      "step": 512
+    },
+    {
+      "epoch": 0.09912085788812675,
+      "grad_norm": 0.7167019248008728,
+      "learning_rate": 9.884169884169885e-06,
+      "loss": 0.5745,
+      "step": 513
+    },
+    {
+      "epoch": 0.0993140759346923,
+      "grad_norm": 0.7086433172225952,
+      "learning_rate": 9.903474903474905e-06,
+      "loss": 0.571,
+      "step": 514
+    },
+    {
+      "epoch": 0.09950729398125785,
+      "grad_norm": 0.7631819844245911,
+      "learning_rate": 9.922779922779924e-06,
+      "loss": 0.6967,
+      "step": 515
+    },
+    {
+      "epoch": 0.0997005120278234,
+      "grad_norm": 0.7007446885108948,
+      "learning_rate": 9.942084942084944e-06,
+      "loss": 0.5941,
+      "step": 516
+    },
+    {
+      "epoch": 0.09989373007438895,
+      "grad_norm": 0.7603797912597656,
+      "learning_rate": 9.961389961389962e-06,
+      "loss": 0.669,
+      "step": 517
+    },
+    {
+      "epoch": 0.10008694812095449,
+      "grad_norm": 0.7833032608032227,
+      "learning_rate": 9.980694980694981e-06,
+      "loss": 0.6543,
+      "step": 518
+    },
+    {
+      "epoch": 0.10028016616752004,
+      "grad_norm": 0.7879395484924316,
+      "learning_rate": 1e-05,
+      "loss": 0.6379,
+      "step": 519
+    },
+    {
+      "epoch": 0.1004733842140856,
+      "grad_norm": 0.7661585807800293,
+      "learning_rate": 9.99999886278974e-06,
+      "loss": 0.656,
+      "step": 520
+    },
+    {
+      "epoch": 0.10066660226065115,
+      "grad_norm": 0.7644357085227966,
+      "learning_rate": 9.999995451159479e-06,
+      "loss": 0.6666,
+      "step": 521
+    },
+    {
+      "epoch": 0.1008598203072167,
+      "grad_norm": 0.7764670252799988,
+      "learning_rate": 9.999989765110767e-06,
+      "loss": 0.6906,
+      "step": 522
+    },
+    {
+      "epoch": 0.10105303835378224,
+      "grad_norm": 0.776177167892456,
+      "learning_rate": 9.999981804646191e-06,
+      "loss": 0.7302,
+      "step": 523
+    },
+    {
+      "epoch": 0.10124625640034779,
+      "grad_norm": 0.6752447485923767,
+      "learning_rate": 9.999971569769372e-06,
+      "loss": 0.5705,
+      "step": 524
+    },
+    {
+      "epoch": 0.10143947444691334,
+      "grad_norm": 0.7966018319129944,
+      "learning_rate": 9.999959060484965e-06,
+      "loss": 0.6779,
+      "step": 525
+    },
+    {
+      "epoch": 0.1016326924934789,
+      "grad_norm": 0.7739578485488892,
+      "learning_rate": 9.99994427679866e-06,
+      "loss": 0.6943,
+      "step": 526
+    },
+    {
+      "epoch": 0.10182591054004445,
+      "grad_norm": 0.7972790598869324,
+      "learning_rate": 9.999927218717186e-06,
+      "loss": 0.6083,
+      "step": 527
+    },
+    {
+      "epoch": 0.10201912858660998,
+      "grad_norm": 0.8121469616889954,
+      "learning_rate": 9.999907886248297e-06,
+      "loss": 0.6826,
+      "step": 528
+    },
+    {
+      "epoch": 0.10221234663317554,
+      "grad_norm": 0.7667931318283081,
+      "learning_rate": 9.999886279400791e-06,
+      "loss": 0.6461,
+      "step": 529
+    },
+    {
+      "epoch": 0.10240556467974109,
+      "grad_norm": 0.7282577753067017,
+      "learning_rate": 9.999862398184495e-06,
+      "loss": 0.6622,
+      "step": 530
+    },
+    {
+      "epoch": 0.10259878272630664,
+      "grad_norm": 0.8754635453224182,
+      "learning_rate": 9.999836242610272e-06,
+      "loss": 0.6921,
+      "step": 531
+    },
+    {
+      "epoch": 0.10279200077287219,
+      "grad_norm": 0.7410080432891846,
+      "learning_rate": 9.99980781269002e-06,
+      "loss": 0.5857,
+      "step": 532
+    },
+    {
+      "epoch": 0.10298521881943773,
+      "grad_norm": 0.7450238466262817,
+      "learning_rate": 9.99977710843667e-06,
+      "loss": 0.5998,
+      "step": 533
+    },
+    {
+      "epoch": 0.10317843686600328,
+      "grad_norm": 0.7782400250434875,
+      "learning_rate": 9.999744129864193e-06,
+      "loss": 0.666,
+      "step": 534
+    },
+    {
+      "epoch": 0.10337165491256883,
+      "grad_norm": 0.7188642024993896,
+      "learning_rate": 9.999708876987587e-06,
+      "loss": 0.6114,
+      "step": 535
+    },
+    {
+      "epoch": 0.10356487295913439,
+      "grad_norm": 0.8192875385284424,
+      "learning_rate": 9.999671349822887e-06,
+      "loss": 0.6506,
+      "step": 536
+    },
+    {
+      "epoch": 0.10375809100569994,
+      "grad_norm": 0.7616800665855408,
+      "learning_rate": 9.999631548387165e-06,
+      "loss": 0.6531,
+      "step": 537
+    },
+    {
+      "epoch": 0.10395130905226548,
+      "grad_norm": 0.7818080186843872,
+      "learning_rate": 9.999589472698528e-06,
+      "loss": 0.6793,
+      "step": 538
+    },
+    {
+      "epoch": 0.10414452709883103,
+      "grad_norm": 0.6944840550422668,
+      "learning_rate": 9.999545122776112e-06,
+      "loss": 0.559,
+      "step": 539
+    },
+    {
+      "epoch": 0.10433774514539658,
+      "grad_norm": 0.7725492119789124,
+      "learning_rate": 9.999498498640094e-06,
+      "loss": 0.6518,
+      "step": 540
+    },
+    {
+      "epoch": 0.10453096319196213,
+      "grad_norm": 0.7020063996315002,
+      "learning_rate": 9.999449600311681e-06,
+      "loss": 0.6528,
+      "step": 541
+    },
+    {
+      "epoch": 0.10472418123852768,
+      "grad_norm": 0.8014616370201111,
+      "learning_rate": 9.999398427813117e-06,
+      "loss": 0.6619,
+      "step": 542
+    },
+    {
+      "epoch": 0.10491739928509322,
+      "grad_norm": 0.727471113204956,
+      "learning_rate": 9.999344981167678e-06,
+      "loss": 0.596,
+      "step": 543
+    },
+    {
+      "epoch": 0.10511061733165877,
+      "grad_norm": 0.7258539199829102,
+      "learning_rate": 9.999289260399679e-06,
+      "loss": 0.6378,
+      "step": 544
+    },
+    {
+      "epoch": 0.10530383537822433,
+      "grad_norm": 0.7421310544013977,
+      "learning_rate": 9.999231265534462e-06,
+      "loss": 0.5804,
+      "step": 545
+    },
+    {
+      "epoch": 0.10549705342478988,
+      "grad_norm": 0.7222425937652588,
+      "learning_rate": 9.999170996598412e-06,
+      "loss": 0.5923,
+      "step": 546
+    },
+    {
+      "epoch": 0.10569027147135543,
+      "grad_norm": 0.7106494903564453,
+      "learning_rate": 9.999108453618943e-06,
+      "loss": 0.7042,
+      "step": 547
+    },
+    {
+      "epoch": 0.10588348951792097,
+      "grad_norm": 0.7394096851348877,
+      "learning_rate": 9.999043636624503e-06,
+      "loss": 0.684,
+      "step": 548
+    },
+    {
+      "epoch": 0.10607670756448652,
+      "grad_norm": 0.779862105846405,
+      "learning_rate": 9.998976545644579e-06,
+      "loss": 0.6145,
+      "step": 549
+    },
+    {
+      "epoch": 0.10626992561105207,
+      "grad_norm": 0.7379212379455566,
+      "learning_rate": 9.99890718070969e-06,
+      "loss": 0.661,
+      "step": 550
+    },
+    {
+      "epoch": 0.10646314365761762,
+      "grad_norm": 0.7846714854240417,
+      "learning_rate": 9.998835541851385e-06,
+      "loss": 0.6985,
+      "step": 551
+    },
+    {
+      "epoch": 0.10665636170418318,
+      "grad_norm": 0.8030348420143127,
+      "learning_rate": 9.998761629102254e-06,
+      "loss": 0.587,
+      "step": 552
+    },
+    {
+      "epoch": 0.10684957975074871,
+      "grad_norm": 0.8147480487823486,
+      "learning_rate": 9.998685442495921e-06,
+      "loss": 0.6627,
+      "step": 553
+    },
+    {
+      "epoch": 0.10704279779731427,
+      "grad_norm": 0.7512419819831848,
+      "learning_rate": 9.998606982067039e-06,
+      "loss": 0.6047,
+      "step": 554
+    },
+    {
+      "epoch": 0.10723601584387982,
+      "grad_norm": 0.8134625554084778,
+      "learning_rate": 9.998526247851297e-06,
+      "loss": 0.7137,
+      "step": 555
+    },
+    {
+      "epoch": 0.10742923389044537,
+      "grad_norm": 0.7168849110603333,
+      "learning_rate": 9.998443239885423e-06,
+      "loss": 0.7142,
+      "step": 556
+    },
+    {
+      "epoch": 0.10762245193701092,
+      "grad_norm": 0.7309911251068115,
+      "learning_rate": 9.998357958207175e-06,
+      "loss": 0.6239,
+      "step": 557
+    },
+    {
+      "epoch": 0.10781566998357646,
+      "grad_norm": 0.736663281917572,
+      "learning_rate": 9.998270402855346e-06,
+      "loss": 0.7206,
+      "step": 558
+    },
+    {
+      "epoch": 0.10800888803014201,
+      "grad_norm": 0.7385422587394714,
+      "learning_rate": 9.998180573869764e-06,
+      "loss": 0.6474,
+      "step": 559
+    },
+    {
+      "epoch": 0.10820210607670756,
+      "grad_norm": 0.8205428123474121,
+      "learning_rate": 9.998088471291289e-06,
+      "loss": 0.6772,
+      "step": 560
+    },
+    {
+      "epoch": 0.10839532412327312,
+      "grad_norm": 0.6974841356277466,
+      "learning_rate": 9.99799409516182e-06,
+      "loss": 0.5969,
+      "step": 561
+    },
+    {
+      "epoch": 0.10858854216983867,
+      "grad_norm": 0.7291163206100464,
+      "learning_rate": 9.997897445524287e-06,
+      "loss": 0.6567,
+      "step": 562
+    },
+    {
+      "epoch": 0.1087817602164042,
+      "grad_norm": 0.8007906079292297,
+      "learning_rate": 9.99779852242265e-06,
+      "loss": 0.6343,
+      "step": 563
+    },
+    {
+      "epoch": 0.10897497826296976,
+      "grad_norm": 0.7016527652740479,
+      "learning_rate": 9.997697325901911e-06,
+      "loss": 0.6596,
+      "step": 564
+    },
+    {
+      "epoch": 0.10916819630953531,
+      "grad_norm": 0.7910249829292297,
+      "learning_rate": 9.997593856008103e-06,
+      "loss": 0.644,
+      "step": 565
+    },
+    {
+      "epoch": 0.10936141435610086,
+      "grad_norm": 0.7845814228057861,
+      "learning_rate": 9.997488112788293e-06,
+      "loss": 0.7068,
+      "step": 566
+    },
+    {
+      "epoch": 0.10955463240266641,
+      "grad_norm": 0.7159735560417175,
+      "learning_rate": 9.99738009629058e-06,
+      "loss": 0.6291,
+      "step": 567
+    },
+    {
+      "epoch": 0.10974785044923197,
+      "grad_norm": 0.8179486393928528,
+      "learning_rate": 9.9972698065641e-06,
+      "loss": 0.6645,
+      "step": 568
+    },
+    {
+      "epoch": 0.1099410684957975,
+      "grad_norm": 0.6620157361030579,
+      "learning_rate": 9.997157243659021e-06,
+      "loss": 0.6351,
+      "step": 569
+    },
+    {
+      "epoch": 0.11013428654236306,
+      "grad_norm": 0.7134221196174622,
+      "learning_rate": 9.99704240762655e-06,
+      "loss": 0.6526,
+      "step": 570
+    },
+    {
+      "epoch": 0.11032750458892861,
+      "grad_norm": 0.7630340456962585,
+      "learning_rate": 9.996925298518918e-06,
+      "loss": 0.6056,
+      "step": 571
+    },
+    {
+      "epoch": 0.11052072263549416,
+      "grad_norm": 0.6866285800933838,
+      "learning_rate": 9.9968059163894e-06,
+      "loss": 0.6393,
+      "step": 572
+    },
+    {
+      "epoch": 0.11071394068205971,
+      "grad_norm": 0.8204681873321533,
+      "learning_rate": 9.9966842612923e-06,
+      "loss": 0.7267,
+      "step": 573
+    },
+    {
+      "epoch": 0.11090715872862525,
+      "grad_norm": 0.7321428656578064,
+      "learning_rate": 9.996560333282959e-06,
+      "loss": 0.6183,
+      "step": 574
+    },
+    {
+      "epoch": 0.1111003767751908,
+      "grad_norm": 0.7335955500602722,
+      "learning_rate": 9.996434132417746e-06,
+      "loss": 0.596,
+      "step": 575
+    },
+    {
+      "epoch": 0.11129359482175635,
+      "grad_norm": 0.7591193318367004,
+      "learning_rate": 9.99630565875407e-06,
+      "loss": 0.6043,
+      "step": 576
+    },
+    {
+      "epoch": 0.1114868128683219,
+      "grad_norm": 0.6604169011116028,
+      "learning_rate": 9.996174912350373e-06,
+      "loss": 0.5662,
+      "step": 577
+    },
+    {
+      "epoch": 0.11168003091488746,
+      "grad_norm": 0.6825835704803467,
+      "learning_rate": 9.996041893266127e-06,
+      "loss": 0.5786,
+      "step": 578
+    },
+    {
+      "epoch": 0.111873248961453,
+      "grad_norm": 0.739543080329895,
+      "learning_rate": 9.995906601561842e-06,
+      "loss": 0.6276,
+      "step": 579
+    },
+    {
+      "epoch": 0.11206646700801855,
+      "grad_norm": 0.7757747173309326,
+      "learning_rate": 9.995769037299059e-06,
+      "loss": 0.5961,
+      "step": 580
+    },
+    {
+      "epoch": 0.1122596850545841,
+      "grad_norm": 0.7062150239944458,
+      "learning_rate": 9.995629200540353e-06,
+      "loss": 0.6355,
+      "step": 581
+    },
+    {
+      "epoch": 0.11245290310114965,
+      "grad_norm": 0.8962317109107971,
+      "learning_rate": 9.995487091349336e-06,
+      "loss": 0.7759,
+      "step": 582
+    },
+    {
+      "epoch": 0.1126461211477152,
+      "grad_norm": 0.6767408847808838,
+      "learning_rate": 9.995342709790649e-06,
+      "loss": 0.5932,
+      "step": 583
+    },
+    {
+      "epoch": 0.11283933919428074,
+      "grad_norm": 0.7329311966896057,
+      "learning_rate": 9.99519605592997e-06,
+      "loss": 0.6012,
+      "step": 584
+    },
+    {
+      "epoch": 0.1130325572408463,
+      "grad_norm": 0.8223921060562134,
+      "learning_rate": 9.995047129834008e-06,
+      "loss": 0.6406,
+      "step": 585
+    },
+    {
+      "epoch": 0.11322577528741185,
+      "grad_norm": 0.7667902708053589,
+      "learning_rate": 9.99489593157051e-06,
+      "loss": 0.6981,
+      "step": 586
+    },
+    {
+      "epoch": 0.1134189933339774,
+      "grad_norm": 0.7068151235580444,
+      "learning_rate": 9.994742461208251e-06,
+      "loss": 0.653,
+      "step": 587
+    },
+    {
+      "epoch": 0.11361221138054295,
+      "grad_norm": 0.7372594475746155,
+      "learning_rate": 9.994586718817044e-06,
+      "loss": 0.5767,
+      "step": 588
+    },
+    {
+      "epoch": 0.11380542942710849,
+      "grad_norm": 0.7604547142982483,
+      "learning_rate": 9.994428704467733e-06,
+      "loss": 0.6242,
+      "step": 589
+    },
+    {
+      "epoch": 0.11399864747367404,
+      "grad_norm": 0.734743058681488,
+      "learning_rate": 9.994268418232195e-06,
+      "loss": 0.5884,
+      "step": 590
+    },
+    {
+      "epoch": 0.11419186552023959,
+      "grad_norm": 0.796778678894043,
+      "learning_rate": 9.994105860183344e-06,
+      "loss": 0.6194,
+      "step": 591
+    },
+    {
+      "epoch": 0.11438508356680514,
+      "grad_norm": 0.7543562054634094,
+      "learning_rate": 9.993941030395123e-06,
+      "loss": 0.6246,
+      "step": 592
+    },
+    {
+      "epoch": 0.1145783016133707,
+      "grad_norm": 0.7375812530517578,
+      "learning_rate": 9.993773928942512e-06,
+      "loss": 0.6395,
+      "step": 593
+    },
+    {
+      "epoch": 0.11477151965993623,
+      "grad_norm": 0.7615134716033936,
+      "learning_rate": 9.993604555901521e-06,
+      "loss": 0.5984,
+      "step": 594
+    },
+    {
+      "epoch": 0.11496473770650179,
+      "grad_norm": 0.78591388463974,
+      "learning_rate": 9.993432911349198e-06,
+      "loss": 0.6622,
+      "step": 595
+    },
+    {
+      "epoch": 0.11515795575306734,
+      "grad_norm": 0.7809131741523743,
+      "learning_rate": 9.993258995363619e-06,
+      "loss": 0.7616,
+      "step": 596
+    },
+    {
+      "epoch": 0.11535117379963289,
+      "grad_norm": 0.7256309390068054,
+      "learning_rate": 9.993082808023895e-06,
+      "loss": 0.5682,
+      "step": 597
+    },
+    {
+      "epoch": 0.11554439184619844,
+      "grad_norm": 0.748842179775238,
+      "learning_rate": 9.992904349410174e-06,
+      "loss": 0.562,
+      "step": 598
+    },
+    {
+      "epoch": 0.11573760989276398,
+      "grad_norm": 0.7882017493247986,
+      "learning_rate": 9.99272361960363e-06,
+      "loss": 0.6592,
+      "step": 599
+    },
+    {
+      "epoch": 0.11593082793932953,
+      "grad_norm": 0.7553358674049377,
+      "learning_rate": 9.992540618686479e-06,
+      "loss": 0.6628,
+      "step": 600
+    },
+    {
+      "epoch": 0.11612404598589508,
+      "grad_norm": 0.7014592885971069,
+      "learning_rate": 9.99235534674196e-06,
+      "loss": 0.648,
+      "step": 601
+    },
+    {
+      "epoch": 0.11631726403246063,
+      "grad_norm": 0.8125733733177185,
+      "learning_rate": 9.992167803854354e-06,
+      "loss": 0.7001,
+      "step": 602
+    },
+    {
+      "epoch": 0.11651048207902619,
+      "grad_norm": 0.8237801790237427,
+      "learning_rate": 9.99197799010897e-06,
+      "loss": 0.6613,
+      "step": 603
+    },
+    {
+      "epoch": 0.11670370012559173,
+      "grad_norm": 0.8320958018302917,
+      "learning_rate": 9.991785905592149e-06,
+      "loss": 0.6533,
+      "step": 604
+    },
+    {
+      "epoch": 0.11689691817215728,
+      "grad_norm": 0.7681236863136292,
+      "learning_rate": 9.99159155039127e-06,
+      "loss": 0.6468,
+      "step": 605
+    },
+    {
+      "epoch": 0.11709013621872283,
+      "grad_norm": 0.7672390937805176,
+      "learning_rate": 9.991394924594742e-06,
+      "loss": 0.6235,
+      "step": 606
+    },
+    {
+      "epoch": 0.11728335426528838,
+      "grad_norm": 0.760800302028656,
+      "learning_rate": 9.991196028292008e-06,
+      "loss": 0.6564,
+      "step": 607
+    },
+    {
+      "epoch": 0.11747657231185393,
+      "grad_norm": 0.7138139009475708,
+      "learning_rate": 9.99099486157354e-06,
+      "loss": 0.6394,
+      "step": 608
+    },
+    {
+      "epoch": 0.11766979035841947,
+      "grad_norm": 0.7287961840629578,
+      "learning_rate": 9.990791424530846e-06,
+      "loss": 0.6256,
+      "step": 609
+    },
+    {
+      "epoch": 0.11786300840498502,
+      "grad_norm": 0.7088921070098877,
+      "learning_rate": 9.990585717256467e-06,
+      "loss": 0.6663,
+      "step": 610
+    },
+    {
+      "epoch": 0.11805622645155057,
+      "grad_norm": 0.7732431292533875,
+      "learning_rate": 9.990377739843978e-06,
+      "loss": 0.6697,
+      "step": 611
+    },
+    {
+      "epoch": 0.11824944449811613,
+      "grad_norm": 0.6658932566642761,
+      "learning_rate": 9.99016749238798e-06,
+      "loss": 0.6343,
+      "step": 612
+    },
+    {
+      "epoch": 0.11844266254468168,
+      "grad_norm": 0.7996414303779602,
+      "learning_rate": 9.989954974984115e-06,
+      "loss": 0.6232,
+      "step": 613
+    },
+    {
+      "epoch": 0.11863588059124722,
+      "grad_norm": 0.7707719802856445,
+      "learning_rate": 9.989740187729052e-06,
+      "loss": 0.6322,
+      "step": 614
+    },
+    {
+      "epoch": 0.11882909863781277,
+      "grad_norm": 0.7807687520980835,
+      "learning_rate": 9.989523130720495e-06,
+      "loss": 0.6735,
+      "step": 615
+    },
+    {
+      "epoch": 0.11902231668437832,
+      "grad_norm": 0.8232107162475586,
+      "learning_rate": 9.98930380405718e-06,
+      "loss": 0.7323,
+      "step": 616
+    },
+    {
+      "epoch": 0.11921553473094387,
+      "grad_norm": 0.7644092440605164,
+      "learning_rate": 9.989082207838876e-06,
+      "loss": 0.7264,
+      "step": 617
+    },
+    {
+      "epoch": 0.11940875277750942,
+      "grad_norm": 0.7119615077972412,
+      "learning_rate": 9.988858342166381e-06,
+      "loss": 0.6829,
+      "step": 618
+    },
+    {
+      "epoch": 0.11960197082407496,
+      "grad_norm": 0.748570442199707,
+      "learning_rate": 9.98863220714153e-06,
+      "loss": 0.6335,
+      "step": 619
+    },
+    {
+      "epoch": 0.11979518887064051,
+      "grad_norm": 0.8471317887306213,
+      "learning_rate": 9.988403802867186e-06,
+      "loss": 0.6934,
+      "step": 620
+    },
+    {
+      "epoch": 0.11998840691720607,
+      "grad_norm": 0.6740917563438416,
+      "learning_rate": 9.988173129447251e-06,
+      "loss": 0.6051,
+      "step": 621
+    },
+    {
+      "epoch": 0.12018162496377162,
+      "grad_norm": 0.7320598363876343,
+      "learning_rate": 9.98794018698665e-06,
+      "loss": 0.6217,
+      "step": 622
+    },
+    {
+      "epoch": 0.12037484301033717,
+      "grad_norm": 0.7318382859230042,
+      "learning_rate": 9.987704975591348e-06,
+      "loss": 0.6585,
+      "step": 623
+    },
+    {
+      "epoch": 0.12056806105690271,
+      "grad_norm": 0.7271731495857239,
+      "learning_rate": 9.987467495368336e-06,
+      "loss": 0.6291,
+      "step": 624
+    },
+    {
+      "epoch": 0.12076127910346826,
+      "grad_norm": 0.7863685488700867,
+      "learning_rate": 9.987227746425642e-06,
+      "loss": 0.6155,
+      "step": 625
+    },
+    {
+      "epoch": 0.12095449715003381,
+      "grad_norm": 0.716116189956665,
+      "learning_rate": 9.986985728872324e-06,
+      "loss": 0.5949,
+      "step": 626
+    },
+    {
+      "epoch": 0.12114771519659936,
+      "grad_norm": 0.6721984148025513,
+      "learning_rate": 9.98674144281847e-06,
+      "loss": 0.5723,
+      "step": 627
+    },
+    {
+      "epoch": 0.12134093324316492,
+      "grad_norm": 0.8436257839202881,
+      "learning_rate": 9.986494888375206e-06,
+      "loss": 0.7152,
+      "step": 628
+    },
+    {
+      "epoch": 0.12153415128973045,
+      "grad_norm": 0.8127397298812866,
+      "learning_rate": 9.986246065654681e-06,
+      "loss": 0.6708,
+      "step": 629
+    },
+    {
+      "epoch": 0.121727369336296,
+      "grad_norm": 0.7631845474243164,
+      "learning_rate": 9.985994974770084e-06,
+      "loss": 0.6266,
+      "step": 630
+    },
+    {
+      "epoch": 0.12192058738286156,
+      "grad_norm": 0.6958643794059753,
+      "learning_rate": 9.985741615835628e-06,
+      "loss": 0.5709,
+      "step": 631
+    },
+    {
+      "epoch": 0.12211380542942711,
+      "grad_norm": 0.7603067755699158,
+      "learning_rate": 9.985485988966567e-06,
+      "loss": 0.6165,
+      "step": 632
+    },
+    {
+      "epoch": 0.12230702347599266,
+      "grad_norm": 0.7445492744445801,
+      "learning_rate": 9.985228094279177e-06,
+      "loss": 0.6852,
+      "step": 633
+    },
+    {
+      "epoch": 0.1225002415225582,
+      "grad_norm": 0.8421475291252136,
+      "learning_rate": 9.984967931890776e-06,
+      "loss": 0.7237,
+      "step": 634
+    },
+    {
+      "epoch": 0.12269345956912375,
+      "grad_norm": 0.6901448369026184,
+      "learning_rate": 9.984705501919703e-06,
+      "loss": 0.6052,
+      "step": 635
+    },
+    {
+      "epoch": 0.1228866776156893,
+      "grad_norm": 0.8226179480552673,
+      "learning_rate": 9.984440804485332e-06,
+      "loss": 0.6906,
+      "step": 636
+    },
+    {
+      "epoch": 0.12307989566225486,
+      "grad_norm": 0.7424408197402954,
+      "learning_rate": 9.984173839708075e-06,
+      "loss": 0.6502,
+      "step": 637
+    },
+    {
+      "epoch": 0.12327311370882041,
+      "grad_norm": 0.7147969603538513,
+      "learning_rate": 9.983904607709365e-06,
+      "loss": 0.5881,
+      "step": 638
+    },
+    {
+      "epoch": 0.12346633175538595,
+      "grad_norm": 0.7510132193565369,
+      "learning_rate": 9.983633108611675e-06,
+      "loss": 0.6448,
+      "step": 639
+    },
+    {
+      "epoch": 0.1236595498019515,
+      "grad_norm": 0.6837280988693237,
+      "learning_rate": 9.983359342538504e-06,
+      "loss": 0.6279,
+      "step": 640
+    },
+    {
+      "epoch": 0.12385276784851705,
+      "grad_norm": 0.773493766784668,
+      "learning_rate": 9.983083309614382e-06,
+      "loss": 0.7016,
+      "step": 641
+    },
+    {
+      "epoch": 0.1240459858950826,
+      "grad_norm": 0.7472678422927856,
+      "learning_rate": 9.982805009964877e-06,
+      "loss": 0.5828,
+      "step": 642
+    },
+    {
+      "epoch": 0.12423920394164815,
+      "grad_norm": 0.7765797972679138,
+      "learning_rate": 9.982524443716578e-06,
+      "loss": 0.6976,
+      "step": 643
+    },
+    {
+      "epoch": 0.12443242198821369,
+      "grad_norm": 0.767511248588562,
+      "learning_rate": 9.982241610997113e-06,
+      "loss": 0.5924,
+      "step": 644
+    },
+    {
+      "epoch": 0.12462564003477924,
+      "grad_norm": 0.753898561000824,
+      "learning_rate": 9.981956511935138e-06,
+      "loss": 0.6223,
+      "step": 645
+    },
+    {
+      "epoch": 0.1248188580813448,
+      "grad_norm": 0.7366109490394592,
+      "learning_rate": 9.981669146660338e-06,
+      "loss": 0.6425,
+      "step": 646
+    },
+    {
+      "epoch": 0.12501207612791035,
+      "grad_norm": 0.7310240864753723,
+      "learning_rate": 9.981379515303432e-06,
+      "loss": 0.6266,
+      "step": 647
+    },
+    {
+      "epoch": 0.1252052941744759,
+      "grad_norm": 0.7246080636978149,
+      "learning_rate": 9.98108761799617e-06,
+      "loss": 0.6037,
+      "step": 648
+    },
+    {
+      "epoch": 0.12539851222104145,
+      "grad_norm": 0.7528015971183777,
+      "learning_rate": 9.98079345487133e-06,
+      "loss": 0.6696,
+      "step": 649
+    },
+    {
+      "epoch": 0.125591730267607,
+      "grad_norm": 0.7126857042312622,
+      "learning_rate": 9.980497026062722e-06,
+      "loss": 0.5934,
+      "step": 650
+    },
+    {
+      "epoch": 0.12578494831417256,
+      "grad_norm": 0.85768061876297,
+      "learning_rate": 9.980198331705186e-06,
+      "loss": 0.6014,
+      "step": 651
+    },
+    {
+      "epoch": 0.12597816636073808,
+      "grad_norm": 0.6953104734420776,
+      "learning_rate": 9.979897371934596e-06,
+      "loss": 0.6412,
+      "step": 652
+    },
+    {
+      "epoch": 0.12617138440730363,
+      "grad_norm": 0.7586202025413513,
+      "learning_rate": 9.979594146887851e-06,
+      "loss": 0.6196,
+      "step": 653
+    },
+    {
+      "epoch": 0.12636460245386918,
+      "grad_norm": 0.751448929309845,
+      "learning_rate": 9.979288656702887e-06,
+      "loss": 0.6697,
+      "step": 654
+    },
+    {
+      "epoch": 0.12655782050043474,
+      "grad_norm": 0.7141832709312439,
+      "learning_rate": 9.978980901518663e-06,
+      "loss": 0.6558,
+      "step": 655
+    },
+    {
+      "epoch": 0.1267510385470003,
+      "grad_norm": 0.7841053009033203,
+      "learning_rate": 9.978670881475173e-06,
+      "loss": 0.6687,
+      "step": 656
+    },
+    {
+      "epoch": 0.12694425659356584,
+      "grad_norm": 0.749960720539093,
+      "learning_rate": 9.978358596713442e-06,
+      "loss": 0.65,
+      "step": 657
+    },
+    {
+      "epoch": 0.1271374746401314,
+      "grad_norm": 0.7296048402786255,
+      "learning_rate": 9.978044047375519e-06,
+      "loss": 0.6388,
+      "step": 658
+    },
+    {
+      "epoch": 0.12733069268669694,
+      "grad_norm": 0.7258001565933228,
+      "learning_rate": 9.977727233604492e-06,
+      "loss": 0.6783,
+      "step": 659
+    },
+    {
+      "epoch": 0.1275239107332625,
+      "grad_norm": 0.640749454498291,
+      "learning_rate": 9.977408155544472e-06,
+      "loss": 0.5395,
+      "step": 660
+    },
+    {
+      "epoch": 0.12771712877982805,
+      "grad_norm": 0.7203798294067383,
+      "learning_rate": 9.977086813340604e-06,
+      "loss": 0.6057,
+      "step": 661
+    },
+    {
+      "epoch": 0.12791034682639357,
+      "grad_norm": 0.7271174788475037,
+      "learning_rate": 9.976763207139059e-06,
+      "loss": 0.5973,
+      "step": 662
+    },
+    {
+      "epoch": 0.12810356487295912,
+      "grad_norm": 0.7223747968673706,
+      "learning_rate": 9.976437337087045e-06,
+      "loss": 0.6355,
+      "step": 663
+    },
+    {
+      "epoch": 0.12829678291952468,
+      "grad_norm": 0.7359287738800049,
+      "learning_rate": 9.97610920333279e-06,
+      "loss": 0.6001,
+      "step": 664
+    },
+    {
+      "epoch": 0.12849000096609023,
+      "grad_norm": 0.739283561706543,
+      "learning_rate": 9.97577880602556e-06,
+      "loss": 0.6352,
+      "step": 665
+    },
+    {
+      "epoch": 0.12868321901265578,
+      "grad_norm": 0.7196642160415649,
+      "learning_rate": 9.975446145315646e-06,
+      "loss": 0.6733,
+      "step": 666
+    },
+    {
+      "epoch": 0.12887643705922133,
+      "grad_norm": 0.772009551525116,
+      "learning_rate": 9.975111221354372e-06,
+      "loss": 0.666,
+      "step": 667
+    },
+    {
+      "epoch": 0.12906965510578688,
+      "grad_norm": 0.7174028754234314,
+      "learning_rate": 9.974774034294086e-06,
+      "loss": 0.6483,
+      "step": 668
+    },
+    {
+      "epoch": 0.12926287315235244,
+      "grad_norm": 0.6723750829696655,
+      "learning_rate": 9.974434584288172e-06,
+      "loss": 0.6142,
+      "step": 669
+    },
+    {
+      "epoch": 0.129456091198918,
+      "grad_norm": 0.7130601406097412,
+      "learning_rate": 9.974092871491039e-06,
+      "loss": 0.5752,
+      "step": 670
+    },
+    {
+      "epoch": 0.12964930924548354,
+      "grad_norm": 0.6578310132026672,
+      "learning_rate": 9.973748896058129e-06,
+      "loss": 0.6062,
+      "step": 671
+    },
+    {
+      "epoch": 0.12984252729204906,
+      "grad_norm": 0.7860562205314636,
+      "learning_rate": 9.973402658145908e-06,
+      "loss": 0.6763,
+      "step": 672
+    },
+    {
+      "epoch": 0.13003574533861462,
+      "grad_norm": 0.7496817111968994,
+      "learning_rate": 9.973054157911877e-06,
+      "loss": 0.6441,
+      "step": 673
+    },
+    {
+      "epoch": 0.13022896338518017,
+      "grad_norm": 0.7753590941429138,
+      "learning_rate": 9.97270339551456e-06,
+      "loss": 0.6824,
+      "step": 674
+    },
+    {
+      "epoch": 0.13042218143174572,
+      "grad_norm": 0.7125529050827026,
+      "learning_rate": 9.972350371113514e-06,
+      "loss": 0.6246,
+      "step": 675
+    },
+    {
+      "epoch": 0.13061539947831127,
+      "grad_norm": 0.7891725301742554,
+      "learning_rate": 9.971995084869327e-06,
+      "loss": 0.6355,
+      "step": 676
+    },
+    {
+      "epoch": 0.13080861752487682,
+      "grad_norm": 0.8140126466751099,
+      "learning_rate": 9.97163753694361e-06,
+      "loss": 0.7053,
+      "step": 677
+    },
+    {
+      "epoch": 0.13100183557144238,
+      "grad_norm": 0.689814031124115,
+      "learning_rate": 9.971277727499008e-06,
+      "loss": 0.6119,
+      "step": 678
+    },
+    {
+      "epoch": 0.13119505361800793,
+      "grad_norm": 0.7719193696975708,
+      "learning_rate": 9.970915656699192e-06,
+      "loss": 0.76,
+      "step": 679
+    },
+    {
+      "epoch": 0.13138827166457348,
+      "grad_norm": 0.651871919631958,
+      "learning_rate": 9.970551324708859e-06,
+      "loss": 0.6341,
+      "step": 680
+    },
+    {
+      "epoch": 0.13158148971113903,
+      "grad_norm": 0.7516567707061768,
+      "learning_rate": 9.970184731693744e-06,
+      "loss": 0.637,
+      "step": 681
+    },
+    {
+      "epoch": 0.13177470775770456,
+      "grad_norm": 0.6947675943374634,
+      "learning_rate": 9.9698158778206e-06,
+      "loss": 0.6273,
+      "step": 682
+    },
+    {
+      "epoch": 0.1319679258042701,
+      "grad_norm": 0.672650933265686,
+      "learning_rate": 9.969444763257213e-06,
+      "loss": 0.6451,
+      "step": 683
+    },
+    {
+      "epoch": 0.13216114385083566,
+      "grad_norm": 0.7337961196899414,
+      "learning_rate": 9.969071388172397e-06,
+      "loss": 0.661,
+      "step": 684
+    },
+    {
+      "epoch": 0.1323543618974012,
+      "grad_norm": 0.7471703886985779,
+      "learning_rate": 9.968695752735996e-06,
+      "loss": 0.6803,
+      "step": 685
+    },
+    {
+      "epoch": 0.13254757994396676,
+      "grad_norm": 0.7267631888389587,
+      "learning_rate": 9.968317857118881e-06,
+      "loss": 0.6822,
+      "step": 686
+    },
+    {
+      "epoch": 0.13274079799053232,
+      "grad_norm": 0.7490827441215515,
+      "learning_rate": 9.967937701492949e-06,
+      "loss": 0.6043,
+      "step": 687
+    },
+    {
+      "epoch": 0.13293401603709787,
+      "grad_norm": 0.6996826529502869,
+      "learning_rate": 9.967555286031127e-06,
+      "loss": 0.7038,
+      "step": 688
+    },
+    {
+      "epoch": 0.13312723408366342,
+      "grad_norm": 0.714138388633728,
+      "learning_rate": 9.96717061090737e-06,
+      "loss": 0.6084,
+      "step": 689
+    },
+    {
+      "epoch": 0.13332045213022897,
+      "grad_norm": 0.6431013941764832,
+      "learning_rate": 9.966783676296661e-06,
+      "loss": 0.5479,
+      "step": 690
+    },
+    {
+      "epoch": 0.13351367017679452,
+      "grad_norm": 0.7326186895370483,
+      "learning_rate": 9.96639448237501e-06,
+      "loss": 0.6315,
+      "step": 691
+    },
+    {
+      "epoch": 0.13370688822336005,
+      "grad_norm": 0.6937079429626465,
+      "learning_rate": 9.966003029319455e-06,
+      "loss": 0.6329,
+      "step": 692
+    },
+    {
+      "epoch": 0.1339001062699256,
+      "grad_norm": 0.7571533918380737,
+      "learning_rate": 9.965609317308062e-06,
+      "loss": 0.7062,
+      "step": 693
+    },
+    {
+      "epoch": 0.13409332431649115,
+      "grad_norm": 0.7926445603370667,
+      "learning_rate": 9.965213346519925e-06,
+      "loss": 0.6375,
+      "step": 694
+    },
+    {
+      "epoch": 0.1342865423630567,
+      "grad_norm": 0.8156589865684509,
+      "learning_rate": 9.964815117135162e-06,
+      "loss": 0.5984,
+      "step": 695
+    },
+    {
+      "epoch": 0.13447976040962226,
+      "grad_norm": 0.7559399604797363,
+      "learning_rate": 9.964414629334925e-06,
+      "loss": 0.641,
+      "step": 696
+    },
+    {
+      "epoch": 0.1346729784561878,
+      "grad_norm": 0.7245908975601196,
+      "learning_rate": 9.964011883301387e-06,
+      "loss": 0.6214,
+      "step": 697
+    },
+    {
+      "epoch": 0.13486619650275336,
+      "grad_norm": 0.6850938200950623,
+      "learning_rate": 9.963606879217752e-06,
+      "loss": 0.6206,
+      "step": 698
+    },
+    {
+      "epoch": 0.1350594145493189,
+      "grad_norm": 0.7440631985664368,
+      "learning_rate": 9.963199617268249e-06,
+      "loss": 0.657,
+      "step": 699
+    },
+    {
+      "epoch": 0.13525263259588446,
+      "grad_norm": 0.7553303241729736,
+      "learning_rate": 9.962790097638137e-06,
+      "loss": 0.6211,
+      "step": 700
+    },
+    {
+      "epoch": 0.13544585064245002,
+      "grad_norm": 0.7305346131324768,
+      "learning_rate": 9.962378320513697e-06,
+      "loss": 0.6,
+      "step": 701
+    },
+    {
+      "epoch": 0.13563906868901554,
+      "grad_norm": 0.7052547931671143,
+      "learning_rate": 9.961964286082243e-06,
+      "loss": 0.598,
+      "step": 702
+    },
+    {
+      "epoch": 0.1358322867355811,
+      "grad_norm": 0.7291136384010315,
+      "learning_rate": 9.961547994532108e-06,
+      "loss": 0.6471,
+      "step": 703
+    },
+    {
+      "epoch": 0.13602550478214664,
+      "grad_norm": 0.8169398307800293,
+      "learning_rate": 9.961129446052662e-06,
+      "loss": 0.6796,
+      "step": 704
+    },
+    {
+      "epoch": 0.1362187228287122,
+      "grad_norm": 0.7460918426513672,
+      "learning_rate": 9.960708640834294e-06,
+      "loss": 0.6418,
+      "step": 705
+    },
+    {
+      "epoch": 0.13641194087527775,
+      "grad_norm": 0.7199252247810364,
+      "learning_rate": 9.960285579068419e-06,
+      "loss": 0.6418,
+      "step": 706
+    },
+    {
+      "epoch": 0.1366051589218433,
+      "grad_norm": 0.6634106040000916,
+      "learning_rate": 9.959860260947485e-06,
+      "loss": 0.6378,
+      "step": 707
+    },
+    {
+      "epoch": 0.13679837696840885,
+      "grad_norm": 0.7238237857818604,
+      "learning_rate": 9.959432686664959e-06,
+      "loss": 0.6087,
+      "step": 708
+    },
+    {
+      "epoch": 0.1369915950149744,
+      "grad_norm": 0.7191987037658691,
+      "learning_rate": 9.959002856415341e-06,
+      "loss": 0.581,
+      "step": 709
+    },
+    {
+      "epoch": 0.13718481306153996,
+      "grad_norm": 0.7042459845542908,
+      "learning_rate": 9.958570770394151e-06,
+      "loss": 0.6479,
+      "step": 710
+    },
+    {
+      "epoch": 0.1373780311081055,
+      "grad_norm": 0.6830907464027405,
+      "learning_rate": 9.95813642879794e-06,
+      "loss": 0.6421,
+      "step": 711
+    },
+    {
+      "epoch": 0.13757124915467103,
+      "grad_norm": 0.7301251888275146,
+      "learning_rate": 9.957699831824281e-06,
+      "loss": 0.6815,
+      "step": 712
+    },
+    {
+      "epoch": 0.13776446720123658,
+      "grad_norm": 0.6892144680023193,
+      "learning_rate": 9.957260979671779e-06,
+      "loss": 0.5987,
+      "step": 713
+    },
+    {
+      "epoch": 0.13795768524780214,
+      "grad_norm": 0.6976742148399353,
+      "learning_rate": 9.956819872540057e-06,
+      "loss": 0.646,
+      "step": 714
+    },
+    {
+      "epoch": 0.1381509032943677,
+      "grad_norm": 0.7400932312011719,
+      "learning_rate": 9.956376510629768e-06,
+      "loss": 0.6323,
+      "step": 715
+    },
+    {
+      "epoch": 0.13834412134093324,
+      "grad_norm": 0.7762027382850647,
+      "learning_rate": 9.955930894142593e-06,
+      "loss": 0.74,
+      "step": 716
+    },
+    {
+      "epoch": 0.1385373393874988,
+      "grad_norm": 0.6694858074188232,
+      "learning_rate": 9.955483023281234e-06,
+      "loss": 0.633,
+      "step": 717
+    },
+    {
+      "epoch": 0.13873055743406434,
+      "grad_norm": 0.6513929963111877,
+      "learning_rate": 9.955032898249419e-06,
+      "loss": 0.5835,
+      "step": 718
+    },
+    {
+      "epoch": 0.1389237754806299,
+      "grad_norm": 0.7037979364395142,
+      "learning_rate": 9.954580519251906e-06,
+      "loss": 0.6147,
+      "step": 719
+    },
+    {
+      "epoch": 0.13911699352719545,
+      "grad_norm": 0.6925401091575623,
+      "learning_rate": 9.954125886494473e-06,
+      "loss": 0.6193,
+      "step": 720
+    },
+    {
+      "epoch": 0.139310211573761,
+      "grad_norm": 0.7778042554855347,
+      "learning_rate": 9.953669000183922e-06,
+      "loss": 0.7344,
+      "step": 721
+    },
+    {
+      "epoch": 0.13950342962032655,
+      "grad_norm": 0.8259244561195374,
+      "learning_rate": 9.95320986052809e-06,
+      "loss": 0.7063,
+      "step": 722
+    },
+    {
+      "epoch": 0.13969664766689208,
+      "grad_norm": 0.6982406377792358,
+      "learning_rate": 9.95274846773583e-06,
+      "loss": 0.6236,
+      "step": 723
+    },
+    {
+      "epoch": 0.13988986571345763,
+      "grad_norm": 0.7595060467720032,
+      "learning_rate": 9.95228482201702e-06,
+      "loss": 0.6665,
+      "step": 724
+    },
+    {
+      "epoch": 0.14008308376002318,
+      "grad_norm": 0.6423277854919434,
+      "learning_rate": 9.951818923582565e-06,
+      "loss": 0.6416,
+      "step": 725
+    },
+    {
+      "epoch": 0.14027630180658873,
+      "grad_norm": 0.6666793823242188,
+      "learning_rate": 9.951350772644399e-06,
+      "loss": 0.5775,
+      "step": 726
+    },
+    {
+      "epoch": 0.14046951985315428,
+      "grad_norm": 0.7121587991714478,
+      "learning_rate": 9.950880369415471e-06,
+      "loss": 0.6138,
+      "step": 727
+    },
+    {
+      "epoch": 0.14066273789971984,
+      "grad_norm": 0.6590285897254944,
+      "learning_rate": 9.950407714109764e-06,
+      "loss": 0.6009,
+      "step": 728
+    },
+    {
+      "epoch": 0.1408559559462854,
+      "grad_norm": 0.667664110660553,
+      "learning_rate": 9.94993280694228e-06,
+      "loss": 0.6378,
+      "step": 729
+    },
+    {
+      "epoch": 0.14104917399285094,
+      "grad_norm": 0.7518263459205627,
+      "learning_rate": 9.949455648129046e-06,
+      "loss": 0.6065,
+      "step": 730
+    },
+    {
+      "epoch": 0.1412423920394165,
+      "grad_norm": 0.7076751589775085,
+      "learning_rate": 9.948976237887113e-06,
+      "loss": 0.5889,
+      "step": 731
+    },
+    {
+      "epoch": 0.14143561008598204,
+      "grad_norm": 0.7487091422080994,
+      "learning_rate": 9.94849457643456e-06,
+      "loss": 0.6454,
+      "step": 732
+    },
+    {
+      "epoch": 0.14162882813254757,
+      "grad_norm": 0.6946980357170105,
+      "learning_rate": 9.948010663990487e-06,
+      "loss": 0.6351,
+      "step": 733
+    },
+    {
+      "epoch": 0.14182204617911312,
+      "grad_norm": 0.6840375661849976,
+      "learning_rate": 9.947524500775016e-06,
+      "loss": 0.5837,
+      "step": 734
+    },
+    {
+      "epoch": 0.14201526422567867,
+      "grad_norm": 0.6377614736557007,
+      "learning_rate": 9.947036087009296e-06,
+      "loss": 0.5897,
+      "step": 735
+    },
+    {
+      "epoch": 0.14220848227224422,
+      "grad_norm": 0.6717380881309509,
+      "learning_rate": 9.946545422915499e-06,
+      "loss": 0.6293,
+      "step": 736
+    },
+    {
+      "epoch": 0.14240170031880978,
+      "grad_norm": 0.6944979429244995,
+      "learning_rate": 9.946052508716819e-06,
+      "loss": 0.6157,
+      "step": 737
+    },
+    {
+      "epoch": 0.14259491836537533,
+      "grad_norm": 0.6753131747245789,
+      "learning_rate": 9.945557344637475e-06,
+      "loss": 0.6814,
+      "step": 738
+    },
+    {
+      "epoch": 0.14278813641194088,
+      "grad_norm": 0.7655063271522522,
+      "learning_rate": 9.945059930902713e-06,
+      "loss": 0.6721,
+      "step": 739
+    },
+    {
+      "epoch": 0.14298135445850643,
+      "grad_norm": 0.7334776520729065,
+      "learning_rate": 9.944560267738792e-06,
+      "loss": 0.6543,
+      "step": 740
+    },
+    {
+      "epoch": 0.14317457250507198,
+      "grad_norm": 0.731667160987854,
+      "learning_rate": 9.944058355373005e-06,
+      "loss": 0.6479,
+      "step": 741
+    },
+    {
+      "epoch": 0.14336779055163754,
+      "grad_norm": 0.7681917548179626,
+      "learning_rate": 9.943554194033664e-06,
+      "loss": 0.6611,
+      "step": 742
+    },
+    {
+      "epoch": 0.14356100859820306,
+      "grad_norm": 0.682420015335083,
+      "learning_rate": 9.943047783950102e-06,
+      "loss": 0.6446,
+      "step": 743
+    },
+    {
+      "epoch": 0.1437542266447686,
+      "grad_norm": 0.754017174243927,
+      "learning_rate": 9.94253912535268e-06,
+      "loss": 0.6555,
+      "step": 744
+    },
+    {
+      "epoch": 0.14394744469133416,
+      "grad_norm": 0.7175251245498657,
+      "learning_rate": 9.942028218472776e-06,
+      "loss": 0.6039,
+      "step": 745
+    },
+    {
+      "epoch": 0.14414066273789972,
+      "grad_norm": 0.7785794138908386,
+      "learning_rate": 9.941515063542795e-06,
+      "loss": 0.6808,
+      "step": 746
+    },
+    {
+      "epoch": 0.14433388078446527,
+      "grad_norm": 0.7380942702293396,
+      "learning_rate": 9.940999660796161e-06,
+      "loss": 0.6349,
+      "step": 747
+    },
+    {
+      "epoch": 0.14452709883103082,
+      "grad_norm": 0.6802393794059753,
+      "learning_rate": 9.940482010467325e-06,
+      "loss": 0.6504,
+      "step": 748
+    },
+    {
+      "epoch": 0.14472031687759637,
+      "grad_norm": 0.6630122661590576,
+      "learning_rate": 9.939962112791757e-06,
+      "loss": 0.6022,
+      "step": 749
+    },
+    {
+      "epoch": 0.14491353492416192,
+      "grad_norm": 0.6767376661300659,
+      "learning_rate": 9.939439968005949e-06,
+      "loss": 0.6662,
+      "step": 750
+    },
+    {
+      "epoch": 0.14510675297072748,
+      "grad_norm": 0.7800834774971008,
+      "learning_rate": 9.938915576347416e-06,
+      "loss": 0.6926,
+      "step": 751
+    },
+    {
+      "epoch": 0.14529997101729303,
+      "grad_norm": 0.7076945900917053,
+      "learning_rate": 9.9383889380547e-06,
+      "loss": 0.6486,
+      "step": 752
+    },
+    {
+      "epoch": 0.14549318906385855,
+      "grad_norm": 0.7234511375427246,
+      "learning_rate": 9.937860053367354e-06,
+      "loss": 0.6686,
+      "step": 753
+    },
+    {
+      "epoch": 0.1456864071104241,
+      "grad_norm": 0.6710833311080933,
+      "learning_rate": 9.937328922525964e-06,
+      "loss": 0.6351,
+      "step": 754
+    },
+    {
+      "epoch": 0.14587962515698966,
+      "grad_norm": 0.7199174761772156,
+      "learning_rate": 9.936795545772131e-06,
+      "loss": 0.5968,
+      "step": 755
+    },
+    {
+      "epoch": 0.1460728432035552,
+      "grad_norm": 0.7537652850151062,
+      "learning_rate": 9.936259923348479e-06,
+      "loss": 0.6472,
+      "step": 756
+    },
+    {
+      "epoch": 0.14626606125012076,
+      "grad_norm": 0.7288406491279602,
+      "learning_rate": 9.935722055498655e-06,
+      "loss": 0.6208,
+      "step": 757
+    },
+    {
+      "epoch": 0.1464592792966863,
+      "grad_norm": 0.7358474135398865,
+      "learning_rate": 9.935181942467329e-06,
+      "loss": 0.6334,
+      "step": 758
+    },
+    {
+      "epoch": 0.14665249734325186,
+      "grad_norm": 0.6807871460914612,
+      "learning_rate": 9.934639584500186e-06,
+      "loss": 0.6118,
+      "step": 759
+    },
+    {
+      "epoch": 0.14684571538981742,
+      "grad_norm": 0.6565280556678772,
+      "learning_rate": 9.934094981843935e-06,
+      "loss": 0.5745,
+      "step": 760
+    },
+    {
+      "epoch": 0.14703893343638297,
+      "grad_norm": 0.8088560700416565,
+      "learning_rate": 9.933548134746313e-06,
+      "loss": 0.7629,
+      "step": 761
+    },
+    {
+      "epoch": 0.14723215148294852,
+      "grad_norm": 0.7415410280227661,
+      "learning_rate": 9.932999043456067e-06,
+      "loss": 0.6866,
+      "step": 762
+    },
+    {
+      "epoch": 0.14742536952951404,
+      "grad_norm": 0.7080822587013245,
+      "learning_rate": 9.932447708222972e-06,
+      "loss": 0.6502,
+      "step": 763
+    },
+    {
+      "epoch": 0.1476185875760796,
+      "grad_norm": 0.7238027453422546,
+      "learning_rate": 9.931894129297822e-06,
+      "loss": 0.6606,
+      "step": 764
+    },
+    {
+      "epoch": 0.14781180562264515,
+      "grad_norm": 0.7542324066162109,
+      "learning_rate": 9.931338306932429e-06,
+      "loss": 0.654,
+      "step": 765
+    },
+    {
+      "epoch": 0.1480050236692107,
+      "grad_norm": 0.7494989037513733,
+      "learning_rate": 9.93078024137963e-06,
+      "loss": 0.6169,
+      "step": 766
+    },
+    {
+      "epoch": 0.14819824171577625,
+      "grad_norm": 0.7121487259864807,
+      "learning_rate": 9.930219932893278e-06,
+      "loss": 0.7149,
+      "step": 767
+    },
+    {
+      "epoch": 0.1483914597623418,
+      "grad_norm": 0.69615638256073,
+      "learning_rate": 9.92965738172825e-06,
+      "loss": 0.6147,
+      "step": 768
+    },
+    {
+      "epoch": 0.14858467780890736,
+      "grad_norm": 0.7801277041435242,
+      "learning_rate": 9.929092588140442e-06,
+      "loss": 0.6722,
+      "step": 769
+    },
+    {
+      "epoch": 0.1487778958554729,
+      "grad_norm": 0.7837132215499878,
+      "learning_rate": 9.928525552386768e-06,
+      "loss": 0.6295,
+      "step": 770
+    },
+    {
+      "epoch": 0.14897111390203846,
+      "grad_norm": 0.72540283203125,
+      "learning_rate": 9.927956274725165e-06,
+      "loss": 0.668,
+      "step": 771
+    },
+    {
+      "epoch": 0.149164331948604,
+      "grad_norm": 0.6560445427894592,
+      "learning_rate": 9.927384755414588e-06,
+      "loss": 0.6263,
+      "step": 772
+    },
+    {
+      "epoch": 0.14935754999516954,
+      "grad_norm": 0.7203543186187744,
+      "learning_rate": 9.926810994715012e-06,
+      "loss": 0.683,
+      "step": 773
+    },
+    {
+      "epoch": 0.1495507680417351,
+      "grad_norm": 0.7603803873062134,
+      "learning_rate": 9.92623499288743e-06,
+      "loss": 0.7011,
+      "step": 774
+    },
+    {
+      "epoch": 0.14974398608830064,
+      "grad_norm": 0.7369344830513,
+      "learning_rate": 9.92565675019386e-06,
+      "loss": 0.635,
+      "step": 775
+    },
+    {
+      "epoch": 0.1499372041348662,
+      "grad_norm": 0.773717999458313,
+      "learning_rate": 9.92507626689733e-06,
+      "loss": 0.662,
+      "step": 776
+    },
+    {
+      "epoch": 0.15013042218143174,
+      "grad_norm": 0.809746503829956,
+      "learning_rate": 9.924493543261898e-06,
+      "loss": 0.6106,
+      "step": 777
+    },
+    {
+      "epoch": 0.1503236402279973,
+      "grad_norm": 0.7721818685531616,
+      "learning_rate": 9.923908579552633e-06,
+      "loss": 0.6658,
+      "step": 778
+    },
+    {
+      "epoch": 0.15051685827456285,
+      "grad_norm": 0.7636384963989258,
+      "learning_rate": 9.923321376035624e-06,
+      "loss": 0.6602,
+      "step": 779
+    },
+    {
+      "epoch": 0.1507100763211284,
+      "grad_norm": 0.694009006023407,
+      "learning_rate": 9.922731932977984e-06,
+      "loss": 0.6585,
+      "step": 780
+    },
+    {
+      "epoch": 0.15090329436769395,
+      "grad_norm": 0.8573810458183289,
+      "learning_rate": 9.92214025064784e-06,
+      "loss": 0.6432,
+      "step": 781
+    },
+    {
+      "epoch": 0.1510965124142595,
+      "grad_norm": 0.7059459686279297,
+      "learning_rate": 9.92154632931434e-06,
+      "loss": 0.6472,
+      "step": 782
+    },
+    {
+      "epoch": 0.15128973046082503,
+      "grad_norm": 0.6879667043685913,
+      "learning_rate": 9.920950169247646e-06,
+      "loss": 0.6377,
+      "step": 783
+    },
+    {
+      "epoch": 0.15148294850739058,
+      "grad_norm": 0.6491668224334717,
+      "learning_rate": 9.920351770718945e-06,
+      "loss": 0.6395,
+      "step": 784
+    },
+    {
+      "epoch": 0.15167616655395613,
+      "grad_norm": 0.6684377789497375,
+      "learning_rate": 9.919751134000437e-06,
+      "loss": 0.5478,
+      "step": 785
+    },
+    {
+      "epoch": 0.15186938460052168,
+      "grad_norm": 0.7460950613021851,
+      "learning_rate": 9.919148259365343e-06,
+      "loss": 0.6335,
+      "step": 786
+    },
+    {
+      "epoch": 0.15206260264708724,
+      "grad_norm": 0.7680968642234802,
+      "learning_rate": 9.9185431470879e-06,
+      "loss": 0.6216,
+      "step": 787
+    },
+    {
+      "epoch": 0.1522558206936528,
+      "grad_norm": 0.7967392206192017,
+      "learning_rate": 9.917935797443367e-06,
+      "loss": 0.6607,
+      "step": 788
+    },
+    {
+      "epoch": 0.15244903874021834,
+      "grad_norm": 0.6718658804893494,
+      "learning_rate": 9.917326210708014e-06,
+      "loss": 0.6649,
+      "step": 789
+    },
+    {
+      "epoch": 0.1526422567867839,
+      "grad_norm": 0.7101138830184937,
+      "learning_rate": 9.916714387159136e-06,
+      "loss": 0.5892,
+      "step": 790
+    },
+    {
+      "epoch": 0.15283547483334944,
+      "grad_norm": 0.7239555716514587,
+      "learning_rate": 9.916100327075038e-06,
+      "loss": 0.6153,
+      "step": 791
+    },
+    {
+      "epoch": 0.153028692879915,
+      "grad_norm": 0.8129069805145264,
+      "learning_rate": 9.915484030735048e-06,
+      "loss": 0.6413,
+      "step": 792
+    },
+    {
+      "epoch": 0.15322191092648052,
+      "grad_norm": 0.7073642611503601,
+      "learning_rate": 9.91486549841951e-06,
+      "loss": 0.6278,
+      "step": 793
+    },
+    {
+      "epoch": 0.15341512897304607,
+      "grad_norm": 0.7051330208778381,
+      "learning_rate": 9.914244730409783e-06,
+      "loss": 0.6284,
+      "step": 794
+    },
+    {
+      "epoch": 0.15360834701961162,
+      "grad_norm": 0.7022778987884521,
+      "learning_rate": 9.913621726988247e-06,
+      "loss": 0.5305,
+      "step": 795
+    },
+    {
+      "epoch": 0.15380156506617718,
+      "grad_norm": 0.6574632525444031,
+      "learning_rate": 9.912996488438294e-06,
+      "loss": 0.6231,
+      "step": 796
+    },
+    {
+      "epoch": 0.15399478311274273,
+      "grad_norm": 0.7044629454612732,
+      "learning_rate": 9.912369015044334e-06,
+      "loss": 0.6261,
+      "step": 797
+    },
+    {
+      "epoch": 0.15418800115930828,
+      "grad_norm": 0.7238366007804871,
+      "learning_rate": 9.911739307091797e-06,
+      "loss": 0.6751,
+      "step": 798
+    },
+    {
+      "epoch": 0.15438121920587383,
+      "grad_norm": 0.7676690220832825,
+      "learning_rate": 9.911107364867128e-06,
+      "loss": 0.6343,
+      "step": 799
+    },
+    {
+      "epoch": 0.15457443725243938,
+      "grad_norm": 0.7298150062561035,
+      "learning_rate": 9.910473188657785e-06,
+      "loss": 0.6317,
+      "step": 800
+    },
+    {
+      "epoch": 0.15476765529900494,
+      "grad_norm": 0.7356631755828857,
+      "learning_rate": 9.909836778752247e-06,
+      "loss": 0.6935,
+      "step": 801
+    },
+    {
+      "epoch": 0.1549608733455705,
+      "grad_norm": 0.7278707027435303,
+      "learning_rate": 9.909198135440005e-06,
+      "loss": 0.7407,
+      "step": 802
+    },
+    {
+      "epoch": 0.155154091392136,
+      "grad_norm": 0.7673940658569336,
+      "learning_rate": 9.908557259011568e-06,
+      "loss": 0.7554,
+      "step": 803
+    },
+    {
+      "epoch": 0.15534730943870156,
+      "grad_norm": 0.817825973033905,
+      "learning_rate": 9.90791414975846e-06,
+      "loss": 0.7267,
+      "step": 804
+    },
+    {
+      "epoch": 0.15554052748526712,
+      "grad_norm": 0.696797251701355,
+      "learning_rate": 9.907268807973222e-06,
+      "loss": 0.6305,
+      "step": 805
+    },
+    {
+      "epoch": 0.15573374553183267,
+      "grad_norm": 0.6991902589797974,
+      "learning_rate": 9.90662123394941e-06,
+      "loss": 0.6193,
+      "step": 806
+    },
+    {
+      "epoch": 0.15592696357839822,
+      "grad_norm": 0.6753354668617249,
+      "learning_rate": 9.905971427981596e-06,
+      "loss": 0.6217,
+      "step": 807
+    },
+    {
+      "epoch": 0.15612018162496377,
+      "grad_norm": 0.7213683724403381,
+      "learning_rate": 9.905319390365364e-06,
+      "loss": 0.6484,
+      "step": 808
+    },
+    {
+      "epoch": 0.15631339967152932,
+      "grad_norm": 0.7073823809623718,
+      "learning_rate": 9.904665121397317e-06,
+      "loss": 0.6216,
+      "step": 809
+    },
+    {
+      "epoch": 0.15650661771809488,
+      "grad_norm": 0.7155327796936035,
+      "learning_rate": 9.90400862137507e-06,
+      "loss": 0.6385,
+      "step": 810
+    },
+    {
+      "epoch": 0.15669983576466043,
+      "grad_norm": 0.7934998273849487,
+      "learning_rate": 9.903349890597256e-06,
+      "loss": 0.6984,
+      "step": 811
+    },
+    {
+      "epoch": 0.15689305381122598,
+      "grad_norm": 0.7773085832595825,
+      "learning_rate": 9.902688929363521e-06,
+      "loss": 0.6594,
+      "step": 812
+    },
+    {
+      "epoch": 0.1570862718577915,
+      "grad_norm": 0.6723622679710388,
+      "learning_rate": 9.902025737974527e-06,
+      "loss": 0.5784,
+      "step": 813
+    },
+    {
+      "epoch": 0.15727948990435706,
+      "grad_norm": 0.727562427520752,
+      "learning_rate": 9.901360316731947e-06,
+      "loss": 0.6189,
+      "step": 814
+    },
+    {
+      "epoch": 0.1574727079509226,
+      "grad_norm": 0.7750793099403381,
+      "learning_rate": 9.90069266593847e-06,
+      "loss": 0.6406,
+      "step": 815
+    },
+    {
+      "epoch": 0.15766592599748816,
+      "grad_norm": 0.7210448980331421,
+      "learning_rate": 9.900022785897802e-06,
+      "loss": 0.6306,
+      "step": 816
+    },
+    {
+      "epoch": 0.1578591440440537,
+      "grad_norm": 0.6876199841499329,
+      "learning_rate": 9.899350676914659e-06,
+      "loss": 0.614,
+      "step": 817
+    },
+    {
+      "epoch": 0.15805236209061926,
+      "grad_norm": 0.7228716611862183,
+      "learning_rate": 9.898676339294775e-06,
+      "loss": 0.6548,
+      "step": 818
+    },
+    {
+      "epoch": 0.15824558013718482,
+      "grad_norm": 0.7910527586936951,
+      "learning_rate": 9.897999773344892e-06,
+      "loss": 0.6728,
+      "step": 819
+    },
+    {
+      "epoch": 0.15843879818375037,
+      "grad_norm": 0.6828069686889648,
+      "learning_rate": 9.897320979372772e-06,
+      "loss": 0.5957,
+      "step": 820
+    },
+    {
+      "epoch": 0.15863201623031592,
+      "grad_norm": 0.6932550072669983,
+      "learning_rate": 9.896639957687188e-06,
+      "loss": 0.7035,
+      "step": 821
+    },
+    {
+      "epoch": 0.15882523427688147,
+      "grad_norm": 0.6872929334640503,
+      "learning_rate": 9.895956708597926e-06,
+      "loss": 0.6205,
+      "step": 822
+    },
+    {
+      "epoch": 0.15901845232344702,
+      "grad_norm": 0.6943690776824951,
+      "learning_rate": 9.89527123241578e-06,
+      "loss": 0.6235,
+      "step": 823
+    },
+    {
+      "epoch": 0.15921167037001255,
+      "grad_norm": 0.7092728018760681,
+      "learning_rate": 9.894583529452567e-06,
+      "loss": 0.6343,
+      "step": 824
+    },
+    {
+      "epoch": 0.1594048884165781,
+      "grad_norm": 0.6512922048568726,
+      "learning_rate": 9.893893600021112e-06,
+      "loss": 0.5834,
+      "step": 825
+    },
+    {
+      "epoch": 0.15959810646314365,
+      "grad_norm": 0.7075539827346802,
+      "learning_rate": 9.893201444435251e-06,
+      "loss": 0.6261,
+      "step": 826
+    },
+    {
+      "epoch": 0.1597913245097092,
+      "grad_norm": 0.6823797225952148,
+      "learning_rate": 9.892507063009837e-06,
+      "loss": 0.6058,
+      "step": 827
+    },
+    {
+      "epoch": 0.15998454255627476,
+      "grad_norm": 0.6512995362281799,
+      "learning_rate": 9.891810456060731e-06,
+      "loss": 0.5525,
+      "step": 828
+    },
+    {
+      "epoch": 0.1601777606028403,
+      "grad_norm": 0.7646318078041077,
+      "learning_rate": 9.891111623904809e-06,
+      "loss": 0.633,
+      "step": 829
+    },
+    {
+      "epoch": 0.16037097864940586,
+      "grad_norm": 0.656969428062439,
+      "learning_rate": 9.89041056685996e-06,
+      "loss": 0.6432,
+      "step": 830
+    },
+    {
+      "epoch": 0.1605641966959714,
+      "grad_norm": 0.7921852469444275,
+      "learning_rate": 9.889707285245081e-06,
+      "loss": 0.6577,
+      "step": 831
+    },
+    {
+      "epoch": 0.16075741474253696,
+      "grad_norm": 0.7091190814971924,
+      "learning_rate": 9.889001779380086e-06,
+      "loss": 0.6753,
+      "step": 832
+    },
+    {
+      "epoch": 0.16095063278910252,
+      "grad_norm": 0.6968415975570679,
+      "learning_rate": 9.888294049585897e-06,
+      "loss": 0.6252,
+      "step": 833
+    },
+    {
+      "epoch": 0.16114385083566804,
+      "grad_norm": 0.7402783632278442,
+      "learning_rate": 9.887584096184448e-06,
+      "loss": 0.7502,
+      "step": 834
+    },
+    {
+      "epoch": 0.1613370688822336,
+      "grad_norm": 0.7449678182601929,
+      "learning_rate": 9.88687191949869e-06,
+      "loss": 0.6438,
+      "step": 835
+    },
+    {
+      "epoch": 0.16153028692879914,
+      "grad_norm": 0.8242966532707214,
+      "learning_rate": 9.886157519852576e-06,
+      "loss": 0.6985,
+      "step": 836
+    },
+    {
+      "epoch": 0.1617235049753647,
+      "grad_norm": 0.6974904537200928,
+      "learning_rate": 9.885440897571076e-06,
+      "loss": 0.5987,
+      "step": 837
+    },
+    {
+      "epoch": 0.16191672302193025,
+      "grad_norm": 0.7118916511535645,
+      "learning_rate": 9.884722052980173e-06,
+      "loss": 0.6592,
+      "step": 838
+    },
+    {
+      "epoch": 0.1621099410684958,
+      "grad_norm": 0.7121632695198059,
+      "learning_rate": 9.884000986406856e-06,
+      "loss": 0.6533,
+      "step": 839
+    },
+    {
+      "epoch": 0.16230315911506135,
+      "grad_norm": 0.7060639262199402,
+      "learning_rate": 9.883277698179127e-06,
+      "loss": 0.5852,
+      "step": 840
+    },
+    {
+      "epoch": 0.1624963771616269,
+      "grad_norm": 0.6703097224235535,
+      "learning_rate": 9.882552188625999e-06,
+      "loss": 0.6923,
+      "step": 841
+    },
+    {
+      "epoch": 0.16268959520819246,
+      "grad_norm": 0.7657973170280457,
+      "learning_rate": 9.881824458077491e-06,
+      "loss": 0.6635,
+      "step": 842
+    },
+    {
+      "epoch": 0.162882813254758,
+      "grad_norm": 0.7365768551826477,
+      "learning_rate": 9.881094506864639e-06,
+      "loss": 0.6525,
+      "step": 843
+    },
+    {
+      "epoch": 0.16307603130132353,
+      "grad_norm": 0.7529985308647156,
+      "learning_rate": 9.880362335319487e-06,
+      "loss": 0.6992,
+      "step": 844
+    },
+    {
+      "epoch": 0.16326924934788908,
+      "grad_norm": 0.7844207882881165,
+      "learning_rate": 9.879627943775089e-06,
+      "loss": 0.5837,
+      "step": 845
+    },
+    {
+      "epoch": 0.16346246739445464,
+      "grad_norm": 0.6962164044380188,
+      "learning_rate": 9.878891332565504e-06,
+      "loss": 0.6741,
+      "step": 846
+    },
+    {
+      "epoch": 0.1636556854410202,
+      "grad_norm": 0.7505472898483276,
+      "learning_rate": 9.878152502025808e-06,
+      "loss": 0.7214,
+      "step": 847
+    },
+    {
+      "epoch": 0.16384890348758574,
+      "grad_norm": 0.7505288124084473,
+      "learning_rate": 9.877411452492083e-06,
+      "loss": 0.6995,
+      "step": 848
+    },
+    {
+      "epoch": 0.1640421215341513,
+      "grad_norm": 0.7398354411125183,
+      "learning_rate": 9.876668184301419e-06,
+      "loss": 0.5941,
+      "step": 849
+    },
+    {
+      "epoch": 0.16423533958071684,
+      "grad_norm": 0.7051838636398315,
+      "learning_rate": 9.875922697791917e-06,
+      "loss": 0.6421,
+      "step": 850
+    },
+    {
+      "epoch": 0.1644285576272824,
+      "grad_norm": 0.7330453991889954,
+      "learning_rate": 9.87517499330269e-06,
+      "loss": 0.6813,
+      "step": 851
+    },
+    {
+      "epoch": 0.16462177567384795,
+      "grad_norm": 0.6850847005844116,
+      "learning_rate": 9.874425071173853e-06,
+      "loss": 0.6148,
+      "step": 852
+    },
+    {
+      "epoch": 0.1648149937204135,
+      "grad_norm": 0.730191171169281,
+      "learning_rate": 9.873672931746537e-06,
+      "loss": 0.6712,
+      "step": 853
+    },
+    {
+      "epoch": 0.16500821176697902,
+      "grad_norm": 0.7278591394424438,
+      "learning_rate": 9.872918575362875e-06,
+      "loss": 0.737,
+      "step": 854
+    },
+    {
+      "epoch": 0.16520142981354458,
+      "grad_norm": 0.7418088316917419,
+      "learning_rate": 9.872162002366013e-06,
+      "loss": 0.6541,
+      "step": 855
+    },
+    {
+      "epoch": 0.16539464786011013,
+      "grad_norm": 0.7046576738357544,
+      "learning_rate": 9.871403213100106e-06,
+      "loss": 0.616,
+      "step": 856
+    },
+    {
+      "epoch": 0.16558786590667568,
+      "grad_norm": 0.7050150632858276,
+      "learning_rate": 9.870642207910313e-06,
+      "loss": 0.6482,
+      "step": 857
+    },
+    {
+      "epoch": 0.16578108395324123,
+      "grad_norm": 0.7141119837760925,
+      "learning_rate": 9.869878987142803e-06,
+      "loss": 0.6773,
+      "step": 858
+    },
+    {
+      "epoch": 0.16597430199980678,
+      "grad_norm": 0.6973040103912354,
+      "learning_rate": 9.869113551144754e-06,
+      "loss": 0.5943,
+      "step": 859
+    },
+    {
+      "epoch": 0.16616752004637234,
+      "grad_norm": 0.6604650020599365,
+      "learning_rate": 9.868345900264351e-06,
+      "loss": 0.647,
+      "step": 860
+    },
+    {
+      "epoch": 0.1663607380929379,
+      "grad_norm": 0.7217727899551392,
+      "learning_rate": 9.867576034850786e-06,
+      "loss": 0.6816,
+      "step": 861
+    },
+    {
+      "epoch": 0.16655395613950344,
+      "grad_norm": 0.6817984580993652,
+      "learning_rate": 9.866803955254257e-06,
+      "loss": 0.6273,
+      "step": 862
+    },
+    {
+      "epoch": 0.166747174186069,
+      "grad_norm": 0.7331098318099976,
+      "learning_rate": 9.866029661825974e-06,
+      "loss": 0.6975,
+      "step": 863
+    },
+    {
+      "epoch": 0.16694039223263452,
+      "grad_norm": 0.756726086139679,
+      "learning_rate": 9.865253154918145e-06,
+      "loss": 0.665,
+      "step": 864
+    },
+    {
+      "epoch": 0.16713361027920007,
+      "grad_norm": 0.761269211769104,
+      "learning_rate": 9.864474434883997e-06,
+      "loss": 0.7861,
+      "step": 865
+    },
+    {
+      "epoch": 0.16732682832576562,
+      "grad_norm": 0.8456854820251465,
+      "learning_rate": 9.863693502077754e-06,
+      "loss": 0.5765,
+      "step": 866
+    },
+    {
+      "epoch": 0.16752004637233117,
+      "grad_norm": 0.6804026365280151,
+      "learning_rate": 9.86291035685465e-06,
+      "loss": 0.6642,
+      "step": 867
+    },
+    {
+      "epoch": 0.16771326441889672,
+      "grad_norm": 0.7447768449783325,
+      "learning_rate": 9.862124999570925e-06,
+      "loss": 0.6205,
+      "step": 868
+    },
+    {
+      "epoch": 0.16790648246546228,
+      "grad_norm": 0.7686854600906372,
+      "learning_rate": 9.861337430583828e-06,
+      "loss": 0.6427,
+      "step": 869
+    },
+    {
+      "epoch": 0.16809970051202783,
+      "grad_norm": 0.6906816959381104,
+      "learning_rate": 9.86054765025161e-06,
+      "loss": 0.6462,
+      "step": 870
+    },
+    {
+      "epoch": 0.16829291855859338,
+      "grad_norm": 0.7194653153419495,
+      "learning_rate": 9.85975565893353e-06,
+      "loss": 0.6506,
+      "step": 871
+    },
+    {
+      "epoch": 0.16848613660515893,
+      "grad_norm": 0.771949052810669,
+      "learning_rate": 9.85896145698985e-06,
+      "loss": 0.6704,
+      "step": 872
+    },
+    {
+      "epoch": 0.16867935465172448,
+      "grad_norm": 0.7580637335777283,
+      "learning_rate": 9.858165044781843e-06,
+      "loss": 0.6975,
+      "step": 873
+    },
+    {
+      "epoch": 0.16887257269829,
+      "grad_norm": 0.6940417885780334,
+      "learning_rate": 9.857366422671784e-06,
+      "loss": 0.6479,
+      "step": 874
+    },
+    {
+      "epoch": 0.16906579074485556,
+      "grad_norm": 0.6890699863433838,
+      "learning_rate": 9.85656559102295e-06,
+      "loss": 0.6237,
+      "step": 875
+    },
+    {
+      "epoch": 0.1692590087914211,
+      "grad_norm": 0.7273939847946167,
+      "learning_rate": 9.85576255019963e-06,
+      "loss": 0.6262,
+      "step": 876
+    },
+    {
+      "epoch": 0.16945222683798666,
+      "grad_norm": 0.8277177810668945,
+      "learning_rate": 9.854957300567114e-06,
+      "loss": 0.7293,
+      "step": 877
+    },
+    {
+      "epoch": 0.16964544488455222,
+      "grad_norm": 0.7387230396270752,
+      "learning_rate": 9.854149842491695e-06,
+      "loss": 0.6104,
+      "step": 878
+    },
+    {
+      "epoch": 0.16983866293111777,
+      "grad_norm": 0.7462281584739685,
+      "learning_rate": 9.853340176340676e-06,
+      "loss": 0.7166,
+      "step": 879
+    },
+    {
+      "epoch": 0.17003188097768332,
+      "grad_norm": 0.7888413071632385,
+      "learning_rate": 9.852528302482361e-06,
+      "loss": 0.7019,
+      "step": 880
+    },
+    {
+      "epoch": 0.17022509902424887,
+      "grad_norm": 0.7410791516304016,
+      "learning_rate": 9.851714221286056e-06,
+      "loss": 0.6238,
+      "step": 881
+    },
+    {
+      "epoch": 0.17041831707081442,
+      "grad_norm": 0.726798951625824,
+      "learning_rate": 9.850897933122075e-06,
+      "loss": 0.6734,
+      "step": 882
+    },
+    {
+      "epoch": 0.17061153511737998,
+      "grad_norm": 0.738061785697937,
+      "learning_rate": 9.850079438361734e-06,
+      "loss": 0.6208,
+      "step": 883
+    },
+    {
+      "epoch": 0.1708047531639455,
+      "grad_norm": 0.7193439602851868,
+      "learning_rate": 9.849258737377354e-06,
+      "loss": 0.6321,
+      "step": 884
+    },
+    {
+      "epoch": 0.17099797121051105,
+      "grad_norm": 0.6632267236709595,
+      "learning_rate": 9.848435830542259e-06,
+      "loss": 0.4936,
+      "step": 885
+    },
+    {
+      "epoch": 0.1711911892570766,
+      "grad_norm": 1.0126943588256836,
+      "learning_rate": 9.847610718230775e-06,
+      "loss": 0.678,
+      "step": 886
+    },
+    {
+      "epoch": 0.17138440730364216,
+      "grad_norm": 0.7017629146575928,
+      "learning_rate": 9.846783400818234e-06,
+      "loss": 0.676,
+      "step": 887
+    },
+    {
+      "epoch": 0.1715776253502077,
+      "grad_norm": 0.7504415512084961,
+      "learning_rate": 9.84595387868097e-06,
+      "loss": 0.6828,
+      "step": 888
+    },
+    {
+      "epoch": 0.17177084339677326,
+      "grad_norm": 0.6952869296073914,
+      "learning_rate": 9.845122152196315e-06,
+      "loss": 0.6405,
+      "step": 889
+    },
+    {
+      "epoch": 0.1719640614433388,
+      "grad_norm": 0.6827472448348999,
+      "learning_rate": 9.844288221742614e-06,
+      "loss": 0.671,
+      "step": 890
+    },
+    {
+      "epoch": 0.17215727948990436,
+      "grad_norm": 0.6543591618537903,
+      "learning_rate": 9.843452087699206e-06,
+      "loss": 0.6438,
+      "step": 891
+    },
+    {
+      "epoch": 0.17235049753646992,
+      "grad_norm": 0.6773020029067993,
+      "learning_rate": 9.842613750446434e-06,
+      "loss": 0.5858,
+      "step": 892
+    },
+    {
+      "epoch": 0.17254371558303547,
+      "grad_norm": 0.7412886023521423,
+      "learning_rate": 9.841773210365646e-06,
+      "loss": 0.6282,
+      "step": 893
+    },
+    {
+      "epoch": 0.172736933629601,
+      "grad_norm": 0.7286463379859924,
+      "learning_rate": 9.84093046783919e-06,
+      "loss": 0.6568,
+      "step": 894
+    },
+    {
+      "epoch": 0.17293015167616654,
+      "grad_norm": 0.8202593326568604,
+      "learning_rate": 9.840085523250417e-06,
+      "loss": 0.7048,
+      "step": 895
+    },
+    {
+      "epoch": 0.1731233697227321,
+      "grad_norm": 0.6944611668586731,
+      "learning_rate": 9.839238376983676e-06,
+      "loss": 0.7231,
+      "step": 896
+    },
+    {
+      "epoch": 0.17331658776929765,
+      "grad_norm": 0.7896881699562073,
+      "learning_rate": 9.838389029424324e-06,
+      "loss": 0.6826,
+      "step": 897
+    },
+    {
+      "epoch": 0.1735098058158632,
+      "grad_norm": 0.6913198232650757,
+      "learning_rate": 9.837537480958712e-06,
+      "loss": 0.648,
+      "step": 898
+    },
+    {
+      "epoch": 0.17370302386242875,
+      "grad_norm": 0.7233971357345581,
+      "learning_rate": 9.8366837319742e-06,
+      "loss": 0.6879,
+      "step": 899
+    },
+    {
+      "epoch": 0.1738962419089943,
+      "grad_norm": 0.6996638774871826,
+      "learning_rate": 9.83582778285914e-06,
+      "loss": 0.6349,
+      "step": 900
+    },
+    {
+      "epoch": 0.17408945995555986,
+      "grad_norm": 0.736670196056366,
+      "learning_rate": 9.834969634002894e-06,
+      "loss": 0.6103,
+      "step": 901
+    },
+    {
+      "epoch": 0.1742826780021254,
+      "grad_norm": 0.7034814953804016,
+      "learning_rate": 9.834109285795819e-06,
+      "loss": 0.6078,
+      "step": 902
+    },
+    {
+      "epoch": 0.17447589604869096,
+      "grad_norm": 0.6867201328277588,
+      "learning_rate": 9.833246738629271e-06,
+      "loss": 0.5497,
+      "step": 903
+    },
+    {
+      "epoch": 0.17466911409525648,
+      "grad_norm": 0.6953839063644409,
+      "learning_rate": 9.832381992895613e-06,
+      "loss": 0.6362,
+      "step": 904
+    },
+    {
+      "epoch": 0.17486233214182204,
+      "grad_norm": 0.7174314856529236,
+      "learning_rate": 9.831515048988203e-06,
+      "loss": 0.5792,
+      "step": 905
+    },
+    {
+      "epoch": 0.1750555501883876,
+      "grad_norm": 0.6981253623962402,
+      "learning_rate": 9.830645907301396e-06,
+      "loss": 0.5864,
+      "step": 906
+    },
+    {
+      "epoch": 0.17524876823495314,
+      "grad_norm": 0.7186115384101868,
+      "learning_rate": 9.82977456823056e-06,
+      "loss": 0.5949,
+      "step": 907
+    },
+    {
+      "epoch": 0.1754419862815187,
+      "grad_norm": 0.6886196732521057,
+      "learning_rate": 9.828901032172042e-06,
+      "loss": 0.6154,
+      "step": 908
+    },
+    {
+      "epoch": 0.17563520432808424,
+      "grad_norm": 0.6701784133911133,
+      "learning_rate": 9.828025299523206e-06,
+      "loss": 0.5803,
+      "step": 909
+    },
+    {
+      "epoch": 0.1758284223746498,
+      "grad_norm": 0.7454864382743835,
+      "learning_rate": 9.82714737068241e-06,
+      "loss": 0.7303,
+      "step": 910
+    },
+    {
+      "epoch": 0.17602164042121535,
+      "grad_norm": 0.7165485620498657,
+      "learning_rate": 9.826267246049006e-06,
+      "loss": 0.6457,
+      "step": 911
+    },
+    {
+      "epoch": 0.1762148584677809,
+      "grad_norm": 0.6941471695899963,
+      "learning_rate": 9.82538492602335e-06,
+      "loss": 0.5841,
+      "step": 912
+    },
+    {
+      "epoch": 0.17640807651434645,
+      "grad_norm": 0.7169017791748047,
+      "learning_rate": 9.824500411006799e-06,
+      "loss": 0.6238,
+      "step": 913
+    },
+    {
+      "epoch": 0.17660129456091198,
+      "grad_norm": 0.7490513920783997,
+      "learning_rate": 9.8236137014017e-06,
+      "loss": 0.619,
+      "step": 914
+    },
+    {
+      "epoch": 0.17679451260747753,
+      "grad_norm": 0.6750937700271606,
+      "learning_rate": 9.822724797611405e-06,
+      "loss": 0.613,
+      "step": 915
+    },
+    {
+      "epoch": 0.17698773065404308,
+      "grad_norm": 0.689081072807312,
+      "learning_rate": 9.82183370004026e-06,
+      "loss": 0.6935,
+      "step": 916
+    },
+    {
+      "epoch": 0.17718094870060863,
+      "grad_norm": 0.6709827184677124,
+      "learning_rate": 9.820940409093616e-06,
+      "loss": 0.6269,
+      "step": 917
+    },
+    {
+      "epoch": 0.17737416674717418,
+      "grad_norm": 0.7775185108184814,
+      "learning_rate": 9.820044925177811e-06,
+      "loss": 0.6699,
+      "step": 918
+    },
+    {
+      "epoch": 0.17756738479373974,
+      "grad_norm": 0.689876139163971,
+      "learning_rate": 9.819147248700192e-06,
+      "loss": 0.6449,
+      "step": 919
+    },
+    {
+      "epoch": 0.1777606028403053,
+      "grad_norm": 0.719153642654419,
+      "learning_rate": 9.818247380069093e-06,
+      "loss": 0.6774,
+      "step": 920
+    },
+    {
+      "epoch": 0.17795382088687084,
+      "grad_norm": 0.8346672058105469,
+      "learning_rate": 9.817345319693854e-06,
+      "loss": 0.6221,
+      "step": 921
+    },
+    {
+      "epoch": 0.1781470389334364,
+      "grad_norm": 0.7349758148193359,
+      "learning_rate": 9.816441067984804e-06,
+      "loss": 0.7034,
+      "step": 922
+    },
+    {
+      "epoch": 0.17834025698000194,
+      "grad_norm": 0.7223201990127563,
+      "learning_rate": 9.815534625353275e-06,
+      "loss": 0.6216,
+      "step": 923
+    },
+    {
+      "epoch": 0.1785334750265675,
+      "grad_norm": 0.6294049024581909,
+      "learning_rate": 9.814625992211595e-06,
+      "loss": 0.6124,
+      "step": 924
+    },
+    {
+      "epoch": 0.17872669307313302,
+      "grad_norm": 0.7914825677871704,
+      "learning_rate": 9.813715168973083e-06,
+      "loss": 0.7306,
+      "step": 925
+    },
+    {
+      "epoch": 0.17891991111969857,
+      "grad_norm": 0.7039434909820557,
+      "learning_rate": 9.81280215605206e-06,
+      "loss": 0.5913,
+      "step": 926
+    },
+    {
+      "epoch": 0.17911312916626412,
+      "grad_norm": 0.7697689533233643,
+      "learning_rate": 9.811886953863841e-06,
+      "loss": 0.7215,
+      "step": 927
+    },
+    {
+      "epoch": 0.17930634721282968,
+      "grad_norm": 0.6863982677459717,
+      "learning_rate": 9.810969562824737e-06,
+      "loss": 0.6289,
+      "step": 928
+    },
+    {
+      "epoch": 0.17949956525939523,
+      "grad_norm": 0.7153484225273132,
+      "learning_rate": 9.810049983352055e-06,
+      "loss": 0.6141,
+      "step": 929
+    },
+    {
+      "epoch": 0.17969278330596078,
+      "grad_norm": 0.6959081888198853,
+      "learning_rate": 9.809128215864096e-06,
+      "loss": 0.6246,
+      "step": 930
+    },
+    {
+      "epoch": 0.17988600135252633,
+      "grad_norm": 0.707690954208374,
+      "learning_rate": 9.808204260780158e-06,
+      "loss": 0.6179,
+      "step": 931
+    },
+    {
+      "epoch": 0.18007921939909188,
+      "grad_norm": 0.7225919961929321,
+      "learning_rate": 9.807278118520533e-06,
+      "loss": 0.6127,
+      "step": 932
+    },
+    {
+      "epoch": 0.18027243744565744,
+      "grad_norm": 0.6621707677841187,
+      "learning_rate": 9.806349789506508e-06,
+      "loss": 0.6275,
+      "step": 933
+    },
+    {
+      "epoch": 0.180465655492223,
+      "grad_norm": 0.7168663144111633,
+      "learning_rate": 9.805419274160368e-06,
+      "loss": 0.6621,
+      "step": 934
+    },
+    {
+      "epoch": 0.1806588735387885,
+      "grad_norm": 0.6859185099601746,
+      "learning_rate": 9.804486572905387e-06,
+      "loss": 0.6626,
+      "step": 935
+    },
+    {
+      "epoch": 0.18085209158535406,
+      "grad_norm": 0.7123717665672302,
+      "learning_rate": 9.803551686165835e-06,
+      "loss": 0.6183,
+      "step": 936
+    },
+    {
+      "epoch": 0.18104530963191962,
+      "grad_norm": 0.6558787226676941,
+      "learning_rate": 9.80261461436698e-06,
+      "loss": 0.5685,
+      "step": 937
+    },
+    {
+      "epoch": 0.18123852767848517,
+      "grad_norm": 0.6965833306312561,
+      "learning_rate": 9.80167535793508e-06,
+      "loss": 0.5815,
+      "step": 938
+    },
+    {
+      "epoch": 0.18143174572505072,
+      "grad_norm": 0.7115914225578308,
+      "learning_rate": 9.800733917297384e-06,
+      "loss": 0.5854,
+      "step": 939
+    },
+    {
+      "epoch": 0.18162496377161627,
+      "grad_norm": 0.728861927986145,
+      "learning_rate": 9.799790292882145e-06,
+      "loss": 0.6996,
+      "step": 940
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 0.7389002442359924,
+      "learning_rate": 9.798844485118599e-06,
+      "loss": 0.6939,
+      "step": 941
+    },
+    {
+      "epoch": 0.18201139986474738,
+      "grad_norm": 0.6196081042289734,
+      "learning_rate": 9.797896494436981e-06,
+      "loss": 0.6435,
+      "step": 942
+    },
+    {
+      "epoch": 0.18220461791131293,
+      "grad_norm": 0.7334592938423157,
+      "learning_rate": 9.796946321268513e-06,
+      "loss": 0.5789,
+      "step": 943
+    },
+    {
+      "epoch": 0.18239783595787848,
+      "grad_norm": 0.7880694270133972,
+      "learning_rate": 9.795993966045418e-06,
+      "loss": 0.6785,
+      "step": 944
+    },
+    {
+      "epoch": 0.182591054004444,
+      "grad_norm": 0.7435905337333679,
+      "learning_rate": 9.795039429200904e-06,
+      "loss": 0.6534,
+      "step": 945
+    },
+    {
+      "epoch": 0.18278427205100956,
+      "grad_norm": 0.7651240229606628,
+      "learning_rate": 9.794082711169177e-06,
+      "loss": 0.6267,
+      "step": 946
+    },
+    {
+      "epoch": 0.1829774900975751,
+      "grad_norm": 0.6815835237503052,
+      "learning_rate": 9.79312381238543e-06,
+      "loss": 0.6596,
+      "step": 947
+    },
+    {
+      "epoch": 0.18317070814414066,
+      "grad_norm": 0.7414952516555786,
+      "learning_rate": 9.792162733285855e-06,
+      "loss": 0.65,
+      "step": 948
+    },
+    {
+      "epoch": 0.1833639261907062,
+      "grad_norm": 0.6607598066329956,
+      "learning_rate": 9.791199474307626e-06,
+      "loss": 0.609,
+      "step": 949
+    },
+    {
+      "epoch": 0.18355714423727176,
+      "grad_norm": 0.7302612066268921,
+      "learning_rate": 9.790234035888921e-06,
+      "loss": 0.6337,
+      "step": 950
+    },
+    {
+      "epoch": 0.18375036228383732,
+      "grad_norm": 0.7319920063018799,
+      "learning_rate": 9.789266418468898e-06,
+      "loss": 0.6472,
+      "step": 951
+    },
+    {
+      "epoch": 0.18394358033040287,
+      "grad_norm": 0.7246800065040588,
+      "learning_rate": 9.78829662248771e-06,
+      "loss": 0.6426,
+      "step": 952
+    },
+    {
+      "epoch": 0.18413679837696842,
+      "grad_norm": 0.7328210473060608,
+      "learning_rate": 9.787324648386505e-06,
+      "loss": 0.6748,
+      "step": 953
+    },
+    {
+      "epoch": 0.18433001642353397,
+      "grad_norm": 0.6572837233543396,
+      "learning_rate": 9.786350496607418e-06,
+      "loss": 0.5875,
+      "step": 954
+    },
+    {
+      "epoch": 0.1845232344700995,
+      "grad_norm": 0.6258249878883362,
+      "learning_rate": 9.785374167593571e-06,
+      "loss": 0.6158,
+      "step": 955
+    },
+    {
+      "epoch": 0.18471645251666505,
+      "grad_norm": 0.778590977191925,
+      "learning_rate": 9.784395661789088e-06,
+      "loss": 0.662,
+      "step": 956
+    },
+    {
+      "epoch": 0.1849096705632306,
+      "grad_norm": 0.6642240285873413,
+      "learning_rate": 9.783414979639069e-06,
+      "loss": 0.5642,
+      "step": 957
+    },
+    {
+      "epoch": 0.18510288860979615,
+      "grad_norm": 0.7142787575721741,
+      "learning_rate": 9.782432121589613e-06,
+      "loss": 0.652,
+      "step": 958
+    },
+    {
+      "epoch": 0.1852961066563617,
+      "grad_norm": 0.7112715244293213,
+      "learning_rate": 9.781447088087809e-06,
+      "loss": 0.6482,
+      "step": 959
+    },
+    {
+      "epoch": 0.18548932470292726,
+      "grad_norm": 0.7406582236289978,
+      "learning_rate": 9.78045987958173e-06,
+      "loss": 0.6258,
+      "step": 960
+    },
+    {
+      "epoch": 0.1856825427494928,
+      "grad_norm": 0.6886512041091919,
+      "learning_rate": 9.779470496520442e-06,
+      "loss": 0.6783,
+      "step": 961
+    },
+    {
+      "epoch": 0.18587576079605836,
+      "grad_norm": 0.7022621035575867,
+      "learning_rate": 9.778478939353998e-06,
+      "loss": 0.6004,
+      "step": 962
+    },
+    {
+      "epoch": 0.1860689788426239,
+      "grad_norm": 0.9996697306632996,
+      "learning_rate": 9.777485208533447e-06,
+      "loss": 0.6278,
+      "step": 963
+    },
+    {
+      "epoch": 0.18626219688918946,
+      "grad_norm": 0.6829773783683777,
+      "learning_rate": 9.776489304510816e-06,
+      "loss": 0.6281,
+      "step": 964
+    },
+    {
+      "epoch": 0.186455414935755,
+      "grad_norm": 0.6927675604820251,
+      "learning_rate": 9.77549122773913e-06,
+      "loss": 0.657,
+      "step": 965
+    },
+    {
+      "epoch": 0.18664863298232054,
+      "grad_norm": 0.6441894173622131,
+      "learning_rate": 9.774490978672393e-06,
+      "loss": 0.6567,
+      "step": 966
+    },
+    {
+      "epoch": 0.1868418510288861,
+      "grad_norm": 0.7478853464126587,
+      "learning_rate": 9.773488557765606e-06,
+      "loss": 0.6507,
+      "step": 967
+    },
+    {
+      "epoch": 0.18703506907545164,
+      "grad_norm": 0.7151575088500977,
+      "learning_rate": 9.772483965474755e-06,
+      "loss": 0.6596,
+      "step": 968
+    },
+    {
+      "epoch": 0.1872282871220172,
+      "grad_norm": 0.7514240145683289,
+      "learning_rate": 9.771477202256812e-06,
+      "loss": 0.6582,
+      "step": 969
+    },
+    {
+      "epoch": 0.18742150516858275,
+      "grad_norm": 0.6776999235153198,
+      "learning_rate": 9.770468268569738e-06,
+      "loss": 0.6506,
+      "step": 970
+    },
+    {
+      "epoch": 0.1876147232151483,
+      "grad_norm": 0.6822491884231567,
+      "learning_rate": 9.769457164872477e-06,
+      "loss": 0.6432,
+      "step": 971
+    },
+    {
+      "epoch": 0.18780794126171385,
+      "grad_norm": 0.7168062329292297,
+      "learning_rate": 9.768443891624972e-06,
+      "loss": 0.5795,
+      "step": 972
+    },
+    {
+      "epoch": 0.1880011593082794,
+      "grad_norm": 0.7101817727088928,
+      "learning_rate": 9.767428449288138e-06,
+      "loss": 0.5788,
+      "step": 973
+    },
+    {
+      "epoch": 0.18819437735484495,
+      "grad_norm": 0.7040407657623291,
+      "learning_rate": 9.766410838323886e-06,
+      "loss": 0.6487,
+      "step": 974
+    },
+    {
+      "epoch": 0.18838759540141048,
+      "grad_norm": 0.7166803479194641,
+      "learning_rate": 9.765391059195111e-06,
+      "loss": 0.6394,
+      "step": 975
+    },
+    {
+      "epoch": 0.18858081344797603,
+      "grad_norm": 0.7549383044242859,
+      "learning_rate": 9.764369112365693e-06,
+      "loss": 0.6464,
+      "step": 976
+    },
+    {
+      "epoch": 0.18877403149454158,
+      "grad_norm": 0.7393662333488464,
+      "learning_rate": 9.763344998300501e-06,
+      "loss": 0.6588,
+      "step": 977
+    },
+    {
+      "epoch": 0.18896724954110714,
+      "grad_norm": 0.6731768250465393,
+      "learning_rate": 9.76231871746539e-06,
+      "loss": 0.6643,
+      "step": 978
+    },
+    {
+      "epoch": 0.1891604675876727,
+      "grad_norm": 0.6625547409057617,
+      "learning_rate": 9.761290270327192e-06,
+      "loss": 0.6115,
+      "step": 979
+    },
+    {
+      "epoch": 0.18935368563423824,
+      "grad_norm": 0.7536241412162781,
+      "learning_rate": 9.76025965735374e-06,
+      "loss": 0.6371,
+      "step": 980
+    },
+    {
+      "epoch": 0.1895469036808038,
+      "grad_norm": 0.7777042388916016,
+      "learning_rate": 9.759226879013839e-06,
+      "loss": 0.6834,
+      "step": 981
+    },
+    {
+      "epoch": 0.18974012172736934,
+      "grad_norm": 0.6960698366165161,
+      "learning_rate": 9.758191935777283e-06,
+      "loss": 0.6589,
+      "step": 982
+    },
+    {
+      "epoch": 0.1899333397739349,
+      "grad_norm": 0.7119461297988892,
+      "learning_rate": 9.757154828114852e-06,
+      "loss": 0.6247,
+      "step": 983
+    },
+    {
+      "epoch": 0.19012655782050045,
+      "grad_norm": 0.6847522854804993,
+      "learning_rate": 9.75611555649831e-06,
+      "loss": 0.5692,
+      "step": 984
+    },
+    {
+      "epoch": 0.19031977586706597,
+      "grad_norm": 0.7183122038841248,
+      "learning_rate": 9.755074121400402e-06,
+      "loss": 0.5878,
+      "step": 985
+    },
+    {
+      "epoch": 0.19051299391363152,
+      "grad_norm": 0.7369206547737122,
+      "learning_rate": 9.754030523294867e-06,
+      "loss": 0.5936,
+      "step": 986
+    },
+    {
+      "epoch": 0.19070621196019708,
+      "grad_norm": 0.7457402944564819,
+      "learning_rate": 9.752984762656417e-06,
+      "loss": 0.5817,
+      "step": 987
+    },
+    {
+      "epoch": 0.19089943000676263,
+      "grad_norm": 0.7065507173538208,
+      "learning_rate": 9.751936839960751e-06,
+      "loss": 0.621,
+      "step": 988
+    },
+    {
+      "epoch": 0.19109264805332818,
+      "grad_norm": 0.6595908999443054,
+      "learning_rate": 9.750886755684554e-06,
+      "loss": 0.6516,
+      "step": 989
+    },
+    {
+      "epoch": 0.19128586609989373,
+      "grad_norm": 0.6642166376113892,
+      "learning_rate": 9.749834510305493e-06,
+      "loss": 0.6323,
+      "step": 990
+    },
+    {
+      "epoch": 0.19147908414645928,
+      "grad_norm": 0.6469762921333313,
+      "learning_rate": 9.748780104302215e-06,
+      "loss": 0.5777,
+      "step": 991
+    },
+    {
+      "epoch": 0.19167230219302483,
+      "grad_norm": 0.7565948963165283,
+      "learning_rate": 9.747723538154356e-06,
+      "loss": 0.654,
+      "step": 992
+    },
+    {
+      "epoch": 0.1918655202395904,
+      "grad_norm": 0.6737761497497559,
+      "learning_rate": 9.746664812342531e-06,
+      "loss": 0.638,
+      "step": 993
+    },
+    {
+      "epoch": 0.19205873828615594,
+      "grad_norm": 0.8291156888008118,
+      "learning_rate": 9.745603927348335e-06,
+      "loss": 0.6959,
+      "step": 994
+    },
+    {
+      "epoch": 0.19225195633272146,
+      "grad_norm": 0.7155655026435852,
+      "learning_rate": 9.744540883654348e-06,
+      "loss": 0.6434,
+      "step": 995
+    },
+    {
+      "epoch": 0.19244517437928702,
+      "grad_norm": 0.6846557259559631,
+      "learning_rate": 9.743475681744135e-06,
+      "loss": 0.5974,
+      "step": 996
+    },
+    {
+      "epoch": 0.19263839242585257,
+      "grad_norm": 0.755359947681427,
+      "learning_rate": 9.742408322102234e-06,
+      "loss": 0.6992,
+      "step": 997
+    },
+    {
+      "epoch": 0.19283161047241812,
+      "grad_norm": 0.7169364094734192,
+      "learning_rate": 9.741338805214175e-06,
+      "loss": 0.695,
+      "step": 998
+    },
+    {
+      "epoch": 0.19302482851898367,
+      "grad_norm": 0.7717985510826111,
+      "learning_rate": 9.740267131566462e-06,
+      "loss": 0.6643,
+      "step": 999
+    },
+    {
+      "epoch": 0.19321804656554922,
+      "grad_norm": 0.7222839593887329,
+      "learning_rate": 9.739193301646582e-06,
+      "loss": 0.6541,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19341126461211477,
+      "grad_norm": 0.6785653233528137,
+      "learning_rate": 9.738117315943004e-06,
+      "loss": 0.6275,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19360448265868033,
+      "grad_norm": 0.7301161289215088,
+      "learning_rate": 9.737039174945175e-06,
+      "loss": 0.6242,
+      "step": 1002
+    },
+    {
+      "epoch": 0.19379770070524588,
+      "grad_norm": 0.6698400378227234,
+      "learning_rate": 9.735958879143527e-06,
+      "loss": 0.6242,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19399091875181143,
+      "grad_norm": 0.6522667407989502,
+      "learning_rate": 9.734876429029469e-06,
+      "loss": 0.6233,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19418413679837696,
+      "grad_norm": 0.7018197178840637,
+      "learning_rate": 9.733791825095389e-06,
+      "loss": 0.6226,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1943773548449425,
+      "grad_norm": 0.6871991753578186,
+      "learning_rate": 9.732705067834655e-06,
+      "loss": 0.5791,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19457057289150806,
+      "grad_norm": 0.6892812252044678,
+      "learning_rate": 9.73161615774162e-06,
+      "loss": 0.651,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1947637909380736,
+      "grad_norm": 0.7220892906188965,
+      "learning_rate": 9.730525095311604e-06,
+      "loss": 0.6651,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19495700898463916,
+      "grad_norm": 0.8108465671539307,
+      "learning_rate": 9.729431881040922e-06,
+      "loss": 0.6489,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19515022703120471,
+      "grad_norm": 0.6641649603843689,
+      "learning_rate": 9.728336515426856e-06,
+      "loss": 0.6818,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19534344507777027,
+      "grad_norm": 0.7121703624725342,
+      "learning_rate": 9.727238998967674e-06,
+      "loss": 0.6847,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19553666312433582,
+      "grad_norm": 0.7284754514694214,
+      "learning_rate": 9.726139332162613e-06,
+      "loss": 0.6805,
+      "step": 1012
+    },
+    {
+      "epoch": 0.19572988117090137,
+      "grad_norm": 0.6468107104301453,
+      "learning_rate": 9.725037515511897e-06,
+      "loss": 0.6076,
+      "step": 1013
+    },
+    {
+      "epoch": 0.19592309921746692,
+      "grad_norm": 0.669460654258728,
+      "learning_rate": 9.723933549516726e-06,
+      "loss": 0.5843,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19611631726403245,
+      "grad_norm": 0.6437966227531433,
+      "learning_rate": 9.722827434679275e-06,
+      "loss": 0.6328,
+      "step": 1015
+    },
+    {
+      "epoch": 0.196309535310598,
+      "grad_norm": 0.7481193542480469,
+      "learning_rate": 9.721719171502699e-06,
+      "loss": 0.6264,
+      "step": 1016
+    },
+    {
+      "epoch": 0.19650275335716355,
+      "grad_norm": 0.7158156633377075,
+      "learning_rate": 9.720608760491127e-06,
+      "loss": 0.6302,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1966959714037291,
+      "grad_norm": 0.6981252431869507,
+      "learning_rate": 9.71949620214967e-06,
+      "loss": 0.6512,
+      "step": 1018
+    },
+    {
+      "epoch": 0.19688918945029465,
+      "grad_norm": 0.667368471622467,
+      "learning_rate": 9.718381496984413e-06,
+      "loss": 0.6169,
+      "step": 1019
+    },
+    {
+      "epoch": 0.1970824074968602,
+      "grad_norm": 0.6622874736785889,
+      "learning_rate": 9.717264645502414e-06,
+      "loss": 0.6511,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19727562554342576,
+      "grad_norm": 0.6562559008598328,
+      "learning_rate": 9.716145648211717e-06,
+      "loss": 0.5724,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1974688435899913,
+      "grad_norm": 0.7347162365913391,
+      "learning_rate": 9.71502450562133e-06,
+      "loss": 0.6414,
+      "step": 1022
+    },
+    {
+      "epoch": 0.19766206163655686,
+      "grad_norm": 0.7133053541183472,
+      "learning_rate": 9.713901218241247e-06,
+      "loss": 0.6633,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19785527968312241,
+      "grad_norm": 0.6382784247398376,
+      "learning_rate": 9.712775786582433e-06,
+      "loss": 0.6002,
+      "step": 1024
+    },
+    {
+      "epoch": 0.19804849772968797,
+      "grad_norm": 0.6543134450912476,
+      "learning_rate": 9.711648211156828e-06,
+      "loss": 0.596,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1982417157762535,
+      "grad_norm": 0.7354627251625061,
+      "learning_rate": 9.710518492477348e-06,
+      "loss": 0.6126,
+      "step": 1026
+    },
+    {
+      "epoch": 0.19843493382281904,
+      "grad_norm": 0.6763684749603271,
+      "learning_rate": 9.709386631057885e-06,
+      "loss": 0.6316,
+      "step": 1027
+    },
+    {
+      "epoch": 0.1986281518693846,
+      "grad_norm": 0.7201732993125916,
+      "learning_rate": 9.708252627413303e-06,
+      "loss": 0.6059,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19882136991595015,
+      "grad_norm": 0.6838993430137634,
+      "learning_rate": 9.707116482059447e-06,
+      "loss": 0.6586,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1990145879625157,
+      "grad_norm": 0.70381098985672,
+      "learning_rate": 9.705978195513125e-06,
+      "loss": 0.7066,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19920780600908125,
+      "grad_norm": 0.6609904170036316,
+      "learning_rate": 9.70483776829213e-06,
+      "loss": 0.6542,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1994010240556468,
+      "grad_norm": 0.7389576435089111,
+      "learning_rate": 9.70369520091522e-06,
+      "loss": 0.5947,
+      "step": 1032
+    },
+    {
+      "epoch": 0.19959424210221235,
+      "grad_norm": 0.663948655128479,
+      "learning_rate": 9.702550493902136e-06,
+      "loss": 0.7133,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1997874601487779,
+      "grad_norm": 0.6392188668251038,
+      "learning_rate": 9.701403647773584e-06,
+      "loss": 0.5387,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19998067819534346,
+      "grad_norm": 0.7064595222473145,
+      "learning_rate": 9.700254663051245e-06,
+      "loss": 0.6729,
+      "step": 1035
+    },
+    {
+      "epoch": 0.20017389624190898,
+      "grad_norm": 0.725175678730011,
+      "learning_rate": 9.699103540257776e-06,
+      "loss": 0.6487,
+      "step": 1036
+    },
+    {
+      "epoch": 0.20036711428847453,
+      "grad_norm": 0.7198014259338379,
+      "learning_rate": 9.697950279916805e-06,
+      "loss": 0.6929,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2005603323350401,
+      "grad_norm": 0.768462061882019,
+      "learning_rate": 9.696794882552932e-06,
+      "loss": 0.6333,
+      "step": 1038
+    },
+    {
+      "epoch": 0.20075355038160564,
+      "grad_norm": 0.6988205909729004,
+      "learning_rate": 9.695637348691725e-06,
+      "loss": 0.6548,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2009467684281712,
+      "grad_norm": 0.6590883135795593,
+      "learning_rate": 9.69447767885973e-06,
+      "loss": 0.6292,
+      "step": 1040
+    },
+    {
+      "epoch": 0.20113998647473674,
+      "grad_norm": 0.6760185956954956,
+      "learning_rate": 9.693315873584464e-06,
+      "loss": 0.6513,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2013332045213023,
+      "grad_norm": 0.6331936120986938,
+      "learning_rate": 9.692151933394412e-06,
+      "loss": 0.5782,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20152642256786785,
+      "grad_norm": 0.7679101228713989,
+      "learning_rate": 9.690985858819033e-06,
+      "loss": 0.6544,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2017196406144334,
+      "grad_norm": 0.6922922730445862,
+      "learning_rate": 9.689817650388754e-06,
+      "loss": 0.621,
+      "step": 1044
+    },
+    {
+      "epoch": 0.20191285866099895,
+      "grad_norm": 0.6999267935752869,
+      "learning_rate": 9.688647308634977e-06,
+      "loss": 0.5854,
+      "step": 1045
+    },
+    {
+      "epoch": 0.20210607670756447,
+      "grad_norm": 0.6538942456245422,
+      "learning_rate": 9.68747483409007e-06,
+      "loss": 0.5802,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20229929475413003,
+      "grad_norm": 0.7141908407211304,
+      "learning_rate": 9.68630022728737e-06,
+      "loss": 0.5738,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20249251280069558,
+      "grad_norm": 0.7174897789955139,
+      "learning_rate": 9.685123488761195e-06,
+      "loss": 0.6798,
+      "step": 1048
+    },
+    {
+      "epoch": 0.20268573084726113,
+      "grad_norm": 0.6853818297386169,
+      "learning_rate": 9.683944619046817e-06,
+      "loss": 0.6249,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20287894889382668,
+      "grad_norm": 0.6704294085502625,
+      "learning_rate": 9.68276361868049e-06,
+      "loss": 0.5628,
+      "step": 1050
+    },
+    {
+      "epoch": 0.20307216694039223,
+      "grad_norm": 0.6642906069755554,
+      "learning_rate": 9.68158048819943e-06,
+      "loss": 0.6616,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2032653849869578,
+      "grad_norm": 0.7046515941619873,
+      "learning_rate": 9.680395228141825e-06,
+      "loss": 0.6105,
+      "step": 1052
+    },
+    {
+      "epoch": 0.20345860303352334,
+      "grad_norm": 0.6480652689933777,
+      "learning_rate": 9.67920783904683e-06,
+      "loss": 0.7255,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2036518210800889,
+      "grad_norm": 0.6735095381736755,
+      "learning_rate": 9.67801832145457e-06,
+      "loss": 0.6769,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20384503912665444,
+      "grad_norm": 0.7383208870887756,
+      "learning_rate": 9.676826675906139e-06,
+      "loss": 0.6605,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20403825717321997,
+      "grad_norm": 0.6329072117805481,
+      "learning_rate": 9.675632902943595e-06,
+      "loss": 0.581,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20423147521978552,
+      "grad_norm": 0.6839841604232788,
+      "learning_rate": 9.674437003109967e-06,
+      "loss": 0.6111,
+      "step": 1057
+    },
+    {
+      "epoch": 0.20442469326635107,
+      "grad_norm": 0.6835163235664368,
+      "learning_rate": 9.673238976949253e-06,
+      "loss": 0.6533,
+      "step": 1058
+    },
+    {
+      "epoch": 0.20461791131291662,
+      "grad_norm": 0.749506413936615,
+      "learning_rate": 9.672038825006415e-06,
+      "loss": 0.6601,
+      "step": 1059
+    },
+    {
+      "epoch": 0.20481112935948217,
+      "grad_norm": 0.6942434310913086,
+      "learning_rate": 9.67083654782738e-06,
+      "loss": 0.6767,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20500434740604773,
+      "grad_norm": 0.735859215259552,
+      "learning_rate": 9.669632145959051e-06,
+      "loss": 0.7378,
+      "step": 1061
+    },
+    {
+      "epoch": 0.20519756545261328,
+      "grad_norm": 0.7022744417190552,
+      "learning_rate": 9.668425619949285e-06,
+      "loss": 0.5824,
+      "step": 1062
+    },
+    {
+      "epoch": 0.20539078349917883,
+      "grad_norm": 0.6515902876853943,
+      "learning_rate": 9.667216970346916e-06,
+      "loss": 0.5956,
+      "step": 1063
+    },
+    {
+      "epoch": 0.20558400154574438,
+      "grad_norm": 0.6899116039276123,
+      "learning_rate": 9.666006197701735e-06,
+      "loss": 0.6656,
+      "step": 1064
+    },
+    {
+      "epoch": 0.20577721959230993,
+      "grad_norm": 0.732843279838562,
+      "learning_rate": 9.664793302564507e-06,
+      "loss": 0.6582,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20597043763887546,
+      "grad_norm": 0.6981117129325867,
+      "learning_rate": 9.66357828548696e-06,
+      "loss": 0.6305,
+      "step": 1066
+    },
+    {
+      "epoch": 0.206163655685441,
+      "grad_norm": 0.6855542659759521,
+      "learning_rate": 9.66236114702178e-06,
+      "loss": 0.6082,
+      "step": 1067
+    },
+    {
+      "epoch": 0.20635687373200656,
+      "grad_norm": 0.6619362235069275,
+      "learning_rate": 9.66114188772263e-06,
+      "loss": 0.5797,
+      "step": 1068
+    },
+    {
+      "epoch": 0.20655009177857211,
+      "grad_norm": 0.7190464735031128,
+      "learning_rate": 9.659920508144127e-06,
+      "loss": 0.5616,
+      "step": 1069
+    },
+    {
+      "epoch": 0.20674330982513767,
+      "grad_norm": 0.7022643089294434,
+      "learning_rate": 9.65869700884186e-06,
+      "loss": 0.6047,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20693652787170322,
+      "grad_norm": 0.6176483631134033,
+      "learning_rate": 9.657471390372379e-06,
+      "loss": 0.6004,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20712974591826877,
+      "grad_norm": 0.8088053464889526,
+      "learning_rate": 9.656243653293199e-06,
+      "loss": 0.6969,
+      "step": 1072
+    },
+    {
+      "epoch": 0.20732296396483432,
+      "grad_norm": 0.6982502937316895,
+      "learning_rate": 9.655013798162796e-06,
+      "loss": 0.6553,
+      "step": 1073
+    },
+    {
+      "epoch": 0.20751618201139987,
+      "grad_norm": 0.6992550492286682,
+      "learning_rate": 9.653781825540613e-06,
+      "loss": 0.682,
+      "step": 1074
+    },
+    {
+      "epoch": 0.20770940005796543,
+      "grad_norm": 0.6730358004570007,
+      "learning_rate": 9.652547735987055e-06,
+      "loss": 0.6761,
+      "step": 1075
+    },
+    {
+      "epoch": 0.20790261810453095,
+      "grad_norm": 0.7835087180137634,
+      "learning_rate": 9.651311530063488e-06,
+      "loss": 0.7146,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2080958361510965,
+      "grad_norm": 0.6247279047966003,
+      "learning_rate": 9.650073208332245e-06,
+      "loss": 0.5277,
+      "step": 1077
+    },
+    {
+      "epoch": 0.20828905419766205,
+      "grad_norm": 0.6868420243263245,
+      "learning_rate": 9.648832771356616e-06,
+      "loss": 0.7027,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2084822722442276,
+      "grad_norm": 0.7575464844703674,
+      "learning_rate": 9.647590219700858e-06,
+      "loss": 0.6745,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20867549029079316,
+      "grad_norm": 0.6909801363945007,
+      "learning_rate": 9.646345553930187e-06,
+      "loss": 0.6869,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2088687083373587,
+      "grad_norm": 0.7814114689826965,
+      "learning_rate": 9.645098774610783e-06,
+      "loss": 0.6791,
+      "step": 1081
+    },
+    {
+      "epoch": 0.20906192638392426,
+      "grad_norm": 0.7519343495368958,
+      "learning_rate": 9.643849882309786e-06,
+      "loss": 0.6919,
+      "step": 1082
+    },
+    {
+      "epoch": 0.20925514443048981,
+      "grad_norm": 0.6684097051620483,
+      "learning_rate": 9.642598877595295e-06,
+      "loss": 0.6295,
+      "step": 1083
+    },
+    {
+      "epoch": 0.20944836247705537,
+      "grad_norm": 0.6317864060401917,
+      "learning_rate": 9.641345761036374e-06,
+      "loss": 0.669,
+      "step": 1084
+    },
+    {
+      "epoch": 0.20964158052362092,
+      "grad_norm": 0.7804228663444519,
+      "learning_rate": 9.640090533203046e-06,
+      "loss": 0.6534,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20983479857018644,
+      "grad_norm": 0.6699924468994141,
+      "learning_rate": 9.638833194666293e-06,
+      "loss": 0.6345,
+      "step": 1086
+    },
+    {
+      "epoch": 0.210028016616752,
+      "grad_norm": 0.7921578884124756,
+      "learning_rate": 9.63757374599806e-06,
+      "loss": 0.5995,
+      "step": 1087
+    },
+    {
+      "epoch": 0.21022123466331755,
+      "grad_norm": 0.8038668632507324,
+      "learning_rate": 9.636312187771249e-06,
+      "loss": 0.6437,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2104144527098831,
+      "grad_norm": 0.7267628312110901,
+      "learning_rate": 9.635048520559722e-06,
+      "loss": 0.6527,
+      "step": 1089
+    },
+    {
+      "epoch": 0.21060767075644865,
+      "grad_norm": 0.6915916800498962,
+      "learning_rate": 9.633782744938303e-06,
+      "loss": 0.6753,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2108008888030142,
+      "grad_norm": 0.7378334403038025,
+      "learning_rate": 9.632514861482772e-06,
+      "loss": 0.6689,
+      "step": 1091
+    },
+    {
+      "epoch": 0.21099410684957975,
+      "grad_norm": 0.6627857685089111,
+      "learning_rate": 9.631244870769869e-06,
+      "loss": 0.6408,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2111873248961453,
+      "grad_norm": 0.6825361251831055,
+      "learning_rate": 9.629972773377294e-06,
+      "loss": 0.5799,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21138054294271086,
+      "grad_norm": 0.6906008124351501,
+      "learning_rate": 9.628698569883702e-06,
+      "loss": 0.5705,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2115737609892764,
+      "grad_norm": 0.6810574531555176,
+      "learning_rate": 9.627422260868708e-06,
+      "loss": 0.5739,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21176697903584193,
+      "grad_norm": 0.6908286809921265,
+      "learning_rate": 9.626143846912887e-06,
+      "loss": 0.6641,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2119601970824075,
+      "grad_norm": 0.7369417548179626,
+      "learning_rate": 9.624863328597767e-06,
+      "loss": 0.6698,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21215341512897304,
+      "grad_norm": 0.727998673915863,
+      "learning_rate": 9.623580706505835e-06,
+      "loss": 0.6097,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2123466331755386,
+      "grad_norm": 0.7268903255462646,
+      "learning_rate": 9.622295981220538e-06,
+      "loss": 0.6582,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21253985122210414,
+      "grad_norm": 0.723845899105072,
+      "learning_rate": 9.621009153326275e-06,
+      "loss": 0.6299,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2127330692686697,
+      "grad_norm": 0.6438921689987183,
+      "learning_rate": 9.619720223408404e-06,
+      "loss": 0.5977,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21292628731523525,
+      "grad_norm": 0.7213731408119202,
+      "learning_rate": 9.618429192053239e-06,
+      "loss": 0.679,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2131195053618008,
+      "grad_norm": 0.7237257361412048,
+      "learning_rate": 9.61713605984805e-06,
+      "loss": 0.6114,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21331272340836635,
+      "grad_norm": 0.6565388441085815,
+      "learning_rate": 9.61584082738106e-06,
+      "loss": 0.6188,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2135059414549319,
+      "grad_norm": 0.7472566366195679,
+      "learning_rate": 9.61454349524145e-06,
+      "loss": 0.6357,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21369915950149743,
+      "grad_norm": 0.7013627886772156,
+      "learning_rate": 9.61324406401936e-06,
+      "loss": 0.624,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21389237754806298,
+      "grad_norm": 0.7195550203323364,
+      "learning_rate": 9.611942534305878e-06,
+      "loss": 0.6799,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21408559559462853,
+      "grad_norm": 0.6914951205253601,
+      "learning_rate": 9.610638906693049e-06,
+      "loss": 0.5486,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21427881364119408,
+      "grad_norm": 0.7639178037643433,
+      "learning_rate": 9.60933318177387e-06,
+      "loss": 0.6419,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21447203168775963,
+      "grad_norm": 0.6939564347267151,
+      "learning_rate": 9.608025360142299e-06,
+      "loss": 0.6106,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2146652497343252,
+      "grad_norm": 0.7501881122589111,
+      "learning_rate": 9.60671544239324e-06,
+      "loss": 0.6739,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21485846778089074,
+      "grad_norm": 0.684212327003479,
+      "learning_rate": 9.605403429122556e-06,
+      "loss": 0.6412,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2150516858274563,
+      "grad_norm": 0.6695783734321594,
+      "learning_rate": 9.604089320927058e-06,
+      "loss": 0.6163,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21524490387402184,
+      "grad_norm": 0.6716512441635132,
+      "learning_rate": 9.602773118404518e-06,
+      "loss": 0.6689,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2154381219205874,
+      "grad_norm": 0.6913531422615051,
+      "learning_rate": 9.60145482215365e-06,
+      "loss": 0.6212,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21563133996715292,
+      "grad_norm": 0.7106210589408875,
+      "learning_rate": 9.600134432774132e-06,
+      "loss": 0.6159,
+      "step": 1116
+    },
+    {
+      "epoch": 0.21582455801371847,
+      "grad_norm": 0.7008172273635864,
+      "learning_rate": 9.59881195086658e-06,
+      "loss": 0.6602,
+      "step": 1117
+    },
+    {
+      "epoch": 0.21601777606028402,
+      "grad_norm": 0.7599744200706482,
+      "learning_rate": 9.597487377032576e-06,
+      "loss": 0.6967,
+      "step": 1118
+    },
+    {
+      "epoch": 0.21621099410684957,
+      "grad_norm": 0.8056255578994751,
+      "learning_rate": 9.596160711874646e-06,
+      "loss": 0.713,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21640421215341513,
+      "grad_norm": 0.7190768122673035,
+      "learning_rate": 9.594831955996272e-06,
+      "loss": 0.6591,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21659743019998068,
+      "grad_norm": 0.7100278735160828,
+      "learning_rate": 9.59350111000188e-06,
+      "loss": 0.6666,
+      "step": 1121
+    },
+    {
+      "epoch": 0.21679064824654623,
+      "grad_norm": 0.7236603498458862,
+      "learning_rate": 9.59216817449685e-06,
+      "loss": 0.629,
+      "step": 1122
+    },
+    {
+      "epoch": 0.21698386629311178,
+      "grad_norm": 0.631443977355957,
+      "learning_rate": 9.590833150087515e-06,
+      "loss": 0.589,
+      "step": 1123
+    },
+    {
+      "epoch": 0.21717708433967733,
+      "grad_norm": 0.8489832878112793,
+      "learning_rate": 9.589496037381156e-06,
+      "loss": 0.5739,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2173703023862429,
+      "grad_norm": 0.6647274494171143,
+      "learning_rate": 9.588156836986007e-06,
+      "loss": 0.6276,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2175635204328084,
+      "grad_norm": 0.629458487033844,
+      "learning_rate": 9.586815549511244e-06,
+      "loss": 0.5635,
+      "step": 1126
+    },
+    {
+      "epoch": 0.21775673847937396,
+      "grad_norm": 0.7327702045440674,
+      "learning_rate": 9.585472175567e-06,
+      "loss": 0.6328,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21794995652593951,
+      "grad_norm": 0.6244614124298096,
+      "learning_rate": 9.584126715764356e-06,
+      "loss": 0.6142,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21814317457250507,
+      "grad_norm": 0.6652092933654785,
+      "learning_rate": 9.582779170715337e-06,
+      "loss": 0.6248,
+      "step": 1129
+    },
+    {
+      "epoch": 0.21833639261907062,
+      "grad_norm": 0.7154573202133179,
+      "learning_rate": 9.581429541032922e-06,
+      "loss": 0.7063,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21852961066563617,
+      "grad_norm": 0.6636493802070618,
+      "learning_rate": 9.580077827331038e-06,
+      "loss": 0.5944,
+      "step": 1131
+    },
+    {
+      "epoch": 0.21872282871220172,
+      "grad_norm": 0.7206150889396667,
+      "learning_rate": 9.578724030224554e-06,
+      "loss": 0.6683,
+      "step": 1132
+    },
+    {
+      "epoch": 0.21891604675876727,
+      "grad_norm": 0.727927029132843,
+      "learning_rate": 9.57736815032929e-06,
+      "loss": 0.6216,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21910926480533283,
+      "grad_norm": 0.7566357254981995,
+      "learning_rate": 9.576010188262019e-06,
+      "loss": 0.5867,
+      "step": 1134
+    },
+    {
+      "epoch": 0.21930248285189838,
+      "grad_norm": 0.7100594639778137,
+      "learning_rate": 9.574650144640454e-06,
+      "loss": 0.626,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21949570089846393,
+      "grad_norm": 0.7419232726097107,
+      "learning_rate": 9.573288020083255e-06,
+      "loss": 0.6392,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21968891894502945,
+      "grad_norm": 0.6545636057853699,
+      "learning_rate": 9.571923815210037e-06,
+      "loss": 0.6642,
+      "step": 1137
+    },
+    {
+      "epoch": 0.219882136991595,
+      "grad_norm": 0.6935176849365234,
+      "learning_rate": 9.570557530641347e-06,
+      "loss": 0.6333,
+      "step": 1138
+    },
+    {
+      "epoch": 0.22007535503816056,
+      "grad_norm": 0.688477635383606,
+      "learning_rate": 9.569189166998693e-06,
+      "loss": 0.5613,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2202685730847261,
+      "grad_norm": 0.759567379951477,
+      "learning_rate": 9.56781872490452e-06,
+      "loss": 0.6683,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22046179113129166,
+      "grad_norm": 0.6750019192695618,
+      "learning_rate": 9.566446204982217e-06,
+      "loss": 0.7004,
+      "step": 1141
+    },
+    {
+      "epoch": 0.22065500917785721,
+      "grad_norm": 0.6627630591392517,
+      "learning_rate": 9.565071607856123e-06,
+      "loss": 0.6228,
+      "step": 1142
+    },
+    {
+      "epoch": 0.22084822722442277,
+      "grad_norm": 0.6812954545021057,
+      "learning_rate": 9.563694934151524e-06,
+      "loss": 0.5958,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22104144527098832,
+      "grad_norm": 0.6947746276855469,
+      "learning_rate": 9.562316184494642e-06,
+      "loss": 0.5884,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22123466331755387,
+      "grad_norm": 0.7650540471076965,
+      "learning_rate": 9.560935359512653e-06,
+      "loss": 0.6613,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22142788136411942,
+      "grad_norm": 0.7075700759887695,
+      "learning_rate": 9.559552459833668e-06,
+      "loss": 0.6608,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22162109941068495,
+      "grad_norm": 0.7310034036636353,
+      "learning_rate": 9.558167486086749e-06,
+      "loss": 0.6725,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2218143174572505,
+      "grad_norm": 0.989868700504303,
+      "learning_rate": 9.556780438901899e-06,
+      "loss": 0.6763,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22200753550381605,
+      "grad_norm": 0.6869245767593384,
+      "learning_rate": 9.555391318910058e-06,
+      "loss": 0.5556,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2222007535503816,
+      "grad_norm": 0.6328395009040833,
+      "learning_rate": 9.554000126743123e-06,
+      "loss": 0.6198,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22239397159694715,
+      "grad_norm": 0.6847072243690491,
+      "learning_rate": 9.55260686303392e-06,
+      "loss": 0.6067,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2225871896435127,
+      "grad_norm": 0.6853692531585693,
+      "learning_rate": 9.551211528416221e-06,
+      "loss": 0.6773,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22278040769007826,
+      "grad_norm": 0.6771348714828491,
+      "learning_rate": 9.549814123524747e-06,
+      "loss": 0.6519,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2229736257366438,
+      "grad_norm": 0.7524189352989197,
+      "learning_rate": 9.548414648995152e-06,
+      "loss": 0.658,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22316684378320936,
+      "grad_norm": 0.7184376120567322,
+      "learning_rate": 9.547013105464036e-06,
+      "loss": 0.6449,
+      "step": 1155
+    },
+    {
+      "epoch": 0.22336006182977491,
+      "grad_norm": 0.7551977634429932,
+      "learning_rate": 9.545609493568938e-06,
+      "loss": 0.6681,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22355327987634044,
+      "grad_norm": 0.6515470743179321,
+      "learning_rate": 9.544203813948336e-06,
+      "loss": 0.6263,
+      "step": 1157
+    },
+    {
+      "epoch": 0.223746497922906,
+      "grad_norm": 0.7196053862571716,
+      "learning_rate": 9.542796067241655e-06,
+      "loss": 0.6634,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22393971596947154,
+      "grad_norm": 0.6914680004119873,
+      "learning_rate": 9.541386254089257e-06,
+      "loss": 0.6503,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2241329340160371,
+      "grad_norm": 0.6578753590583801,
+      "learning_rate": 9.539974375132441e-06,
+      "loss": 0.6302,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22432615206260265,
+      "grad_norm": 0.684166431427002,
+      "learning_rate": 9.53856043101345e-06,
+      "loss": 0.6332,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2245193701091682,
+      "grad_norm": 0.6369361877441406,
+      "learning_rate": 9.537144422375465e-06,
+      "loss": 0.558,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22471258815573375,
+      "grad_norm": 0.7176256775856018,
+      "learning_rate": 9.535726349862604e-06,
+      "loss": 0.6174,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2249058062022993,
+      "grad_norm": 0.7150354981422424,
+      "learning_rate": 9.534306214119928e-06,
+      "loss": 0.6509,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22509902424886485,
+      "grad_norm": 0.6560936570167542,
+      "learning_rate": 9.532884015793432e-06,
+      "loss": 0.6202,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2252922422954304,
+      "grad_norm": 0.6534649729728699,
+      "learning_rate": 9.531459755530053e-06,
+      "loss": 0.657,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22548546034199593,
+      "grad_norm": 0.7739822864532471,
+      "learning_rate": 9.530033433977661e-06,
+      "loss": 0.6452,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22567867838856148,
+      "grad_norm": 0.6942310333251953,
+      "learning_rate": 9.528605051785073e-06,
+      "loss": 0.6455,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22587189643512703,
+      "grad_norm": 0.6991093754768372,
+      "learning_rate": 9.527174609602033e-06,
+      "loss": 0.6723,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2260651144816926,
+      "grad_norm": 0.786305844783783,
+      "learning_rate": 9.525742108079226e-06,
+      "loss": 0.7261,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22625833252825814,
+      "grad_norm": 0.6831268668174744,
+      "learning_rate": 9.524307547868277e-06,
+      "loss": 0.62,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2264515505748237,
+      "grad_norm": 0.8073094487190247,
+      "learning_rate": 9.522870929621742e-06,
+      "loss": 0.65,
+      "step": 1172
+    },
+    {
+      "epoch": 0.22664476862138924,
+      "grad_norm": 0.6997658610343933,
+      "learning_rate": 9.52143225399312e-06,
+      "loss": 0.6764,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2268379866679548,
+      "grad_norm": 0.7320740818977356,
+      "learning_rate": 9.519991521636835e-06,
+      "loss": 0.5959,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22703120471452035,
+      "grad_norm": 0.6965344548225403,
+      "learning_rate": 9.518548733208258e-06,
+      "loss": 0.6865,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2272244227610859,
+      "grad_norm": 0.6489505171775818,
+      "learning_rate": 9.51710388936369e-06,
+      "loss": 0.6104,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22741764080765142,
+      "grad_norm": 0.6967784762382507,
+      "learning_rate": 9.515656990760366e-06,
+      "loss": 0.6686,
+      "step": 1177
+    },
+    {
+      "epoch": 0.22761085885421697,
+      "grad_norm": 0.6577012538909912,
+      "learning_rate": 9.51420803805646e-06,
+      "loss": 0.6377,
+      "step": 1178
+    },
+    {
+      "epoch": 0.22780407690078253,
+      "grad_norm": 0.7022258043289185,
+      "learning_rate": 9.512757031911076e-06,
+      "loss": 0.6086,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22799729494734808,
+      "grad_norm": 0.7283061742782593,
+      "learning_rate": 9.511303972984251e-06,
+      "loss": 0.6597,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22819051299391363,
+      "grad_norm": 0.759132444858551,
+      "learning_rate": 9.509848861936963e-06,
+      "loss": 0.6478,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22838373104047918,
+      "grad_norm": 0.7047739028930664,
+      "learning_rate": 9.508391699431114e-06,
+      "loss": 0.6123,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22857694908704473,
+      "grad_norm": 0.7257245182991028,
+      "learning_rate": 9.506932486129548e-06,
+      "loss": 0.6167,
+      "step": 1183
+    },
+    {
+      "epoch": 0.22877016713361029,
+      "grad_norm": 0.7427594661712646,
+      "learning_rate": 9.505471222696037e-06,
+      "loss": 0.6486,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22896338518017584,
+      "grad_norm": 0.6519560813903809,
+      "learning_rate": 9.504007909795287e-06,
+      "loss": 0.57,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2291566032267414,
+      "grad_norm": 0.6736737489700317,
+      "learning_rate": 9.502542548092934e-06,
+      "loss": 0.6162,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22934982127330691,
+      "grad_norm": 0.6636473536491394,
+      "learning_rate": 9.501075138255546e-06,
+      "loss": 0.6284,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22954303931987247,
+      "grad_norm": 0.699479877948761,
+      "learning_rate": 9.499605680950632e-06,
+      "loss": 0.6542,
+      "step": 1188
+    },
+    {
+      "epoch": 0.22973625736643802,
+      "grad_norm": 0.6886146664619446,
+      "learning_rate": 9.498134176846616e-06,
+      "loss": 0.6024,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22992947541300357,
+      "grad_norm": 0.6369981169700623,
+      "learning_rate": 9.496660626612867e-06,
+      "loss": 0.581,
+      "step": 1190
+    },
+    {
+      "epoch": 0.23012269345956912,
+      "grad_norm": 0.6791765093803406,
+      "learning_rate": 9.495185030919678e-06,
+      "loss": 0.636,
+      "step": 1191
+    },
+    {
+      "epoch": 0.23031591150613467,
+      "grad_norm": 0.6348846554756165,
+      "learning_rate": 9.493707390438275e-06,
+      "loss": 0.572,
+      "step": 1192
+    },
+    {
+      "epoch": 0.23050912955270023,
+      "grad_norm": 0.6705478429794312,
+      "learning_rate": 9.492227705840811e-06,
+      "loss": 0.585,
+      "step": 1193
+    },
+    {
+      "epoch": 0.23070234759926578,
+      "grad_norm": 0.6739753484725952,
+      "learning_rate": 9.490745977800372e-06,
+      "loss": 0.6509,
+      "step": 1194
+    },
+    {
+      "epoch": 0.23089556564583133,
+      "grad_norm": 0.6946831941604614,
+      "learning_rate": 9.489262206990976e-06,
+      "loss": 0.5886,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23108878369239688,
+      "grad_norm": 0.6643811464309692,
+      "learning_rate": 9.487776394087563e-06,
+      "loss": 0.6302,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2312820017389624,
+      "grad_norm": 0.7033334970474243,
+      "learning_rate": 9.486288539766004e-06,
+      "loss": 0.6788,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23147521978552796,
+      "grad_norm": 0.6602683067321777,
+      "learning_rate": 9.484798644703104e-06,
+      "loss": 0.5796,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2316684378320935,
+      "grad_norm": 0.7179946899414062,
+      "learning_rate": 9.48330670957659e-06,
+      "loss": 0.708,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23186165587865906,
+      "grad_norm": 0.6813563704490662,
+      "learning_rate": 9.481812735065123e-06,
+      "loss": 0.7012,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23205487392522461,
+      "grad_norm": 0.7580933570861816,
+      "learning_rate": 9.480316721848285e-06,
+      "loss": 0.7058,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23224809197179017,
+      "grad_norm": 0.7759296298027039,
+      "learning_rate": 9.478818670606588e-06,
+      "loss": 0.6901,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23244131001835572,
+      "grad_norm": 0.6933876872062683,
+      "learning_rate": 9.477318582021477e-06,
+      "loss": 0.6515,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23263452806492127,
+      "grad_norm": 0.7261629104614258,
+      "learning_rate": 9.475816456775313e-06,
+      "loss": 0.7087,
+      "step": 1204
+    },
+    {
+      "epoch": 0.23282774611148682,
+      "grad_norm": 0.7353148460388184,
+      "learning_rate": 9.47431229555139e-06,
+      "loss": 0.6936,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23302096415805237,
+      "grad_norm": 0.6757679581642151,
+      "learning_rate": 9.472806099033928e-06,
+      "loss": 0.663,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2332141822046179,
+      "grad_norm": 0.656043529510498,
+      "learning_rate": 9.471297867908073e-06,
+      "loss": 0.6339,
+      "step": 1207
+    },
+    {
+      "epoch": 0.23340740025118345,
+      "grad_norm": 0.7144488096237183,
+      "learning_rate": 9.469787602859893e-06,
+      "loss": 0.6515,
+      "step": 1208
+    },
+    {
+      "epoch": 0.233600618297749,
+      "grad_norm": 0.685064971446991,
+      "learning_rate": 9.468275304576384e-06,
+      "loss": 0.6148,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23379383634431455,
+      "grad_norm": 0.6732592582702637,
+      "learning_rate": 9.466760973745467e-06,
+      "loss": 0.6328,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2339870543908801,
+      "grad_norm": 0.6932583451271057,
+      "learning_rate": 9.465244611055988e-06,
+      "loss": 0.6359,
+      "step": 1211
+    },
+    {
+      "epoch": 0.23418027243744566,
+      "grad_norm": 0.7521321773529053,
+      "learning_rate": 9.463726217197715e-06,
+      "loss": 0.6166,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2343734904840112,
+      "grad_norm": 0.7030579447746277,
+      "learning_rate": 9.462205792861342e-06,
+      "loss": 0.6545,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23456670853057676,
+      "grad_norm": 0.7132095098495483,
+      "learning_rate": 9.460683338738485e-06,
+      "loss": 0.6463,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2347599265771423,
+      "grad_norm": 0.7034361958503723,
+      "learning_rate": 9.459158855521684e-06,
+      "loss": 0.6401,
+      "step": 1215
+    },
+    {
+      "epoch": 0.23495314462370787,
+      "grad_norm": 0.698146641254425,
+      "learning_rate": 9.457632343904404e-06,
+      "loss": 0.5924,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2351463626702734,
+      "grad_norm": 0.7495508790016174,
+      "learning_rate": 9.456103804581029e-06,
+      "loss": 0.672,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23533958071683894,
+      "grad_norm": 0.6623322367668152,
+      "learning_rate": 9.454573238246867e-06,
+      "loss": 0.6353,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2355327987634045,
+      "grad_norm": 0.7141072154045105,
+      "learning_rate": 9.45304064559815e-06,
+      "loss": 0.6185,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23572601680997005,
+      "grad_norm": 0.7194342613220215,
+      "learning_rate": 9.45150602733203e-06,
+      "loss": 0.6961,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2359192348565356,
+      "grad_norm": 0.6973040103912354,
+      "learning_rate": 9.449969384146578e-06,
+      "loss": 0.6216,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23611245290310115,
+      "grad_norm": 0.6743655800819397,
+      "learning_rate": 9.44843071674079e-06,
+      "loss": 0.638,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2363056709496667,
+      "grad_norm": 0.6546769738197327,
+      "learning_rate": 9.446890025814583e-06,
+      "loss": 0.6039,
+      "step": 1223
+    },
+    {
+      "epoch": 0.23649888899623225,
+      "grad_norm": 0.6159663200378418,
+      "learning_rate": 9.44534731206879e-06,
+      "loss": 0.6083,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2366921070427978,
+      "grad_norm": 0.7463479042053223,
+      "learning_rate": 9.44380257620517e-06,
+      "loss": 0.6938,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23688532508936336,
+      "grad_norm": 0.6805145740509033,
+      "learning_rate": 9.442255818926396e-06,
+      "loss": 0.6773,
+      "step": 1226
+    },
+    {
+      "epoch": 0.23707854313592888,
+      "grad_norm": 0.6652450561523438,
+      "learning_rate": 9.440707040936066e-06,
+      "loss": 0.6213,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23727176118249443,
+      "grad_norm": 0.6333974003791809,
+      "learning_rate": 9.439156242938693e-06,
+      "loss": 0.6127,
+      "step": 1228
+    },
+    {
+      "epoch": 0.23746497922905999,
+      "grad_norm": 0.7357860803604126,
+      "learning_rate": 9.437603425639708e-06,
+      "loss": 0.7228,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23765819727562554,
+      "grad_norm": 0.7023572325706482,
+      "learning_rate": 9.436048589745465e-06,
+      "loss": 0.6966,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2378514153221911,
+      "grad_norm": 0.6706063151359558,
+      "learning_rate": 9.434491735963236e-06,
+      "loss": 0.6783,
+      "step": 1231
+    },
+    {
+      "epoch": 0.23804463336875664,
+      "grad_norm": 0.6695427894592285,
+      "learning_rate": 9.432932865001206e-06,
+      "loss": 0.6359,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2382378514153222,
+      "grad_norm": 0.6954817771911621,
+      "learning_rate": 9.431371977568483e-06,
+      "loss": 0.6233,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23843106946188775,
+      "grad_norm": 0.6429852247238159,
+      "learning_rate": 9.429809074375089e-06,
+      "loss": 0.6087,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2386242875084533,
+      "grad_norm": 0.6513153314590454,
+      "learning_rate": 9.428244156131963e-06,
+      "loss": 0.5786,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23881750555501885,
+      "grad_norm": 0.6744505763053894,
+      "learning_rate": 9.426677223550961e-06,
+      "loss": 0.6196,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2390107236015844,
+      "grad_norm": 0.6530510783195496,
+      "learning_rate": 9.42510827734486e-06,
+      "loss": 0.6149,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23920394164814993,
+      "grad_norm": 0.6796987056732178,
+      "learning_rate": 9.423537318227342e-06,
+      "loss": 0.5701,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23939715969471548,
+      "grad_norm": 0.7942997217178345,
+      "learning_rate": 9.421964346913017e-06,
+      "loss": 0.6538,
+      "step": 1239
+    },
+    {
+      "epoch": 0.23959037774128103,
+      "grad_norm": 0.7620525360107422,
+      "learning_rate": 9.420389364117401e-06,
+      "loss": 0.7396,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23978359578784658,
+      "grad_norm": 0.7043704986572266,
+      "learning_rate": 9.418812370556931e-06,
+      "loss": 0.6659,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23997681383441213,
+      "grad_norm": 0.6823469400405884,
+      "learning_rate": 9.417233366948958e-06,
+      "loss": 0.5383,
+      "step": 1242
+    },
+    {
+      "epoch": 0.24017003188097769,
+      "grad_norm": 0.6695544719696045,
+      "learning_rate": 9.415652354011739e-06,
+      "loss": 0.6469,
+      "step": 1243
+    },
+    {
+      "epoch": 0.24036324992754324,
+      "grad_norm": 0.7239609956741333,
+      "learning_rate": 9.414069332464457e-06,
+      "loss": 0.6696,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2405564679741088,
+      "grad_norm": 0.7016108632087708,
+      "learning_rate": 9.412484303027202e-06,
+      "loss": 0.6206,
+      "step": 1245
+    },
+    {
+      "epoch": 0.24074968602067434,
+      "grad_norm": 0.6473593711853027,
+      "learning_rate": 9.41089726642098e-06,
+      "loss": 0.6063,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2409429040672399,
+      "grad_norm": 0.6433751583099365,
+      "learning_rate": 9.409308223367706e-06,
+      "loss": 0.5753,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24113612211380542,
+      "grad_norm": 0.7342225909233093,
+      "learning_rate": 9.407717174590212e-06,
+      "loss": 0.5895,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24132934016037097,
+      "grad_norm": 0.6566116809844971,
+      "learning_rate": 9.406124120812242e-06,
+      "loss": 0.6387,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24152255820693652,
+      "grad_norm": 0.6862229704856873,
+      "learning_rate": 9.404529062758447e-06,
+      "loss": 0.6591,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24171577625350207,
+      "grad_norm": 0.7207818031311035,
+      "learning_rate": 9.4029320011544e-06,
+      "loss": 0.5924,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24190899430006763,
+      "grad_norm": 0.7077277302742004,
+      "learning_rate": 9.401332936726572e-06,
+      "loss": 0.6684,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24210221234663318,
+      "grad_norm": 0.7245992422103882,
+      "learning_rate": 9.399731870202355e-06,
+      "loss": 0.6193,
+      "step": 1253
+    },
+    {
+      "epoch": 0.24229543039319873,
+      "grad_norm": 0.6966639757156372,
+      "learning_rate": 9.398128802310049e-06,
+      "loss": 0.6896,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24248864843976428,
+      "grad_norm": 0.7056952118873596,
+      "learning_rate": 9.396523733778866e-06,
+      "loss": 0.6787,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24268186648632983,
+      "grad_norm": 0.6863043904304504,
+      "learning_rate": 9.394916665338921e-06,
+      "loss": 0.6346,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24287508453289539,
+      "grad_norm": 0.7104870676994324,
+      "learning_rate": 9.393307597721248e-06,
+      "loss": 0.6853,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2430683025794609,
+      "grad_norm": 0.6768023371696472,
+      "learning_rate": 9.391696531657786e-06,
+      "loss": 0.5778,
+      "step": 1258
+    },
+    {
+      "epoch": 0.24326152062602646,
+      "grad_norm": 0.6683980822563171,
+      "learning_rate": 9.390083467881381e-06,
+      "loss": 0.6241,
+      "step": 1259
+    },
+    {
+      "epoch": 0.243454738672592,
+      "grad_norm": 0.6089362502098083,
+      "learning_rate": 9.388468407125792e-06,
+      "loss": 0.6059,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24364795671915757,
+      "grad_norm": 0.7082406878471375,
+      "learning_rate": 9.386851350125686e-06,
+      "loss": 0.6643,
+      "step": 1261
+    },
+    {
+      "epoch": 0.24384117476572312,
+      "grad_norm": 0.6735268831253052,
+      "learning_rate": 9.385232297616632e-06,
+      "loss": 0.617,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24403439281228867,
+      "grad_norm": 0.6751341819763184,
+      "learning_rate": 9.383611250335115e-06,
+      "loss": 0.6505,
+      "step": 1263
+    },
+    {
+      "epoch": 0.24422761085885422,
+      "grad_norm": 0.6730311512947083,
+      "learning_rate": 9.381988209018522e-06,
+      "loss": 0.631,
+      "step": 1264
+    },
+    {
+      "epoch": 0.24442082890541977,
+      "grad_norm": 0.6552103757858276,
+      "learning_rate": 9.380363174405149e-06,
+      "loss": 0.6657,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24461404695198533,
+      "grad_norm": 0.6176113486289978,
+      "learning_rate": 9.3787361472342e-06,
+      "loss": 0.6332,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24480726499855088,
+      "grad_norm": 0.6744945645332336,
+      "learning_rate": 9.377107128245782e-06,
+      "loss": 0.675,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2450004830451164,
+      "grad_norm": 0.6500509977340698,
+      "learning_rate": 9.375476118180908e-06,
+      "loss": 0.5728,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24519370109168195,
+      "grad_norm": 0.6308650374412537,
+      "learning_rate": 9.373843117781503e-06,
+      "loss": 0.6019,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2453869191382475,
+      "grad_norm": 0.685941755771637,
+      "learning_rate": 9.37220812779039e-06,
+      "loss": 0.6087,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24558013718481306,
+      "grad_norm": 0.6627559661865234,
+      "learning_rate": 9.370571148951301e-06,
+      "loss": 0.7392,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2457733552313786,
+      "grad_norm": 0.6973881125450134,
+      "learning_rate": 9.368932182008871e-06,
+      "loss": 0.6098,
+      "step": 1272
+    },
+    {
+      "epoch": 0.24596657327794416,
+      "grad_norm": 0.6960049271583557,
+      "learning_rate": 9.36729122770864e-06,
+      "loss": 0.6631,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2461597913245097,
+      "grad_norm": 0.7674120664596558,
+      "learning_rate": 9.365648286797052e-06,
+      "loss": 0.6814,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24635300937107527,
+      "grad_norm": 0.6666301488876343,
+      "learning_rate": 9.364003360021454e-06,
+      "loss": 0.6449,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24654622741764082,
+      "grad_norm": 0.6962092518806458,
+      "learning_rate": 9.3623564481301e-06,
+      "loss": 0.6597,
+      "step": 1276
+    },
+    {
+      "epoch": 0.24673944546420637,
+      "grad_norm": 0.6665223836898804,
+      "learning_rate": 9.360707551872144e-06,
+      "loss": 0.5324,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2469326635107719,
+      "grad_norm": 0.7016412615776062,
+      "learning_rate": 9.359056671997637e-06,
+      "loss": 0.6386,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24712588155733745,
+      "grad_norm": 0.7335931062698364,
+      "learning_rate": 9.357403809257544e-06,
+      "loss": 0.6454,
+      "step": 1279
+    },
+    {
+      "epoch": 0.247319099603903,
+      "grad_norm": 0.7252184748649597,
+      "learning_rate": 9.355748964403723e-06,
+      "loss": 0.7087,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24751231765046855,
+      "grad_norm": 0.7524668574333191,
+      "learning_rate": 9.354092138188937e-06,
+      "loss": 0.6658,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2477055356970341,
+      "grad_norm": 0.6784666180610657,
+      "learning_rate": 9.35243333136685e-06,
+      "loss": 0.5838,
+      "step": 1282
+    },
+    {
+      "epoch": 0.24789875374359965,
+      "grad_norm": 0.7665519118309021,
+      "learning_rate": 9.350772544692028e-06,
+      "loss": 0.7147,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2480919717901652,
+      "grad_norm": 0.636006236076355,
+      "learning_rate": 9.349109778919938e-06,
+      "loss": 0.6462,
+      "step": 1284
+    },
+    {
+      "epoch": 0.24828518983673076,
+      "grad_norm": 0.7340297102928162,
+      "learning_rate": 9.34744503480694e-06,
+      "loss": 0.6625,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2484784078832963,
+      "grad_norm": 0.7024398446083069,
+      "learning_rate": 9.345778313110302e-06,
+      "loss": 0.7151,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24867162592986186,
+      "grad_norm": 0.6922248601913452,
+      "learning_rate": 9.344109614588194e-06,
+      "loss": 0.5923,
+      "step": 1287
+    },
+    {
+      "epoch": 0.24886484397642739,
+      "grad_norm": 0.5752539038658142,
+      "learning_rate": 9.342438939999675e-06,
+      "loss": 0.5154,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24905806202299294,
+      "grad_norm": 0.6519090533256531,
+      "learning_rate": 9.34076629010471e-06,
+      "loss": 0.6168,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2492512800695585,
+      "grad_norm": 0.6388266086578369,
+      "learning_rate": 9.33909166566416e-06,
+      "loss": 0.5838,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24944449811612404,
+      "grad_norm": 0.7600163817405701,
+      "learning_rate": 9.337415067439785e-06,
+      "loss": 0.6334,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2496377161626896,
+      "grad_norm": 0.7046670913696289,
+      "learning_rate": 9.335736496194246e-06,
+      "loss": 0.5985,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24983093420925515,
+      "grad_norm": 0.6666012406349182,
+      "learning_rate": 9.334055952691096e-06,
+      "loss": 0.5921,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2500241522558207,
+      "grad_norm": 0.6516761183738708,
+      "learning_rate": 9.332373437694785e-06,
+      "loss": 0.5474,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2502173703023862,
+      "grad_norm": 0.7334963083267212,
+      "learning_rate": 9.330688951970667e-06,
+      "loss": 0.6431,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2504105883489518,
+      "grad_norm": 0.7188529968261719,
+      "learning_rate": 9.329002496284986e-06,
+      "loss": 0.6712,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2506038063955173,
+      "grad_norm": 0.6044237613677979,
+      "learning_rate": 9.327314071404884e-06,
+      "loss": 0.5811,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2507970244420829,
+      "grad_norm": 0.740854024887085,
+      "learning_rate": 9.325623678098396e-06,
+      "loss": 0.6382,
+      "step": 1298
+    },
+    {
+      "epoch": 0.25099024248864843,
+      "grad_norm": 0.7193633913993835,
+      "learning_rate": 9.32393131713446e-06,
+      "loss": 0.6646,
+      "step": 1299
+    },
+    {
+      "epoch": 0.251183460535214,
+      "grad_norm": 0.729121744632721,
+      "learning_rate": 9.3222369892829e-06,
+      "loss": 0.6287,
+      "step": 1300
+    },
+    {
+      "epoch": 0.25137667858177953,
+      "grad_norm": 0.7677316665649414,
+      "learning_rate": 9.32054069531444e-06,
+      "loss": 0.6419,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2515698966283451,
+      "grad_norm": 0.714491069316864,
+      "learning_rate": 9.318842436000697e-06,
+      "loss": 0.6174,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25176311467491064,
+      "grad_norm": 0.7498719692230225,
+      "learning_rate": 9.317142212114183e-06,
+      "loss": 0.6549,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25195633272147616,
+      "grad_norm": 0.7018754482269287,
+      "learning_rate": 9.315440024428303e-06,
+      "loss": 0.6553,
+      "step": 1304
+    },
+    {
+      "epoch": 0.25214955076804174,
+      "grad_norm": 0.7130201458930969,
+      "learning_rate": 9.313735873717352e-06,
+      "loss": 0.6636,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25234276881460727,
+      "grad_norm": 0.6303830742835999,
+      "learning_rate": 9.312029760756527e-06,
+      "loss": 0.5684,
+      "step": 1306
+    },
+    {
+      "epoch": 0.25253598686117285,
+      "grad_norm": 0.6996858716011047,
+      "learning_rate": 9.310321686321905e-06,
+      "loss": 0.6621,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25272920490773837,
+      "grad_norm": 0.6936467885971069,
+      "learning_rate": 9.308611651190467e-06,
+      "loss": 0.6711,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25292242295430395,
+      "grad_norm": 0.6534955501556396,
+      "learning_rate": 9.30689965614008e-06,
+      "loss": 0.5644,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2531156410008695,
+      "grad_norm": 0.6407191157341003,
+      "learning_rate": 9.305185701949501e-06,
+      "loss": 0.6218,
+      "step": 1310
+    },
+    {
+      "epoch": 0.25330885904743505,
+      "grad_norm": 0.6793675422668457,
+      "learning_rate": 9.303469789398381e-06,
+      "loss": 0.6516,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2535020770940006,
+      "grad_norm": 0.6567733287811279,
+      "learning_rate": 9.301751919267264e-06,
+      "loss": 0.6568,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2536952951405661,
+      "grad_norm": 0.6839560270309448,
+      "learning_rate": 9.300032092337577e-06,
+      "loss": 0.6008,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2538885131871317,
+      "grad_norm": 0.6825824975967407,
+      "learning_rate": 9.298310309391646e-06,
+      "loss": 0.6309,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2540817312336972,
+      "grad_norm": 0.7058524489402771,
+      "learning_rate": 9.296586571212682e-06,
+      "loss": 0.6132,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2542749492802628,
+      "grad_norm": 0.722435474395752,
+      "learning_rate": 9.294860878584786e-06,
+      "loss": 0.6172,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2544681673268283,
+      "grad_norm": 0.6743823885917664,
+      "learning_rate": 9.293133232292947e-06,
+      "loss": 0.675,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2546613853733939,
+      "grad_norm": 0.6788048148155212,
+      "learning_rate": 9.291403633123046e-06,
+      "loss": 0.628,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2548546034199594,
+      "grad_norm": 0.6541290879249573,
+      "learning_rate": 9.289672081861847e-06,
+      "loss": 0.6634,
+      "step": 1319
+    },
+    {
+      "epoch": 0.255047821466525,
+      "grad_norm": 0.6932106018066406,
+      "learning_rate": 9.287938579297009e-06,
+      "loss": 0.6328,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2552410395130905,
+      "grad_norm": 0.6388095617294312,
+      "learning_rate": 9.28620312621707e-06,
+      "loss": 0.6351,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2554342575596561,
+      "grad_norm": 0.6587784886360168,
+      "learning_rate": 9.284465723411465e-06,
+      "loss": 0.536,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2556274756062216,
+      "grad_norm": 0.7307528257369995,
+      "learning_rate": 9.282726371670508e-06,
+      "loss": 0.6518,
+      "step": 1323
+    },
+    {
+      "epoch": 0.25582069365278715,
+      "grad_norm": 0.6636239290237427,
+      "learning_rate": 9.280985071785402e-06,
+      "loss": 0.6148,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2560139116993527,
+      "grad_norm": 0.6947236657142639,
+      "learning_rate": 9.279241824548239e-06,
+      "loss": 0.639,
+      "step": 1325
+    },
+    {
+      "epoch": 0.25620712974591825,
+      "grad_norm": 0.6370094418525696,
+      "learning_rate": 9.277496630751993e-06,
+      "loss": 0.5776,
+      "step": 1326
+    },
+    {
+      "epoch": 0.25640034779248383,
+      "grad_norm": 0.6500885486602783,
+      "learning_rate": 9.275749491190524e-06,
+      "loss": 0.6409,
+      "step": 1327
+    },
+    {
+      "epoch": 0.25659356583904935,
+      "grad_norm": 0.6948850154876709,
+      "learning_rate": 9.27400040665858e-06,
+      "loss": 0.6596,
+      "step": 1328
+    },
+    {
+      "epoch": 0.25678678388561493,
+      "grad_norm": 0.6476846933364868,
+      "learning_rate": 9.27224937795179e-06,
+      "loss": 0.6422,
+      "step": 1329
+    },
+    {
+      "epoch": 0.25698000193218046,
+      "grad_norm": 0.6797247529029846,
+      "learning_rate": 9.270496405866671e-06,
+      "loss": 0.5749,
+      "step": 1330
+    },
+    {
+      "epoch": 0.25717321997874604,
+      "grad_norm": 0.6138875484466553,
+      "learning_rate": 9.268741491200622e-06,
+      "loss": 0.6122,
+      "step": 1331
+    },
+    {
+      "epoch": 0.25736643802531156,
+      "grad_norm": 0.6783483028411865,
+      "learning_rate": 9.266984634751922e-06,
+      "loss": 0.627,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25755965607187714,
+      "grad_norm": 0.6843644380569458,
+      "learning_rate": 9.265225837319742e-06,
+      "loss": 0.6555,
+      "step": 1333
+    },
+    {
+      "epoch": 0.25775287411844267,
+      "grad_norm": 0.6997473835945129,
+      "learning_rate": 9.26346509970413e-06,
+      "loss": 0.622,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2579460921650082,
+      "grad_norm": 0.7046927213668823,
+      "learning_rate": 9.261702422706014e-06,
+      "loss": 0.6548,
+      "step": 1335
+    },
+    {
+      "epoch": 0.25813931021157377,
+      "grad_norm": 0.6632881164550781,
+      "learning_rate": 9.259937807127211e-06,
+      "loss": 0.6803,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2583325282581393,
+      "grad_norm": 0.6286334991455078,
+      "learning_rate": 9.258171253770417e-06,
+      "loss": 0.6541,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2585257463047049,
+      "grad_norm": 0.6898749470710754,
+      "learning_rate": 9.256402763439207e-06,
+      "loss": 0.5692,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2587189643512704,
+      "grad_norm": 0.6879926919937134,
+      "learning_rate": 9.254632336938041e-06,
+      "loss": 0.6536,
+      "step": 1339
+    },
+    {
+      "epoch": 0.258912182397836,
+      "grad_norm": 0.6940664649009705,
+      "learning_rate": 9.252859975072255e-06,
+      "loss": 0.5955,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2591054004444015,
+      "grad_norm": 0.6852861642837524,
+      "learning_rate": 9.251085678648072e-06,
+      "loss": 0.5927,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2592986184909671,
+      "grad_norm": 0.6824904680252075,
+      "learning_rate": 9.249309448472588e-06,
+      "loss": 0.6931,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2594918365375326,
+      "grad_norm": 0.6450233459472656,
+      "learning_rate": 9.247531285353784e-06,
+      "loss": 0.637,
+      "step": 1343
+    },
+    {
+      "epoch": 0.25968505458409813,
+      "grad_norm": 0.6728276014328003,
+      "learning_rate": 9.245751190100519e-06,
+      "loss": 0.5545,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2598782726306637,
+      "grad_norm": 0.7501856684684753,
+      "learning_rate": 9.243969163522524e-06,
+      "loss": 0.6304,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26007149067722923,
+      "grad_norm": 0.6281067132949829,
+      "learning_rate": 9.242185206430423e-06,
+      "loss": 0.6233,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2602647087237948,
+      "grad_norm": 0.6946764588356018,
+      "learning_rate": 9.240399319635705e-06,
+      "loss": 0.647,
+      "step": 1347
+    },
+    {
+      "epoch": 0.26045792677036034,
+      "grad_norm": 0.6161202192306519,
+      "learning_rate": 9.238611503950741e-06,
+      "loss": 0.6729,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2606511448169259,
+      "grad_norm": 0.7080413103103638,
+      "learning_rate": 9.23682176018878e-06,
+      "loss": 0.6226,
+      "step": 1349
+    },
+    {
+      "epoch": 0.26084436286349144,
+      "grad_norm": 0.6799114346504211,
+      "learning_rate": 9.235030089163952e-06,
+      "loss": 0.567,
+      "step": 1350
+    },
+    {
+      "epoch": 0.261037580910057,
+      "grad_norm": 0.6607974171638489,
+      "learning_rate": 9.233236491691255e-06,
+      "loss": 0.639,
+      "step": 1351
+    },
+    {
+      "epoch": 0.26123079895662255,
+      "grad_norm": 0.6714608669281006,
+      "learning_rate": 9.231440968586572e-06,
+      "loss": 0.6155,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2614240170031881,
+      "grad_norm": 0.7131973505020142,
+      "learning_rate": 9.229643520666653e-06,
+      "loss": 0.6676,
+      "step": 1353
+    },
+    {
+      "epoch": 0.26161723504975365,
+      "grad_norm": 0.7094300985336304,
+      "learning_rate": 9.227844148749132e-06,
+      "loss": 0.6875,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2618104530963192,
+      "grad_norm": 0.7151849269866943,
+      "learning_rate": 9.226042853652514e-06,
+      "loss": 0.6567,
+      "step": 1355
+    },
+    {
+      "epoch": 0.26200367114288475,
+      "grad_norm": 0.7184725403785706,
+      "learning_rate": 9.22423963619618e-06,
+      "loss": 0.6351,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2621968891894503,
+      "grad_norm": 0.7010936141014099,
+      "learning_rate": 9.222434497200384e-06,
+      "loss": 0.6238,
+      "step": 1357
+    },
+    {
+      "epoch": 0.26239010723601586,
+      "grad_norm": 0.6577677726745605,
+      "learning_rate": 9.220627437486256e-06,
+      "loss": 0.5778,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2625833252825814,
+      "grad_norm": 0.6939574480056763,
+      "learning_rate": 9.218818457875797e-06,
+      "loss": 0.5882,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26277654332914696,
+      "grad_norm": 0.7126010656356812,
+      "learning_rate": 9.217007559191884e-06,
+      "loss": 0.6271,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2629697613757125,
+      "grad_norm": 0.7135534286499023,
+      "learning_rate": 9.215194742258266e-06,
+      "loss": 0.5799,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26316297942227806,
+      "grad_norm": 0.6752845644950867,
+      "learning_rate": 9.213380007899566e-06,
+      "loss": 0.6037,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2633561974688436,
+      "grad_norm": 0.6691304445266724,
+      "learning_rate": 9.211563356941277e-06,
+      "loss": 0.6292,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2635494155154091,
+      "grad_norm": 0.6356368660926819,
+      "learning_rate": 9.209744790209762e-06,
+      "loss": 0.6062,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2637426335619747,
+      "grad_norm": 0.6422574520111084,
+      "learning_rate": 9.207924308532264e-06,
+      "loss": 0.5968,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2639358516085402,
+      "grad_norm": 0.676967442035675,
+      "learning_rate": 9.206101912736884e-06,
+      "loss": 0.6118,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2641290696551058,
+      "grad_norm": 0.7227928042411804,
+      "learning_rate": 9.204277603652607e-06,
+      "loss": 0.6189,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2643222877016713,
+      "grad_norm": 0.7035180330276489,
+      "learning_rate": 9.202451382109278e-06,
+      "loss": 0.6804,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2645155057482369,
+      "grad_norm": 0.6967726349830627,
+      "learning_rate": 9.200623248937619e-06,
+      "loss": 0.6589,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2647087237948024,
+      "grad_norm": 0.6449087858200073,
+      "learning_rate": 9.198793204969217e-06,
+      "loss": 0.5868,
+      "step": 1370
+    },
+    {
+      "epoch": 0.264901941841368,
+      "grad_norm": 0.6756646633148193,
+      "learning_rate": 9.196961251036529e-06,
+      "loss": 0.6199,
+      "step": 1371
+    },
+    {
+      "epoch": 0.26509515988793353,
+      "grad_norm": 0.6303045153617859,
+      "learning_rate": 9.195127387972884e-06,
+      "loss": 0.5902,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2652883779344991,
+      "grad_norm": 0.6270753741264343,
+      "learning_rate": 9.193291616612477e-06,
+      "loss": 0.7038,
+      "step": 1373
+    },
+    {
+      "epoch": 0.26548159598106463,
+      "grad_norm": 0.7049100399017334,
+      "learning_rate": 9.19145393779037e-06,
+      "loss": 0.622,
+      "step": 1374
+    },
+    {
+      "epoch": 0.26567481402763016,
+      "grad_norm": 0.7319111824035645,
+      "learning_rate": 9.189614352342495e-06,
+      "loss": 0.6748,
+      "step": 1375
+    },
+    {
+      "epoch": 0.26586803207419574,
+      "grad_norm": 0.6673938632011414,
+      "learning_rate": 9.187772861105648e-06,
+      "loss": 0.5804,
+      "step": 1376
+    },
+    {
+      "epoch": 0.26606125012076126,
+      "grad_norm": 0.6885422468185425,
+      "learning_rate": 9.185929464917497e-06,
+      "loss": 0.6362,
+      "step": 1377
+    },
+    {
+      "epoch": 0.26625446816732684,
+      "grad_norm": 0.6877424716949463,
+      "learning_rate": 9.184084164616571e-06,
+      "loss": 0.6347,
+      "step": 1378
+    },
+    {
+      "epoch": 0.26644768621389237,
+      "grad_norm": 0.7038718461990356,
+      "learning_rate": 9.18223696104227e-06,
+      "loss": 0.6387,
+      "step": 1379
+    },
+    {
+      "epoch": 0.26664090426045794,
+      "grad_norm": 0.732409656047821,
+      "learning_rate": 9.180387855034858e-06,
+      "loss": 0.6455,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26683412230702347,
+      "grad_norm": 0.6192600727081299,
+      "learning_rate": 9.17853684743546e-06,
+      "loss": 0.6366,
+      "step": 1381
+    },
+    {
+      "epoch": 0.26702734035358905,
+      "grad_norm": 0.6902378797531128,
+      "learning_rate": 9.176683939086072e-06,
+      "loss": 0.645,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2672205584001546,
+      "grad_norm": 0.661967933177948,
+      "learning_rate": 9.174829130829556e-06,
+      "loss": 0.6638,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2674137764467201,
+      "grad_norm": 0.6804383397102356,
+      "learning_rate": 9.172972423509631e-06,
+      "loss": 0.5894,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2676069944932857,
+      "grad_norm": 0.6567407250404358,
+      "learning_rate": 9.171113817970882e-06,
+      "loss": 0.6151,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2678002125398512,
+      "grad_norm": 0.7250513434410095,
+      "learning_rate": 9.169253315058764e-06,
+      "loss": 0.6551,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2679934305864168,
+      "grad_norm": 0.7052079439163208,
+      "learning_rate": 9.167390915619588e-06,
+      "loss": 0.6403,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2681866486329823,
+      "grad_norm": 0.6788066625595093,
+      "learning_rate": 9.16552662050053e-06,
+      "loss": 0.6519,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2683798666795479,
+      "grad_norm": 0.7103116512298584,
+      "learning_rate": 9.163660430549624e-06,
+      "loss": 0.6587,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2685730847261134,
+      "grad_norm": 0.6283394694328308,
+      "learning_rate": 9.161792346615778e-06,
+      "loss": 0.594,
+      "step": 1390
+    },
+    {
+      "epoch": 0.268766302772679,
+      "grad_norm": 0.7346019744873047,
+      "learning_rate": 9.159922369548749e-06,
+      "loss": 0.6528,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2689595208192445,
+      "grad_norm": 0.7051128149032593,
+      "learning_rate": 9.158050500199159e-06,
+      "loss": 0.6369,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2691527388658101,
+      "grad_norm": 0.6564698815345764,
+      "learning_rate": 9.156176739418494e-06,
+      "loss": 0.623,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2693459569123756,
+      "grad_norm": 0.7223008871078491,
+      "learning_rate": 9.154301088059096e-06,
+      "loss": 0.6557,
+      "step": 1394
+    },
+    {
+      "epoch": 0.26953917495894114,
+      "grad_norm": 0.6690618395805359,
+      "learning_rate": 9.152423546974172e-06,
+      "loss": 0.6456,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2697323930055067,
+      "grad_norm": 0.7860140800476074,
+      "learning_rate": 9.150544117017783e-06,
+      "loss": 0.7035,
+      "step": 1396
+    },
+    {
+      "epoch": 0.26992561105207225,
+      "grad_norm": 0.6654738783836365,
+      "learning_rate": 9.148662799044852e-06,
+      "loss": 0.6172,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2701188290986378,
+      "grad_norm": 0.7082982659339905,
+      "learning_rate": 9.146779593911162e-06,
+      "loss": 0.7053,
+      "step": 1398
+    },
+    {
+      "epoch": 0.27031204714520335,
+      "grad_norm": 0.6523699760437012,
+      "learning_rate": 9.144894502473353e-06,
+      "loss": 0.65,
+      "step": 1399
+    },
+    {
+      "epoch": 0.27050526519176893,
+      "grad_norm": 0.6714527606964111,
+      "learning_rate": 9.14300752558892e-06,
+      "loss": 0.6615,
+      "step": 1400
+    },
+    {
+      "epoch": 0.27069848323833445,
+      "grad_norm": 0.6754347085952759,
+      "learning_rate": 9.141118664116223e-06,
+      "loss": 0.659,
+      "step": 1401
+    },
+    {
+      "epoch": 0.27089170128490003,
+      "grad_norm": 0.7002037763595581,
+      "learning_rate": 9.139227918914473e-06,
+      "loss": 0.6301,
+      "step": 1402
+    },
+    {
+      "epoch": 0.27108491933146556,
+      "grad_norm": 0.6453575491905212,
+      "learning_rate": 9.13733529084374e-06,
+      "loss": 0.5786,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2712781373780311,
+      "grad_norm": 0.6752354502677917,
+      "learning_rate": 9.13544078076495e-06,
+      "loss": 0.6162,
+      "step": 1404
+    },
+    {
+      "epoch": 0.27147135542459666,
+      "grad_norm": 0.6364012956619263,
+      "learning_rate": 9.133544389539888e-06,
+      "loss": 0.578,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2716645734711622,
+      "grad_norm": 0.7436330914497375,
+      "learning_rate": 9.131646118031188e-06,
+      "loss": 0.6668,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27185779151772776,
+      "grad_norm": 0.6687547564506531,
+      "learning_rate": 9.129745967102347e-06,
+      "loss": 0.6373,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2720510095642933,
+      "grad_norm": 0.6435561180114746,
+      "learning_rate": 9.127843937617714e-06,
+      "loss": 0.6564,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27224422761085887,
+      "grad_norm": 0.6614880561828613,
+      "learning_rate": 9.125940030442487e-06,
+      "loss": 0.5933,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2724374456574244,
+      "grad_norm": 0.6977416276931763,
+      "learning_rate": 9.124034246442727e-06,
+      "loss": 0.6483,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27263066370399,
+      "grad_norm": 0.6783322095870972,
+      "learning_rate": 9.122126586485346e-06,
+      "loss": 0.6497,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2728238817505555,
+      "grad_norm": 0.6800594329833984,
+      "learning_rate": 9.120217051438106e-06,
+      "loss": 0.6587,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2730170997971211,
+      "grad_norm": 0.7166368961334229,
+      "learning_rate": 9.118305642169625e-06,
+      "loss": 0.6593,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2732103178436866,
+      "grad_norm": 0.6323719024658203,
+      "learning_rate": 9.116392359549372e-06,
+      "loss": 0.633,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2734035358902521,
+      "grad_norm": 0.6394928097724915,
+      "learning_rate": 9.114477204447668e-06,
+      "loss": 0.6204,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2735967539368177,
+      "grad_norm": 0.6801345944404602,
+      "learning_rate": 9.112560177735687e-06,
+      "loss": 0.6294,
+      "step": 1416
+    },
+    {
+      "epoch": 0.27378997198338323,
+      "grad_norm": 0.7697802782058716,
+      "learning_rate": 9.110641280285456e-06,
+      "loss": 0.6222,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2739831900299488,
+      "grad_norm": 0.6736264228820801,
+      "learning_rate": 9.10872051296985e-06,
+      "loss": 0.5771,
+      "step": 1418
+    },
+    {
+      "epoch": 0.27417640807651433,
+      "grad_norm": 0.6663708090782166,
+      "learning_rate": 9.106797876662596e-06,
+      "loss": 0.6536,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2743696261230799,
+      "grad_norm": 0.6007384061813354,
+      "learning_rate": 9.104873372238269e-06,
+      "loss": 0.55,
+      "step": 1420
+    },
+    {
+      "epoch": 0.27456284416964544,
+      "grad_norm": 0.6675135493278503,
+      "learning_rate": 9.102947000572297e-06,
+      "loss": 0.5966,
+      "step": 1421
+    },
+    {
+      "epoch": 0.274756062216211,
+      "grad_norm": 0.7029179334640503,
+      "learning_rate": 9.101018762540954e-06,
+      "loss": 0.5982,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27494928026277654,
+      "grad_norm": 0.6796406507492065,
+      "learning_rate": 9.099088659021367e-06,
+      "loss": 0.6585,
+      "step": 1423
+    },
+    {
+      "epoch": 0.27514249830934207,
+      "grad_norm": 0.6680667400360107,
+      "learning_rate": 9.097156690891508e-06,
+      "loss": 0.6593,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27533571635590764,
+      "grad_norm": 0.6278805136680603,
+      "learning_rate": 9.095222859030199e-06,
+      "loss": 0.6185,
+      "step": 1425
+    },
+    {
+      "epoch": 0.27552893440247317,
+      "grad_norm": 0.7050215601921082,
+      "learning_rate": 9.09328716431711e-06,
+      "loss": 0.6003,
+      "step": 1426
+    },
+    {
+      "epoch": 0.27572215244903875,
+      "grad_norm": 0.7046961784362793,
+      "learning_rate": 9.091349607632755e-06,
+      "loss": 0.6471,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2759153704956043,
+      "grad_norm": 0.6325632929801941,
+      "learning_rate": 9.089410189858502e-06,
+      "loss": 0.6094,
+      "step": 1428
+    },
+    {
+      "epoch": 0.27610858854216985,
+      "grad_norm": 0.7788944840431213,
+      "learning_rate": 9.087468911876558e-06,
+      "loss": 0.695,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2763018065887354,
+      "grad_norm": 0.6748207211494446,
+      "learning_rate": 9.085525774569983e-06,
+      "loss": 0.6655,
+      "step": 1430
+    },
+    {
+      "epoch": 0.27649502463530096,
+      "grad_norm": 0.698729395866394,
+      "learning_rate": 9.083580778822674e-06,
+      "loss": 0.6276,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2766882426818665,
+      "grad_norm": 0.6781487464904785,
+      "learning_rate": 9.081633925519384e-06,
+      "loss": 0.685,
+      "step": 1432
+    },
+    {
+      "epoch": 0.27688146072843206,
+      "grad_norm": 0.6931982636451721,
+      "learning_rate": 9.079685215545702e-06,
+      "loss": 0.641,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2770746787749976,
+      "grad_norm": 0.6546832919120789,
+      "learning_rate": 9.077734649788067e-06,
+      "loss": 0.5967,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2772678968215631,
+      "grad_norm": 0.6401552557945251,
+      "learning_rate": 9.075782229133759e-06,
+      "loss": 0.5935,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2774611148681287,
+      "grad_norm": 0.6719592213630676,
+      "learning_rate": 9.073827954470904e-06,
+      "loss": 0.6605,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2776543329146942,
+      "grad_norm": 0.7003101110458374,
+      "learning_rate": 9.071871826688472e-06,
+      "loss": 0.6446,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2778475509612598,
+      "grad_norm": 0.6478577256202698,
+      "learning_rate": 9.069913846676272e-06,
+      "loss": 0.6003,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2780407690078253,
+      "grad_norm": 0.7326667308807373,
+      "learning_rate": 9.067954015324958e-06,
+      "loss": 0.6346,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2782339870543909,
+      "grad_norm": 0.6476895809173584,
+      "learning_rate": 9.065992333526029e-06,
+      "loss": 0.6146,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2784272051009564,
+      "grad_norm": 0.6810159683227539,
+      "learning_rate": 9.064028802171819e-06,
+      "loss": 0.6332,
+      "step": 1441
+    },
+    {
+      "epoch": 0.278620423147522,
+      "grad_norm": 0.6619317531585693,
+      "learning_rate": 9.062063422155506e-06,
+      "loss": 0.651,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2788136411940875,
+      "grad_norm": 0.6919517517089844,
+      "learning_rate": 9.060096194371118e-06,
+      "loss": 0.6357,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2790068592406531,
+      "grad_norm": 0.6377819776535034,
+      "learning_rate": 9.058127119713509e-06,
+      "loss": 0.6056,
+      "step": 1444
+    },
+    {
+      "epoch": 0.27920007728721863,
+      "grad_norm": 0.6609470844268799,
+      "learning_rate": 9.056156199078383e-06,
+      "loss": 0.6164,
+      "step": 1445
+    },
+    {
+      "epoch": 0.27939329533378415,
+      "grad_norm": 0.7095999121665955,
+      "learning_rate": 9.054183433362277e-06,
+      "loss": 0.6757,
+      "step": 1446
+    },
+    {
+      "epoch": 0.27958651338034973,
+      "grad_norm": 0.6659539341926575,
+      "learning_rate": 9.052208823462573e-06,
+      "loss": 0.6335,
+      "step": 1447
+    },
+    {
+      "epoch": 0.27977973142691526,
+      "grad_norm": 0.665187656879425,
+      "learning_rate": 9.050232370277491e-06,
+      "loss": 0.6073,
+      "step": 1448
+    },
+    {
+      "epoch": 0.27997294947348084,
+      "grad_norm": 0.6914938688278198,
+      "learning_rate": 9.048254074706085e-06,
+      "loss": 0.6581,
+      "step": 1449
+    },
+    {
+      "epoch": 0.28016616752004636,
+      "grad_norm": 0.7124966382980347,
+      "learning_rate": 9.046273937648252e-06,
+      "loss": 0.7081,
+      "step": 1450
+    },
+    {
+      "epoch": 0.28035938556661194,
+      "grad_norm": 0.6893721222877502,
+      "learning_rate": 9.044291960004726e-06,
+      "loss": 0.6665,
+      "step": 1451
+    },
+    {
+      "epoch": 0.28055260361317746,
+      "grad_norm": 0.6703068017959595,
+      "learning_rate": 9.042308142677074e-06,
+      "loss": 0.5892,
+      "step": 1452
+    },
+    {
+      "epoch": 0.28074582165974304,
+      "grad_norm": 0.6970689296722412,
+      "learning_rate": 9.040322486567704e-06,
+      "loss": 0.6196,
+      "step": 1453
+    },
+    {
+      "epoch": 0.28093903970630857,
+      "grad_norm": 0.674347996711731,
+      "learning_rate": 9.038334992579863e-06,
+      "loss": 0.6482,
+      "step": 1454
+    },
+    {
+      "epoch": 0.2811322577528741,
+      "grad_norm": 0.687533438205719,
+      "learning_rate": 9.036345661617626e-06,
+      "loss": 0.6066,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2813254757994397,
+      "grad_norm": 0.6901268362998962,
+      "learning_rate": 9.034354494585911e-06,
+      "loss": 0.6308,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2815186938460052,
+      "grad_norm": 0.7019492983818054,
+      "learning_rate": 9.032361492390464e-06,
+      "loss": 0.6262,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2817119118925708,
+      "grad_norm": 0.6911032199859619,
+      "learning_rate": 9.030366655937875e-06,
+      "loss": 0.6107,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2819051299391363,
+      "grad_norm": 0.6633785367012024,
+      "learning_rate": 9.028369986135561e-06,
+      "loss": 0.6177,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2820983479857019,
+      "grad_norm": 0.6996631622314453,
+      "learning_rate": 9.026371483891775e-06,
+      "loss": 0.5983,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2822915660322674,
+      "grad_norm": 0.665198564529419,
+      "learning_rate": 9.024371150115606e-06,
+      "loss": 0.5811,
+      "step": 1461
+    },
+    {
+      "epoch": 0.282484784078833,
+      "grad_norm": 0.735887885093689,
+      "learning_rate": 9.02236898571697e-06,
+      "loss": 0.6548,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2826780021253985,
+      "grad_norm": 0.7405731081962585,
+      "learning_rate": 9.020364991606625e-06,
+      "loss": 0.6345,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2828712201719641,
+      "grad_norm": 0.678359866142273,
+      "learning_rate": 9.018359168696152e-06,
+      "loss": 0.6348,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2830644382185296,
+      "grad_norm": 0.6538341045379639,
+      "learning_rate": 9.016351517897968e-06,
+      "loss": 0.6377,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28325765626509514,
+      "grad_norm": 0.6717284917831421,
+      "learning_rate": 9.014342040125324e-06,
+      "loss": 0.6327,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2834508743116607,
+      "grad_norm": 0.6848226189613342,
+      "learning_rate": 9.012330736292299e-06,
+      "loss": 0.6611,
+      "step": 1467
+    },
+    {
+      "epoch": 0.28364409235822624,
+      "grad_norm": 0.7523446679115295,
+      "learning_rate": 9.010317607313801e-06,
+      "loss": 0.6732,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2838373104047918,
+      "grad_norm": 0.7628350853919983,
+      "learning_rate": 9.008302654105573e-06,
+      "loss": 0.6556,
+      "step": 1469
+    },
+    {
+      "epoch": 0.28403052845135734,
+      "grad_norm": 0.7232263088226318,
+      "learning_rate": 9.006285877584183e-06,
+      "loss": 0.644,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2842237464979229,
+      "grad_norm": 0.6396799683570862,
+      "learning_rate": 9.004267278667032e-06,
+      "loss": 0.6218,
+      "step": 1471
+    },
+    {
+      "epoch": 0.28441696454448845,
+      "grad_norm": 0.7226306200027466,
+      "learning_rate": 9.002246858272347e-06,
+      "loss": 0.6254,
+      "step": 1472
+    },
+    {
+      "epoch": 0.28461018259105403,
+      "grad_norm": 0.7279789447784424,
+      "learning_rate": 9.000224617319189e-06,
+      "loss": 0.719,
+      "step": 1473
+    },
+    {
+      "epoch": 0.28480340063761955,
+      "grad_norm": 0.7415453195571899,
+      "learning_rate": 8.998200556727439e-06,
+      "loss": 0.7012,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2849966186841851,
+      "grad_norm": 0.7683239579200745,
+      "learning_rate": 8.99617467741781e-06,
+      "loss": 0.6413,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28518983673075066,
+      "grad_norm": 0.6824908256530762,
+      "learning_rate": 8.994146980311845e-06,
+      "loss": 0.6241,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2853830547773162,
+      "grad_norm": 0.6401504874229431,
+      "learning_rate": 8.99211746633191e-06,
+      "loss": 0.5591,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28557627282388176,
+      "grad_norm": 0.6179832220077515,
+      "learning_rate": 8.990086136401199e-06,
+      "loss": 0.6214,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2857694908704473,
+      "grad_norm": 0.6373721957206726,
+      "learning_rate": 8.98805299144373e-06,
+      "loss": 0.5758,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28596270891701286,
+      "grad_norm": 0.6565260291099548,
+      "learning_rate": 8.986018032384349e-06,
+      "loss": 0.6316,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2861559269635784,
+      "grad_norm": 0.6949194073677063,
+      "learning_rate": 8.983981260148728e-06,
+      "loss": 0.7003,
+      "step": 1481
+    },
+    {
+      "epoch": 0.28634914501014397,
+      "grad_norm": 0.6738445162773132,
+      "learning_rate": 8.981942675663363e-06,
+      "loss": 0.5899,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2865423630567095,
+      "grad_norm": 0.6757757663726807,
+      "learning_rate": 8.979902279855572e-06,
+      "loss": 0.6309,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2867355811032751,
+      "grad_norm": 0.6691810488700867,
+      "learning_rate": 8.977860073653497e-06,
+      "loss": 0.6523,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2869287991498406,
+      "grad_norm": 0.6313115954399109,
+      "learning_rate": 8.975816057986107e-06,
+      "loss": 0.5736,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2871220171964061,
+      "grad_norm": 0.6729751825332642,
+      "learning_rate": 8.973770233783192e-06,
+      "loss": 0.625,
+      "step": 1486
+    },
+    {
+      "epoch": 0.2873152352429717,
+      "grad_norm": 0.7062031030654907,
+      "learning_rate": 8.971722601975366e-06,
+      "loss": 0.59,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2875084532895372,
+      "grad_norm": 0.6396594643592834,
+      "learning_rate": 8.969673163494063e-06,
+      "loss": 0.5894,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2877016713361028,
+      "grad_norm": 0.7156132459640503,
+      "learning_rate": 8.96762191927154e-06,
+      "loss": 0.6654,
+      "step": 1489
+    },
+    {
+      "epoch": 0.28789488938266833,
+      "grad_norm": 0.6878542900085449,
+      "learning_rate": 8.965568870240876e-06,
+      "loss": 0.6177,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2880881074292339,
+      "grad_norm": 0.6550641655921936,
+      "learning_rate": 8.963514017335967e-06,
+      "loss": 0.5937,
+      "step": 1491
+    },
+    {
+      "epoch": 0.28828132547579943,
+      "grad_norm": 0.7460904121398926,
+      "learning_rate": 8.96145736149154e-06,
+      "loss": 0.6569,
+      "step": 1492
+    },
+    {
+      "epoch": 0.288474543522365,
+      "grad_norm": 0.6946448683738708,
+      "learning_rate": 8.95939890364313e-06,
+      "loss": 0.6623,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28866776156893054,
+      "grad_norm": 0.6982112526893616,
+      "learning_rate": 8.957338644727095e-06,
+      "loss": 0.6181,
+      "step": 1494
+    },
+    {
+      "epoch": 0.28886097961549606,
+      "grad_norm": 0.6668018102645874,
+      "learning_rate": 8.95527658568062e-06,
+      "loss": 0.6319,
+      "step": 1495
+    },
+    {
+      "epoch": 0.28905419766206164,
+      "grad_norm": 0.7239404916763306,
+      "learning_rate": 8.953212727441697e-06,
+      "loss": 0.5923,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28924741570862716,
+      "grad_norm": 0.6917080283164978,
+      "learning_rate": 8.951147070949147e-06,
+      "loss": 0.606,
+      "step": 1497
+    },
+    {
+      "epoch": 0.28944063375519274,
+      "grad_norm": 0.7642697095870972,
+      "learning_rate": 8.949079617142601e-06,
+      "loss": 0.6179,
+      "step": 1498
+    },
+    {
+      "epoch": 0.28963385180175827,
+      "grad_norm": 0.6476341485977173,
+      "learning_rate": 8.947010366962511e-06,
+      "loss": 0.6303,
+      "step": 1499
+    },
+    {
+      "epoch": 0.28982706984832385,
+      "grad_norm": 0.6693698167800903,
+      "learning_rate": 8.94493932135015e-06,
+      "loss": 0.5574,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2900202878948894,
+      "grad_norm": 0.7053855061531067,
+      "learning_rate": 8.942866481247599e-06,
+      "loss": 0.6346,
+      "step": 1501
+    },
+    {
+      "epoch": 0.29021350594145495,
+      "grad_norm": 0.6754406094551086,
+      "learning_rate": 8.94079184759776e-06,
+      "loss": 0.5746,
+      "step": 1502
+    },
+    {
+      "epoch": 0.2904067239880205,
+      "grad_norm": 0.664268434047699,
+      "learning_rate": 8.938715421344355e-06,
+      "loss": 0.6667,
+      "step": 1503
+    },
+    {
+      "epoch": 0.29059994203458606,
+      "grad_norm": 0.6724767684936523,
+      "learning_rate": 8.936637203431914e-06,
+      "loss": 0.6222,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2907931600811516,
+      "grad_norm": 0.6487547755241394,
+      "learning_rate": 8.934557194805787e-06,
+      "loss": 0.6769,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2909863781277171,
+      "grad_norm": 0.6595832109451294,
+      "learning_rate": 8.932475396412135e-06,
+      "loss": 0.6539,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2911795961742827,
+      "grad_norm": 0.7754471898078918,
+      "learning_rate": 8.930391809197938e-06,
+      "loss": 0.6775,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2913728142208482,
+      "grad_norm": 0.6859375238418579,
+      "learning_rate": 8.928306434110982e-06,
+      "loss": 0.5453,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2915660322674138,
+      "grad_norm": 0.6574501395225525,
+      "learning_rate": 8.926219272099876e-06,
+      "loss": 0.6443,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2917592503139793,
+      "grad_norm": 0.6686326265335083,
+      "learning_rate": 8.924130324114035e-06,
+      "loss": 0.6099,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2919524683605449,
+      "grad_norm": 0.7561066150665283,
+      "learning_rate": 8.922039591103686e-06,
+      "loss": 0.6511,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2921456864071104,
+      "grad_norm": 0.6241776943206787,
+      "learning_rate": 8.919947074019874e-06,
+      "loss": 0.614,
+      "step": 1512
+    },
+    {
+      "epoch": 0.292338904453676,
+      "grad_norm": 0.6215047836303711,
+      "learning_rate": 8.917852773814449e-06,
+      "loss": 0.6323,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2925321225002415,
+      "grad_norm": 0.6106544137001038,
+      "learning_rate": 8.915756691440074e-06,
+      "loss": 0.6008,
+      "step": 1514
+    },
+    {
+      "epoch": 0.29272534054680704,
+      "grad_norm": 0.6516183018684387,
+      "learning_rate": 8.913658827850229e-06,
+      "loss": 0.6363,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2929185585933726,
+      "grad_norm": 0.7012656927108765,
+      "learning_rate": 8.91155918399919e-06,
+      "loss": 0.6108,
+      "step": 1516
+    },
+    {
+      "epoch": 0.29311177663993815,
+      "grad_norm": 0.7504993677139282,
+      "learning_rate": 8.90945776084206e-06,
+      "loss": 0.6894,
+      "step": 1517
+    },
+    {
+      "epoch": 0.29330499468650373,
+      "grad_norm": 0.6714639663696289,
+      "learning_rate": 8.90735455933474e-06,
+      "loss": 0.6472,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29349821273306925,
+      "grad_norm": 0.6767387390136719,
+      "learning_rate": 8.905249580433941e-06,
+      "loss": 0.6072,
+      "step": 1519
+    },
+    {
+      "epoch": 0.29369143077963483,
+      "grad_norm": 0.6911678910255432,
+      "learning_rate": 8.903142825097187e-06,
+      "loss": 0.6029,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29388464882620036,
+      "grad_norm": 0.6555519700050354,
+      "learning_rate": 8.901034294282807e-06,
+      "loss": 0.5796,
+      "step": 1521
+    },
+    {
+      "epoch": 0.29407786687276594,
+      "grad_norm": 0.688406765460968,
+      "learning_rate": 8.898923988949936e-06,
+      "loss": 0.6266,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29427108491933146,
+      "grad_norm": 0.6644681692123413,
+      "learning_rate": 8.89681191005852e-06,
+      "loss": 0.6374,
+      "step": 1523
+    },
+    {
+      "epoch": 0.29446430296589704,
+      "grad_norm": 0.6141144633293152,
+      "learning_rate": 8.894698058569312e-06,
+      "loss": 0.5937,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29465752101246256,
+      "grad_norm": 0.8370233774185181,
+      "learning_rate": 8.892582435443866e-06,
+      "loss": 0.6648,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2948507390590281,
+      "grad_norm": 0.6980727910995483,
+      "learning_rate": 8.890465041644548e-06,
+      "loss": 0.6506,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29504395710559367,
+      "grad_norm": 0.7163034081459045,
+      "learning_rate": 8.888345878134526e-06,
+      "loss": 0.7125,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2952371751521592,
+      "grad_norm": 0.7097302079200745,
+      "learning_rate": 8.88622494587777e-06,
+      "loss": 0.6531,
+      "step": 1528
+    },
+    {
+      "epoch": 0.29543039319872477,
+      "grad_norm": 0.6747966408729553,
+      "learning_rate": 8.884102245839066e-06,
+      "loss": 0.6221,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2956236112452903,
+      "grad_norm": 0.681468665599823,
+      "learning_rate": 8.881977778983992e-06,
+      "loss": 0.6205,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2958168292918559,
+      "grad_norm": 0.6919286847114563,
+      "learning_rate": 8.879851546278933e-06,
+      "loss": 0.6233,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2960100473384214,
+      "grad_norm": 0.6992058753967285,
+      "learning_rate": 8.87772354869108e-06,
+      "loss": 0.6411,
+      "step": 1532
+    },
+    {
+      "epoch": 0.296203265384987,
+      "grad_norm": 0.6766147613525391,
+      "learning_rate": 8.875593787188426e-06,
+      "loss": 0.6264,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2963964834315525,
+      "grad_norm": 0.6689425706863403,
+      "learning_rate": 8.873462262739764e-06,
+      "loss": 0.5586,
+      "step": 1534
+    },
+    {
+      "epoch": 0.29658970147811803,
+      "grad_norm": 0.6772864460945129,
+      "learning_rate": 8.871328976314692e-06,
+      "loss": 0.637,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2967829195246836,
+      "grad_norm": 0.7240951061248779,
+      "learning_rate": 8.869193928883608e-06,
+      "loss": 0.6968,
+      "step": 1536
+    },
+    {
+      "epoch": 0.29697613757124913,
+      "grad_norm": 0.7312039732933044,
+      "learning_rate": 8.86705712141771e-06,
+      "loss": 0.6676,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2971693556178147,
+      "grad_norm": 0.7157521843910217,
+      "learning_rate": 8.864918554889e-06,
+      "loss": 0.737,
+      "step": 1538
+    },
+    {
+      "epoch": 0.29736257366438024,
+      "grad_norm": 0.6912029385566711,
+      "learning_rate": 8.862778230270276e-06,
+      "loss": 0.6222,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2975557917109458,
+      "grad_norm": 0.6883869767189026,
+      "learning_rate": 8.860636148535138e-06,
+      "loss": 0.6138,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29774900975751134,
+      "grad_norm": 0.6661552786827087,
+      "learning_rate": 8.858492310657985e-06,
+      "loss": 0.6539,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2979422278040769,
+      "grad_norm": 0.6614292860031128,
+      "learning_rate": 8.856346717614014e-06,
+      "loss": 0.5791,
+      "step": 1542
+    },
+    {
+      "epoch": 0.29813544585064244,
+      "grad_norm": 0.648363471031189,
+      "learning_rate": 8.854199370379222e-06,
+      "loss": 0.5678,
+      "step": 1543
+    },
+    {
+      "epoch": 0.298328663897208,
+      "grad_norm": 0.6516388654708862,
+      "learning_rate": 8.852050269930404e-06,
+      "loss": 0.5864,
+      "step": 1544
+    },
+    {
+      "epoch": 0.29852188194377355,
+      "grad_norm": 0.7678657174110413,
+      "learning_rate": 8.849899417245149e-06,
+      "loss": 0.6481,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2987150999903391,
+      "grad_norm": 0.6451981067657471,
+      "learning_rate": 8.84774681330185e-06,
+      "loss": 0.589,
+      "step": 1546
+    },
+    {
+      "epoch": 0.29890831803690465,
+      "grad_norm": 0.6721261739730835,
+      "learning_rate": 8.845592459079688e-06,
+      "loss": 0.6563,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2991015360834702,
+      "grad_norm": 0.7379022836685181,
+      "learning_rate": 8.843436355558645e-06,
+      "loss": 0.6487,
+      "step": 1548
+    },
+    {
+      "epoch": 0.29929475413003576,
+      "grad_norm": 0.6845641732215881,
+      "learning_rate": 8.841278503719502e-06,
+      "loss": 0.6397,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2994879721766013,
+      "grad_norm": 0.7007449269294739,
+      "learning_rate": 8.839118904543827e-06,
+      "loss": 0.6607,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29968119022316686,
+      "grad_norm": 0.650937557220459,
+      "learning_rate": 8.83695755901399e-06,
+      "loss": 0.602,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2998744082697324,
+      "grad_norm": 0.6021164059638977,
+      "learning_rate": 8.834794468113152e-06,
+      "loss": 0.5809,
+      "step": 1552
+    },
+    {
+      "epoch": 0.30006762631629796,
+      "grad_norm": 0.7031016945838928,
+      "learning_rate": 8.832629632825269e-06,
+      "loss": 0.7023,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3002608443628635,
+      "grad_norm": 0.7624228596687317,
+      "learning_rate": 8.83046305413509e-06,
+      "loss": 0.5994,
+      "step": 1554
+    },
+    {
+      "epoch": 0.30045406240942907,
+      "grad_norm": 0.6690539121627808,
+      "learning_rate": 8.828294733028155e-06,
+      "loss": 0.6463,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3006472804559946,
+      "grad_norm": 0.7630172967910767,
+      "learning_rate": 8.826124670490804e-06,
+      "loss": 0.6384,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3008404985025601,
+      "grad_norm": 0.7151029109954834,
+      "learning_rate": 8.823952867510158e-06,
+      "loss": 0.6774,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3010337165491257,
+      "grad_norm": 0.7126525044441223,
+      "learning_rate": 8.82177932507414e-06,
+      "loss": 0.614,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3012269345956912,
+      "grad_norm": 0.7123734354972839,
+      "learning_rate": 8.819604044171457e-06,
+      "loss": 0.7596,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3014201526422568,
+      "grad_norm": 0.68149334192276,
+      "learning_rate": 8.817427025791611e-06,
+      "loss": 0.6427,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3016133706888223,
+      "grad_norm": 0.7297238111495972,
+      "learning_rate": 8.815248270924895e-06,
+      "loss": 0.684,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3018065887353879,
+      "grad_norm": 0.7036851644515991,
+      "learning_rate": 8.813067780562387e-06,
+      "loss": 0.6183,
+      "step": 1562
+    },
+    {
+      "epoch": 0.30199980678195343,
+      "grad_norm": 0.6509476900100708,
+      "learning_rate": 8.810885555695958e-06,
+      "loss": 0.6105,
+      "step": 1563
+    },
+    {
+      "epoch": 0.302193024828519,
+      "grad_norm": 0.649760365486145,
+      "learning_rate": 8.80870159731827e-06,
+      "loss": 0.6054,
+      "step": 1564
+    },
+    {
+      "epoch": 0.30238624287508453,
+      "grad_norm": 0.5940372943878174,
+      "learning_rate": 8.806515906422766e-06,
+      "loss": 0.5962,
+      "step": 1565
+    },
+    {
+      "epoch": 0.30257946092165006,
+      "grad_norm": 0.6401147246360779,
+      "learning_rate": 8.804328484003686e-06,
+      "loss": 0.6433,
+      "step": 1566
+    },
+    {
+      "epoch": 0.30277267896821564,
+      "grad_norm": 0.690054714679718,
+      "learning_rate": 8.802139331056053e-06,
+      "loss": 0.6169,
+      "step": 1567
+    },
+    {
+      "epoch": 0.30296589701478116,
+      "grad_norm": 0.6381725072860718,
+      "learning_rate": 8.799948448575677e-06,
+      "loss": 0.6776,
+      "step": 1568
+    },
+    {
+      "epoch": 0.30315911506134674,
+      "grad_norm": 0.7185710072517395,
+      "learning_rate": 8.797755837559156e-06,
+      "loss": 0.6182,
+      "step": 1569
+    },
+    {
+      "epoch": 0.30335233310791226,
+      "grad_norm": 0.7362060546875,
+      "learning_rate": 8.795561499003873e-06,
+      "loss": 0.7006,
+      "step": 1570
+    },
+    {
+      "epoch": 0.30354555115447784,
+      "grad_norm": 0.6968568563461304,
+      "learning_rate": 8.793365433908e-06,
+      "loss": 0.6412,
+      "step": 1571
+    },
+    {
+      "epoch": 0.30373876920104337,
+      "grad_norm": 0.7330650687217712,
+      "learning_rate": 8.791167643270491e-06,
+      "loss": 0.6858,
+      "step": 1572
+    },
+    {
+      "epoch": 0.30393198724760895,
+      "grad_norm": 0.6777971386909485,
+      "learning_rate": 8.788968128091084e-06,
+      "loss": 0.6744,
+      "step": 1573
+    },
+    {
+      "epoch": 0.30412520529417447,
+      "grad_norm": 0.6727851033210754,
+      "learning_rate": 8.786766889370306e-06,
+      "loss": 0.6268,
+      "step": 1574
+    },
+    {
+      "epoch": 0.30431842334074005,
+      "grad_norm": 0.7013323903083801,
+      "learning_rate": 8.784563928109465e-06,
+      "loss": 0.6769,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3045116413873056,
+      "grad_norm": 0.769282877445221,
+      "learning_rate": 8.782359245310654e-06,
+      "loss": 0.6806,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3047048594338711,
+      "grad_norm": 0.6675457954406738,
+      "learning_rate": 8.780152841976745e-06,
+      "loss": 0.6181,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3048980774804367,
+      "grad_norm": 0.6937217116355896,
+      "learning_rate": 8.777944719111399e-06,
+      "loss": 0.5946,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3050912955270022,
+      "grad_norm": 0.6043604016304016,
+      "learning_rate": 8.775734877719053e-06,
+      "loss": 0.5993,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3052845135735678,
+      "grad_norm": 0.6887547969818115,
+      "learning_rate": 8.773523318804934e-06,
+      "loss": 0.5644,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3054777316201333,
+      "grad_norm": 0.6853593587875366,
+      "learning_rate": 8.771310043375039e-06,
+      "loss": 0.6293,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3056709496666989,
+      "grad_norm": 0.7678703665733337,
+      "learning_rate": 8.769095052436154e-06,
+      "loss": 0.6974,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3058641677132644,
+      "grad_norm": 0.668164074420929,
+      "learning_rate": 8.766878346995845e-06,
+      "loss": 0.6622,
+      "step": 1583
+    },
+    {
+      "epoch": 0.30605738575983,
+      "grad_norm": 0.6898372769355774,
+      "learning_rate": 8.764659928062454e-06,
+      "loss": 0.6203,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3062506038063955,
+      "grad_norm": 0.7377843856811523,
+      "learning_rate": 8.762439796645103e-06,
+      "loss": 0.5925,
+      "step": 1585
+    },
+    {
+      "epoch": 0.30644382185296104,
+      "grad_norm": 0.7535669803619385,
+      "learning_rate": 8.760217953753699e-06,
+      "loss": 0.6556,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3066370398995266,
+      "grad_norm": 0.6516215205192566,
+      "learning_rate": 8.757994400398918e-06,
+      "loss": 0.6439,
+      "step": 1587
+    },
+    {
+      "epoch": 0.30683025794609214,
+      "grad_norm": 0.6754837036132812,
+      "learning_rate": 8.755769137592225e-06,
+      "loss": 0.6464,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3070234759926577,
+      "grad_norm": 0.6840673685073853,
+      "learning_rate": 8.753542166345851e-06,
+      "loss": 0.6852,
+      "step": 1589
+    },
+    {
+      "epoch": 0.30721669403922325,
+      "grad_norm": 0.6201848983764648,
+      "learning_rate": 8.751313487672815e-06,
+      "loss": 0.6145,
+      "step": 1590
+    },
+    {
+      "epoch": 0.30740991208578883,
+      "grad_norm": 0.6625637412071228,
+      "learning_rate": 8.749083102586902e-06,
+      "loss": 0.6531,
+      "step": 1591
+    },
+    {
+      "epoch": 0.30760313013235435,
+      "grad_norm": 0.6314133405685425,
+      "learning_rate": 8.746851012102681e-06,
+      "loss": 0.5731,
+      "step": 1592
+    },
+    {
+      "epoch": 0.30779634817891993,
+      "grad_norm": 0.6859167218208313,
+      "learning_rate": 8.744617217235497e-06,
+      "loss": 0.6319,
+      "step": 1593
+    },
+    {
+      "epoch": 0.30798956622548546,
+      "grad_norm": 0.7420565485954285,
+      "learning_rate": 8.742381719001463e-06,
+      "loss": 0.7021,
+      "step": 1594
+    },
+    {
+      "epoch": 0.30818278427205104,
+      "grad_norm": 0.6536823511123657,
+      "learning_rate": 8.740144518417477e-06,
+      "loss": 0.6557,
+      "step": 1595
+    },
+    {
+      "epoch": 0.30837600231861656,
+      "grad_norm": 0.6870988011360168,
+      "learning_rate": 8.737905616501201e-06,
+      "loss": 0.6269,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3085692203651821,
+      "grad_norm": 0.655680775642395,
+      "learning_rate": 8.73566501427108e-06,
+      "loss": 0.5848,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30876243841174766,
+      "grad_norm": 0.7777279615402222,
+      "learning_rate": 8.733422712746324e-06,
+      "loss": 0.6112,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3089556564583132,
+      "grad_norm": 0.7062700390815735,
+      "learning_rate": 8.731178712946924e-06,
+      "loss": 0.5905,
+      "step": 1599
+    },
+    {
+      "epoch": 0.30914887450487877,
+      "grad_norm": 0.7094499468803406,
+      "learning_rate": 8.728933015893639e-06,
+      "loss": 0.6367,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3093420925514443,
+      "grad_norm": 0.6788246631622314,
+      "learning_rate": 8.726685622608e-06,
+      "loss": 0.6204,
+      "step": 1601
+    },
+    {
+      "epoch": 0.30953531059800987,
+      "grad_norm": 0.6332913637161255,
+      "learning_rate": 8.724436534112311e-06,
+      "loss": 0.602,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3097285286445754,
+      "grad_norm": 0.659821093082428,
+      "learning_rate": 8.722185751429648e-06,
+      "loss": 0.6178,
+      "step": 1603
+    },
+    {
+      "epoch": 0.309921746691141,
+      "grad_norm": 0.6489788889884949,
+      "learning_rate": 8.719933275583851e-06,
+      "loss": 0.5776,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3101149647377065,
+      "grad_norm": 0.6902356743812561,
+      "learning_rate": 8.717679107599544e-06,
+      "loss": 0.6203,
+      "step": 1605
+    },
+    {
+      "epoch": 0.310308182784272,
+      "grad_norm": 0.6306488513946533,
+      "learning_rate": 8.715423248502103e-06,
+      "loss": 0.5899,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3105014008308376,
+      "grad_norm": 0.6839290857315063,
+      "learning_rate": 8.71316569931769e-06,
+      "loss": 0.7106,
+      "step": 1607
+    },
+    {
+      "epoch": 0.31069461887740313,
+      "grad_norm": 0.6429887413978577,
+      "learning_rate": 8.710906461073225e-06,
+      "loss": 0.6596,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3108878369239687,
+      "grad_norm": 0.591609537601471,
+      "learning_rate": 8.7086455347964e-06,
+      "loss": 0.5994,
+      "step": 1609
+    },
+    {
+      "epoch": 0.31108105497053423,
+      "grad_norm": 0.7396703362464905,
+      "learning_rate": 8.706382921515672e-06,
+      "loss": 0.593,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3112742730170998,
+      "grad_norm": 0.7240954041481018,
+      "learning_rate": 8.70411862226027e-06,
+      "loss": 0.599,
+      "step": 1611
+    },
+    {
+      "epoch": 0.31146749106366534,
+      "grad_norm": 0.6710633039474487,
+      "learning_rate": 8.701852638060191e-06,
+      "loss": 0.6443,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3116607091102309,
+      "grad_norm": 0.6782061457633972,
+      "learning_rate": 8.699584969946188e-06,
+      "loss": 0.6025,
+      "step": 1613
+    },
+    {
+      "epoch": 0.31185392715679644,
+      "grad_norm": 0.7152571678161621,
+      "learning_rate": 8.697315618949794e-06,
+      "loss": 0.6437,
+      "step": 1614
+    },
+    {
+      "epoch": 0.312047145203362,
+      "grad_norm": 0.7185957431793213,
+      "learning_rate": 8.695044586103297e-06,
+      "loss": 0.629,
+      "step": 1615
+    },
+    {
+      "epoch": 0.31224036324992754,
+      "grad_norm": 0.7054689526557922,
+      "learning_rate": 8.692771872439753e-06,
+      "loss": 0.6579,
+      "step": 1616
+    },
+    {
+      "epoch": 0.31243358129649307,
+      "grad_norm": 0.674817681312561,
+      "learning_rate": 8.690497478992986e-06,
+      "loss": 0.6169,
+      "step": 1617
+    },
+    {
+      "epoch": 0.31262679934305865,
+      "grad_norm": 0.6700680255889893,
+      "learning_rate": 8.688221406797582e-06,
+      "loss": 0.6291,
+      "step": 1618
+    },
+    {
+      "epoch": 0.31282001738962417,
+      "grad_norm": 0.6509378552436829,
+      "learning_rate": 8.685943656888884e-06,
+      "loss": 0.6195,
+      "step": 1619
+    },
+    {
+      "epoch": 0.31301323543618975,
+      "grad_norm": 0.6852338314056396,
+      "learning_rate": 8.683664230303011e-06,
+      "loss": 0.6213,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3132064534827553,
+      "grad_norm": 0.6458637118339539,
+      "learning_rate": 8.681383128076836e-06,
+      "loss": 0.6465,
+      "step": 1621
+    },
+    {
+      "epoch": 0.31339967152932086,
+      "grad_norm": 0.6213224530220032,
+      "learning_rate": 8.679100351247995e-06,
+      "loss": 0.5701,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3135928895758864,
+      "grad_norm": 0.6640623807907104,
+      "learning_rate": 8.676815900854887e-06,
+      "loss": 0.5552,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31378610762245196,
+      "grad_norm": 0.6908933520317078,
+      "learning_rate": 8.674529777936674e-06,
+      "loss": 0.644,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3139793256690175,
+      "grad_norm": 0.6624265909194946,
+      "learning_rate": 8.672241983533274e-06,
+      "loss": 0.6912,
+      "step": 1625
+    },
+    {
+      "epoch": 0.314172543715583,
+      "grad_norm": 0.7143158316612244,
+      "learning_rate": 8.669952518685372e-06,
+      "loss": 0.656,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3143657617621486,
+      "grad_norm": 0.6743907928466797,
+      "learning_rate": 8.667661384434406e-06,
+      "loss": 0.6605,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3145589798087141,
+      "grad_norm": 0.6735371351242065,
+      "learning_rate": 8.665368581822577e-06,
+      "loss": 0.6071,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3147521978552797,
+      "grad_norm": 0.7238439917564392,
+      "learning_rate": 8.663074111892845e-06,
+      "loss": 0.6664,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3149454159018452,
+      "grad_norm": 0.6655930280685425,
+      "learning_rate": 8.66077797568893e-06,
+      "loss": 0.5997,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3151386339484108,
+      "grad_norm": 0.6962398886680603,
+      "learning_rate": 8.658480174255303e-06,
+      "loss": 0.6003,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3153318519949763,
+      "grad_norm": 0.6285104155540466,
+      "learning_rate": 8.656180708637202e-06,
+      "loss": 0.629,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3155250700415419,
+      "grad_norm": 0.677779495716095,
+      "learning_rate": 8.653879579880615e-06,
+      "loss": 0.7072,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3157182880881074,
+      "grad_norm": 0.6885689496994019,
+      "learning_rate": 8.65157678903229e-06,
+      "loss": 0.6157,
+      "step": 1634
+    },
+    {
+      "epoch": 0.315911506134673,
+      "grad_norm": 0.6435257792472839,
+      "learning_rate": 8.649272337139729e-06,
+      "loss": 0.5689,
+      "step": 1635
+    },
+    {
+      "epoch": 0.31610472418123853,
+      "grad_norm": 0.6048800349235535,
+      "learning_rate": 8.646966225251191e-06,
+      "loss": 0.6041,
+      "step": 1636
+    },
+    {
+      "epoch": 0.31629794222780405,
+      "grad_norm": 0.6844281554222107,
+      "learning_rate": 8.644658454415689e-06,
+      "loss": 0.6623,
+      "step": 1637
+    },
+    {
+      "epoch": 0.31649116027436963,
+      "grad_norm": 0.6471332907676697,
+      "learning_rate": 8.642349025682992e-06,
+      "loss": 0.6008,
+      "step": 1638
+    },
+    {
+      "epoch": 0.31668437832093516,
+      "grad_norm": 0.6771491169929504,
+      "learning_rate": 8.640037940103625e-06,
+      "loss": 0.6125,
+      "step": 1639
+    },
+    {
+      "epoch": 0.31687759636750074,
+      "grad_norm": 0.6857588291168213,
+      "learning_rate": 8.637725198728858e-06,
+      "loss": 0.6605,
+      "step": 1640
+    },
+    {
+      "epoch": 0.31707081441406626,
+      "grad_norm": 0.6792883276939392,
+      "learning_rate": 8.635410802610724e-06,
+      "loss": 0.6936,
+      "step": 1641
+    },
+    {
+      "epoch": 0.31726403246063184,
+      "grad_norm": 0.7090012431144714,
+      "learning_rate": 8.633094752802006e-06,
+      "loss": 0.6172,
+      "step": 1642
+    },
+    {
+      "epoch": 0.31745725050719736,
+      "grad_norm": 0.7227849364280701,
+      "learning_rate": 8.630777050356237e-06,
+      "loss": 0.6407,
+      "step": 1643
+    },
+    {
+      "epoch": 0.31765046855376294,
+      "grad_norm": 0.6651045083999634,
+      "learning_rate": 8.628457696327703e-06,
+      "loss": 0.6093,
+      "step": 1644
+    },
+    {
+      "epoch": 0.31784368660032847,
+      "grad_norm": 0.6534207463264465,
+      "learning_rate": 8.62613669177144e-06,
+      "loss": 0.6055,
+      "step": 1645
+    },
+    {
+      "epoch": 0.31803690464689405,
+      "grad_norm": 0.6606358885765076,
+      "learning_rate": 8.623814037743236e-06,
+      "loss": 0.6176,
+      "step": 1646
+    },
+    {
+      "epoch": 0.31823012269345957,
+      "grad_norm": 0.6786819696426392,
+      "learning_rate": 8.621489735299634e-06,
+      "loss": 0.6967,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3184233407400251,
+      "grad_norm": 0.7081559300422668,
+      "learning_rate": 8.619163785497917e-06,
+      "loss": 0.6735,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3186165587865907,
+      "grad_norm": 0.6579135656356812,
+      "learning_rate": 8.616836189396125e-06,
+      "loss": 0.655,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3188097768331562,
+      "grad_norm": 0.645850419998169,
+      "learning_rate": 8.614506948053044e-06,
+      "loss": 0.601,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3190029948797218,
+      "grad_norm": 0.6575025916099548,
+      "learning_rate": 8.612176062528207e-06,
+      "loss": 0.6086,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3191962129262873,
+      "grad_norm": 0.7101923227310181,
+      "learning_rate": 8.609843533881901e-06,
+      "loss": 0.5948,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3193894309728529,
+      "grad_norm": 0.7055867910385132,
+      "learning_rate": 8.607509363175154e-06,
+      "loss": 0.6244,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3195826490194184,
+      "grad_norm": 0.599667489528656,
+      "learning_rate": 8.605173551469741e-06,
+      "loss": 0.5747,
+      "step": 1654
+    },
+    {
+      "epoch": 0.319775867065984,
+      "grad_norm": 0.6390600204467773,
+      "learning_rate": 8.60283609982819e-06,
+      "loss": 0.604,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3199690851125495,
+      "grad_norm": 0.6360976099967957,
+      "learning_rate": 8.600497009313766e-06,
+      "loss": 0.6192,
+      "step": 1656
+    },
+    {
+      "epoch": 0.32016230315911504,
+      "grad_norm": 0.6547744274139404,
+      "learning_rate": 8.598156280990486e-06,
+      "loss": 0.6256,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3203555212056806,
+      "grad_norm": 0.6491039395332336,
+      "learning_rate": 8.595813915923113e-06,
+      "loss": 0.6058,
+      "step": 1658
+    },
+    {
+      "epoch": 0.32054873925224614,
+      "grad_norm": 0.6301624178886414,
+      "learning_rate": 8.593469915177146e-06,
+      "loss": 0.5703,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3207419572988117,
+      "grad_norm": 0.7351461052894592,
+      "learning_rate": 8.591124279818837e-06,
+      "loss": 0.6169,
+      "step": 1660
+    },
+    {
+      "epoch": 0.32093517534537724,
+      "grad_norm": 0.669195294380188,
+      "learning_rate": 8.588777010915179e-06,
+      "loss": 0.5923,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3211283933919428,
+      "grad_norm": 0.7271985411643982,
+      "learning_rate": 8.586428109533906e-06,
+      "loss": 0.6239,
+      "step": 1662
+    },
+    {
+      "epoch": 0.32132161143850835,
+      "grad_norm": 0.6417050957679749,
+      "learning_rate": 8.584077576743497e-06,
+      "loss": 0.5477,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3215148294850739,
+      "grad_norm": 0.7702397704124451,
+      "learning_rate": 8.58172541361317e-06,
+      "loss": 0.7218,
+      "step": 1664
+    },
+    {
+      "epoch": 0.32170804753163945,
+      "grad_norm": 0.7248337268829346,
+      "learning_rate": 8.579371621212888e-06,
+      "loss": 0.6294,
+      "step": 1665
+    },
+    {
+      "epoch": 0.32190126557820503,
+      "grad_norm": 0.6334032416343689,
+      "learning_rate": 8.577016200613353e-06,
+      "loss": 0.6569,
+      "step": 1666
+    },
+    {
+      "epoch": 0.32209448362477056,
+      "grad_norm": 0.6605886816978455,
+      "learning_rate": 8.574659152886009e-06,
+      "loss": 0.6046,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3222877016713361,
+      "grad_norm": 0.7508733868598938,
+      "learning_rate": 8.57230047910304e-06,
+      "loss": 0.6324,
+      "step": 1668
+    },
+    {
+      "epoch": 0.32248091971790166,
+      "grad_norm": 0.6199350953102112,
+      "learning_rate": 8.569940180337369e-06,
+      "loss": 0.5883,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3226741377644672,
+      "grad_norm": 0.6653841137886047,
+      "learning_rate": 8.567578257662658e-06,
+      "loss": 0.6288,
+      "step": 1670
+    },
+    {
+      "epoch": 0.32286735581103276,
+      "grad_norm": 0.7516412138938904,
+      "learning_rate": 8.565214712153307e-06,
+      "loss": 0.6548,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3230605738575983,
+      "grad_norm": 0.6498536467552185,
+      "learning_rate": 8.562849544884458e-06,
+      "loss": 0.6136,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32325379190416387,
+      "grad_norm": 0.7005547881126404,
+      "learning_rate": 8.560482756931984e-06,
+      "loss": 0.5719,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3234470099507294,
+      "grad_norm": 0.6913560032844543,
+      "learning_rate": 8.558114349372502e-06,
+      "loss": 0.6864,
+      "step": 1674
+    },
+    {
+      "epoch": 0.32364022799729497,
+      "grad_norm": 0.6251383423805237,
+      "learning_rate": 8.555744323283364e-06,
+      "loss": 0.6314,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3238334460438605,
+      "grad_norm": 0.7038425803184509,
+      "learning_rate": 8.553372679742655e-06,
+      "loss": 0.6355,
+      "step": 1676
+    },
+    {
+      "epoch": 0.324026664090426,
+      "grad_norm": 0.6905673742294312,
+      "learning_rate": 8.5509994198292e-06,
+      "loss": 0.6406,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3242198821369916,
+      "grad_norm": 0.6522449254989624,
+      "learning_rate": 8.548624544622553e-06,
+      "loss": 0.6314,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3244131001835571,
+      "grad_norm": 0.7171085476875305,
+      "learning_rate": 8.546248055203012e-06,
+      "loss": 0.6287,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3246063182301227,
+      "grad_norm": 0.7585219144821167,
+      "learning_rate": 8.543869952651603e-06,
+      "loss": 0.633,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3247995362766882,
+      "grad_norm": 0.628213107585907,
+      "learning_rate": 8.541490238050082e-06,
+      "loss": 0.6374,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3249927543232538,
+      "grad_norm": 0.6575116515159607,
+      "learning_rate": 8.539108912480951e-06,
+      "loss": 0.6147,
+      "step": 1682
+    },
+    {
+      "epoch": 0.32518597236981933,
+      "grad_norm": 0.7645049691200256,
+      "learning_rate": 8.536725977027433e-06,
+      "loss": 0.648,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3253791904163849,
+      "grad_norm": 0.7149015069007874,
+      "learning_rate": 8.534341432773488e-06,
+      "loss": 0.6193,
+      "step": 1684
+    },
+    {
+      "epoch": 0.32557240846295044,
+      "grad_norm": 0.7104012370109558,
+      "learning_rate": 8.531955280803808e-06,
+      "loss": 0.6323,
+      "step": 1685
+    },
+    {
+      "epoch": 0.325765626509516,
+      "grad_norm": 0.65350341796875,
+      "learning_rate": 8.529567522203815e-06,
+      "loss": 0.6123,
+      "step": 1686
+    },
+    {
+      "epoch": 0.32595884455608154,
+      "grad_norm": 0.697105884552002,
+      "learning_rate": 8.527178158059662e-06,
+      "loss": 0.673,
+      "step": 1687
+    },
+    {
+      "epoch": 0.32615206260264706,
+      "grad_norm": 0.6544195413589478,
+      "learning_rate": 8.524787189458232e-06,
+      "loss": 0.6296,
+      "step": 1688
+    },
+    {
+      "epoch": 0.32634528064921264,
+      "grad_norm": 0.5822769403457642,
+      "learning_rate": 8.522394617487142e-06,
+      "loss": 0.6029,
+      "step": 1689
+    },
+    {
+      "epoch": 0.32653849869577817,
+      "grad_norm": 0.6164411902427673,
+      "learning_rate": 8.520000443234731e-06,
+      "loss": 0.6116,
+      "step": 1690
+    },
+    {
+      "epoch": 0.32673171674234375,
+      "grad_norm": 0.6944220066070557,
+      "learning_rate": 8.517604667790073e-06,
+      "loss": 0.6361,
+      "step": 1691
+    },
+    {
+      "epoch": 0.32692493478890927,
+      "grad_norm": 0.6626195311546326,
+      "learning_rate": 8.515207292242969e-06,
+      "loss": 0.617,
+      "step": 1692
+    },
+    {
+      "epoch": 0.32711815283547485,
+      "grad_norm": 0.6486663222312927,
+      "learning_rate": 8.512808317683945e-06,
+      "loss": 0.6525,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3273113708820404,
+      "grad_norm": 0.6353569626808167,
+      "learning_rate": 8.510407745204257e-06,
+      "loss": 0.5676,
+      "step": 1694
+    },
+    {
+      "epoch": 0.32750458892860596,
+      "grad_norm": 0.697246789932251,
+      "learning_rate": 8.508005575895887e-06,
+      "loss": 0.6052,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3276978069751715,
+      "grad_norm": 0.64223712682724,
+      "learning_rate": 8.505601810851543e-06,
+      "loss": 0.5828,
+      "step": 1696
+    },
+    {
+      "epoch": 0.327891025021737,
+      "grad_norm": 0.6822214722633362,
+      "learning_rate": 8.503196451164663e-06,
+      "loss": 0.6437,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3280842430683026,
+      "grad_norm": 0.6183921694755554,
+      "learning_rate": 8.500789497929404e-06,
+      "loss": 0.599,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3282774611148681,
+      "grad_norm": 0.6892261505126953,
+      "learning_rate": 8.49838095224065e-06,
+      "loss": 0.6677,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3284706791614337,
+      "grad_norm": 0.6412778496742249,
+      "learning_rate": 8.495970815194013e-06,
+      "loss": 0.6908,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3286638972079992,
+      "grad_norm": 0.6095704436302185,
+      "learning_rate": 8.493559087885822e-06,
+      "loss": 0.5869,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3288571152545648,
+      "grad_norm": 0.6210219860076904,
+      "learning_rate": 8.491145771413138e-06,
+      "loss": 0.5682,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3290503333011303,
+      "grad_norm": 0.6679145097732544,
+      "learning_rate": 8.488730866873738e-06,
+      "loss": 0.6089,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3292435513476959,
+      "grad_norm": 0.6045292019844055,
+      "learning_rate": 8.486314375366122e-06,
+      "loss": 0.5677,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3294367693942614,
+      "grad_norm": 0.7063819169998169,
+      "learning_rate": 8.483896297989516e-06,
+      "loss": 0.7326,
+      "step": 1705
+    },
+    {
+      "epoch": 0.329629987440827,
+      "grad_norm": 0.6415681838989258,
+      "learning_rate": 8.481476635843867e-06,
+      "loss": 0.7093,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3298232054873925,
+      "grad_norm": 0.6587389707565308,
+      "learning_rate": 8.479055390029834e-06,
+      "loss": 0.6231,
+      "step": 1707
+    },
+    {
+      "epoch": 0.33001642353395805,
+      "grad_norm": 0.5838842988014221,
+      "learning_rate": 8.47663256164881e-06,
+      "loss": 0.5964,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3302096415805236,
+      "grad_norm": 0.6646946668624878,
+      "learning_rate": 8.474208151802898e-06,
+      "loss": 0.5995,
+      "step": 1709
+    },
+    {
+      "epoch": 0.33040285962708915,
+      "grad_norm": 0.659927248954773,
+      "learning_rate": 8.471782161594923e-06,
+      "loss": 0.6595,
+      "step": 1710
+    },
+    {
+      "epoch": 0.33059607767365473,
+      "grad_norm": 0.6634847521781921,
+      "learning_rate": 8.469354592128432e-06,
+      "loss": 0.6609,
+      "step": 1711
+    },
+    {
+      "epoch": 0.33078929572022026,
+      "grad_norm": 0.6531798839569092,
+      "learning_rate": 8.466925444507685e-06,
+      "loss": 0.6263,
+      "step": 1712
+    },
+    {
+      "epoch": 0.33098251376678584,
+      "grad_norm": 0.642352819442749,
+      "learning_rate": 8.464494719837666e-06,
+      "loss": 0.634,
+      "step": 1713
+    },
+    {
+      "epoch": 0.33117573181335136,
+      "grad_norm": 0.7995596528053284,
+      "learning_rate": 8.462062419224069e-06,
+      "loss": 0.6861,
+      "step": 1714
+    },
+    {
+      "epoch": 0.33136894985991694,
+      "grad_norm": 0.6108599305152893,
+      "learning_rate": 8.45962854377331e-06,
+      "loss": 0.5484,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33156216790648246,
+      "grad_norm": 0.6254993677139282,
+      "learning_rate": 8.457193094592521e-06,
+      "loss": 0.5704,
+      "step": 1716
+    },
+    {
+      "epoch": 0.331755385953048,
+      "grad_norm": 0.7216454148292542,
+      "learning_rate": 8.454756072789551e-06,
+      "loss": 0.6281,
+      "step": 1717
+    },
+    {
+      "epoch": 0.33194860399961357,
+      "grad_norm": 0.7719717025756836,
+      "learning_rate": 8.45231747947296e-06,
+      "loss": 0.6447,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3321418220461791,
+      "grad_norm": 0.6733083724975586,
+      "learning_rate": 8.449877315752025e-06,
+      "loss": 0.6175,
+      "step": 1719
+    },
+    {
+      "epoch": 0.33233504009274467,
+      "grad_norm": 0.69261634349823,
+      "learning_rate": 8.447435582736739e-06,
+      "loss": 0.6469,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3325282581393102,
+      "grad_norm": 0.7032318711280823,
+      "learning_rate": 8.444992281537807e-06,
+      "loss": 0.6317,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3327214761858758,
+      "grad_norm": 0.6324744820594788,
+      "learning_rate": 8.442547413266648e-06,
+      "loss": 0.5907,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3329146942324413,
+      "grad_norm": 0.6593074202537537,
+      "learning_rate": 8.440100979035394e-06,
+      "loss": 0.6788,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3331079122790069,
+      "grad_norm": 0.7253870964050293,
+      "learning_rate": 8.437652979956888e-06,
+      "loss": 0.6507,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3333011303255724,
+      "grad_norm": 0.6776022911071777,
+      "learning_rate": 8.435203417144689e-06,
+      "loss": 0.6163,
+      "step": 1725
+    },
+    {
+      "epoch": 0.333494348372138,
+      "grad_norm": 0.7141279578208923,
+      "learning_rate": 8.432752291713058e-06,
+      "loss": 0.6072,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3336875664187035,
+      "grad_norm": 0.6799734234809875,
+      "learning_rate": 8.43029960477698e-06,
+      "loss": 0.603,
+      "step": 1727
+    },
+    {
+      "epoch": 0.33388078446526903,
+      "grad_norm": 0.6963182687759399,
+      "learning_rate": 8.427845357452138e-06,
+      "loss": 0.5847,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3340740025118346,
+      "grad_norm": 0.632631242275238,
+      "learning_rate": 8.425389550854935e-06,
+      "loss": 0.5954,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33426722055840014,
+      "grad_norm": 0.6513384580612183,
+      "learning_rate": 8.422932186102472e-06,
+      "loss": 0.6057,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3344604386049657,
+      "grad_norm": 0.7083600759506226,
+      "learning_rate": 8.42047326431257e-06,
+      "loss": 0.7149,
+      "step": 1731
+    },
+    {
+      "epoch": 0.33465365665153124,
+      "grad_norm": 0.6206734776496887,
+      "learning_rate": 8.418012786603753e-06,
+      "loss": 0.5826,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3348468746980968,
+      "grad_norm": 0.67019122838974,
+      "learning_rate": 8.415550754095252e-06,
+      "loss": 0.6028,
+      "step": 1733
+    },
+    {
+      "epoch": 0.33504009274466234,
+      "grad_norm": 0.7165263891220093,
+      "learning_rate": 8.413087167907008e-06,
+      "loss": 0.6415,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3352333107912279,
+      "grad_norm": 0.6728923916816711,
+      "learning_rate": 8.410622029159664e-06,
+      "loss": 0.6607,
+      "step": 1735
+    },
+    {
+      "epoch": 0.33542652883779345,
+      "grad_norm": 0.6933238506317139,
+      "learning_rate": 8.408155338974575e-06,
+      "loss": 0.6443,
+      "step": 1736
+    },
+    {
+      "epoch": 0.33561974688435897,
+      "grad_norm": 0.6264029145240784,
+      "learning_rate": 8.4056870984738e-06,
+      "loss": 0.6861,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33581296493092455,
+      "grad_norm": 0.6758835911750793,
+      "learning_rate": 8.403217308780099e-06,
+      "loss": 0.6238,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3360061829774901,
+      "grad_norm": 0.6778144240379333,
+      "learning_rate": 8.400745971016941e-06,
+      "loss": 0.6266,
+      "step": 1739
+    },
+    {
+      "epoch": 0.33619940102405566,
+      "grad_norm": 0.7014786005020142,
+      "learning_rate": 8.398273086308503e-06,
+      "loss": 0.6663,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3363926190706212,
+      "grad_norm": 0.6527827382087708,
+      "learning_rate": 8.395798655779655e-06,
+      "loss": 0.5718,
+      "step": 1741
+    },
+    {
+      "epoch": 0.33658583711718676,
+      "grad_norm": 0.6306667923927307,
+      "learning_rate": 8.393322680555979e-06,
+      "loss": 0.617,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3367790551637523,
+      "grad_norm": 0.6963008046150208,
+      "learning_rate": 8.390845161763756e-06,
+      "loss": 0.6421,
+      "step": 1743
+    },
+    {
+      "epoch": 0.33697227321031786,
+      "grad_norm": 0.7127588987350464,
+      "learning_rate": 8.38836610052997e-06,
+      "loss": 0.621,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3371654912568834,
+      "grad_norm": 0.7142772078514099,
+      "learning_rate": 8.385885497982307e-06,
+      "loss": 0.6576,
+      "step": 1745
+    },
+    {
+      "epoch": 0.33735870930344897,
+      "grad_norm": 0.688413679599762,
+      "learning_rate": 8.383403355249152e-06,
+      "loss": 0.6439,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3375519273500145,
+      "grad_norm": 0.6886526346206665,
+      "learning_rate": 8.380919673459594e-06,
+      "loss": 0.6036,
+      "step": 1747
+    },
+    {
+      "epoch": 0.33774514539658,
+      "grad_norm": 0.6078207492828369,
+      "learning_rate": 8.378434453743421e-06,
+      "loss": 0.678,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3379383634431456,
+      "grad_norm": 0.721482515335083,
+      "learning_rate": 8.375947697231117e-06,
+      "loss": 0.6129,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3381315814897111,
+      "grad_norm": 0.6769798398017883,
+      "learning_rate": 8.373459405053873e-06,
+      "loss": 0.6096,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3383247995362767,
+      "grad_norm": 0.6590465307235718,
+      "learning_rate": 8.370969578343568e-06,
+      "loss": 0.6107,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3385180175828422,
+      "grad_norm": 0.674583911895752,
+      "learning_rate": 8.368478218232787e-06,
+      "loss": 0.5699,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3387112356294078,
+      "grad_norm": 0.71125727891922,
+      "learning_rate": 8.365985325854811e-06,
+      "loss": 0.5894,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3389044536759733,
+      "grad_norm": 0.7267223000526428,
+      "learning_rate": 8.363490902343616e-06,
+      "loss": 0.6291,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3390976717225389,
+      "grad_norm": 0.7898128628730774,
+      "learning_rate": 8.360994948833875e-06,
+      "loss": 0.6425,
+      "step": 1755
+    },
+    {
+      "epoch": 0.33929088976910443,
+      "grad_norm": 0.6875765919685364,
+      "learning_rate": 8.35849746646096e-06,
+      "loss": 0.5632,
+      "step": 1756
+    },
+    {
+      "epoch": 0.33948410781567,
+      "grad_norm": 0.6858415603637695,
+      "learning_rate": 8.355998456360932e-06,
+      "loss": 0.5915,
+      "step": 1757
+    },
+    {
+      "epoch": 0.33967732586223554,
+      "grad_norm": 0.6635237336158752,
+      "learning_rate": 8.353497919670557e-06,
+      "loss": 0.6417,
+      "step": 1758
+    },
+    {
+      "epoch": 0.33987054390880106,
+      "grad_norm": 0.917772114276886,
+      "learning_rate": 8.350995857527282e-06,
+      "loss": 0.6249,
+      "step": 1759
+    },
+    {
+      "epoch": 0.34006376195536664,
+      "grad_norm": 0.6510213017463684,
+      "learning_rate": 8.34849227106926e-06,
+      "loss": 0.6462,
+      "step": 1760
+    },
+    {
+      "epoch": 0.34025698000193216,
+      "grad_norm": 0.7026508450508118,
+      "learning_rate": 8.345987161435334e-06,
+      "loss": 0.6552,
+      "step": 1761
+    },
+    {
+      "epoch": 0.34045019804849774,
+      "grad_norm": 0.6677646636962891,
+      "learning_rate": 8.343480529765035e-06,
+      "loss": 0.6178,
+      "step": 1762
+    },
+    {
+      "epoch": 0.34064341609506327,
+      "grad_norm": 0.6568297147750854,
+      "learning_rate": 8.340972377198591e-06,
+      "loss": 0.5747,
+      "step": 1763
+    },
+    {
+      "epoch": 0.34083663414162885,
+      "grad_norm": 0.7258191704750061,
+      "learning_rate": 8.338462704876921e-06,
+      "loss": 0.6275,
+      "step": 1764
+    },
+    {
+      "epoch": 0.34102985218819437,
+      "grad_norm": 0.703335165977478,
+      "learning_rate": 8.335951513941634e-06,
+      "loss": 0.6013,
+      "step": 1765
+    },
+    {
+      "epoch": 0.34122307023475995,
+      "grad_norm": 0.5971309542655945,
+      "learning_rate": 8.333438805535034e-06,
+      "loss": 0.5947,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3414162882813255,
+      "grad_norm": 0.7022266387939453,
+      "learning_rate": 8.330924580800108e-06,
+      "loss": 0.5896,
+      "step": 1767
+    },
+    {
+      "epoch": 0.341609506327891,
+      "grad_norm": 0.7429389953613281,
+      "learning_rate": 8.32840884088054e-06,
+      "loss": 0.615,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3418027243744566,
+      "grad_norm": 0.6553027033805847,
+      "learning_rate": 8.3258915869207e-06,
+      "loss": 0.6094,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3419959424210221,
+      "grad_norm": 0.615128219127655,
+      "learning_rate": 8.323372820065643e-06,
+      "loss": 0.5733,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3421891604675877,
+      "grad_norm": 0.6827779412269592,
+      "learning_rate": 8.320852541461121e-06,
+      "loss": 0.6095,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3423823785141532,
+      "grad_norm": 0.6371614933013916,
+      "learning_rate": 8.318330752253565e-06,
+      "loss": 0.5669,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3425755965607188,
+      "grad_norm": 0.5939729809761047,
+      "learning_rate": 8.315807453590099e-06,
+      "loss": 0.5388,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3427688146072843,
+      "grad_norm": 0.6204526424407959,
+      "learning_rate": 8.313282646618531e-06,
+      "loss": 0.5954,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3429620326538499,
+      "grad_norm": 0.7201001048088074,
+      "learning_rate": 8.310756332487355e-06,
+      "loss": 0.7006,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3431552507004154,
+      "grad_norm": 0.6835174560546875,
+      "learning_rate": 8.30822851234575e-06,
+      "loss": 0.5944,
+      "step": 1776
+    },
+    {
+      "epoch": 0.343348468746981,
+      "grad_norm": 0.6286168098449707,
+      "learning_rate": 8.305699187343586e-06,
+      "loss": 0.5583,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3435416867935465,
+      "grad_norm": 0.6395207643508911,
+      "learning_rate": 8.303168358631406e-06,
+      "loss": 0.6288,
+      "step": 1778
+    },
+    {
+      "epoch": 0.34373490484011204,
+      "grad_norm": 0.6310188174247742,
+      "learning_rate": 8.30063602736045e-06,
+      "loss": 0.6502,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3439281228866776,
+      "grad_norm": 0.5993399620056152,
+      "learning_rate": 8.298102194682628e-06,
+      "loss": 0.5805,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34412134093324315,
+      "grad_norm": 0.6772288680076599,
+      "learning_rate": 8.295566861750547e-06,
+      "loss": 0.6048,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3443145589798087,
+      "grad_norm": 0.6895396113395691,
+      "learning_rate": 8.293030029717487e-06,
+      "loss": 0.5671,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34450777702637425,
+      "grad_norm": 0.7514630556106567,
+      "learning_rate": 8.290491699737416e-06,
+      "loss": 0.6331,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34470099507293983,
+      "grad_norm": 0.6433349847793579,
+      "learning_rate": 8.287951872964973e-06,
+      "loss": 0.5809,
+      "step": 1784
+    },
+    {
+      "epoch": 0.34489421311950536,
+      "grad_norm": 0.6493355631828308,
+      "learning_rate": 8.285410550555487e-06,
+      "loss": 0.6508,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34508743116607093,
+      "grad_norm": 0.6496013402938843,
+      "learning_rate": 8.28286773366497e-06,
+      "loss": 0.6155,
+      "step": 1786
+    },
+    {
+      "epoch": 0.34528064921263646,
+      "grad_norm": 0.6888113021850586,
+      "learning_rate": 8.280323423450106e-06,
+      "loss": 0.6685,
+      "step": 1787
+    },
+    {
+      "epoch": 0.345473867259202,
+      "grad_norm": 0.6624513268470764,
+      "learning_rate": 8.277777621068258e-06,
+      "loss": 0.5875,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34566708530576756,
+      "grad_norm": 0.6122423410415649,
+      "learning_rate": 8.275230327677476e-06,
+      "loss": 0.5965,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3458603033523331,
+      "grad_norm": 0.6209532618522644,
+      "learning_rate": 8.27268154443648e-06,
+      "loss": 0.6048,
+      "step": 1790
+    },
+    {
+      "epoch": 0.34605352139889867,
+      "grad_norm": 0.7249213457107544,
+      "learning_rate": 8.270131272504674e-06,
+      "loss": 0.6107,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3462467394454642,
+      "grad_norm": 0.7285491824150085,
+      "learning_rate": 8.267579513042133e-06,
+      "loss": 0.6402,
+      "step": 1792
+    },
+    {
+      "epoch": 0.34643995749202977,
+      "grad_norm": 0.6942946314811707,
+      "learning_rate": 8.265026267209615e-06,
+      "loss": 0.5996,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3466331755385953,
+      "grad_norm": 0.685630202293396,
+      "learning_rate": 8.262471536168547e-06,
+      "loss": 0.6563,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3468263935851609,
+      "grad_norm": 0.7015906572341919,
+      "learning_rate": 8.259915321081038e-06,
+      "loss": 0.6132,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3470196116317264,
+      "grad_norm": 0.7102599143981934,
+      "learning_rate": 8.25735762310987e-06,
+      "loss": 0.6267,
+      "step": 1796
+    },
+    {
+      "epoch": 0.347212829678292,
+      "grad_norm": 0.6604264974594116,
+      "learning_rate": 8.2547984434185e-06,
+      "loss": 0.6367,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3474060477248575,
+      "grad_norm": 0.6004226207733154,
+      "learning_rate": 8.252237783171054e-06,
+      "loss": 0.6665,
+      "step": 1798
+    },
+    {
+      "epoch": 0.347599265771423,
+      "grad_norm": 0.6511847972869873,
+      "learning_rate": 8.24967564353234e-06,
+      "loss": 0.6261,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3477924838179886,
+      "grad_norm": 0.6915515661239624,
+      "learning_rate": 8.247112025667831e-06,
+      "loss": 0.6296,
+      "step": 1800
+    },
+    {
+      "epoch": 0.34798570186455413,
+      "grad_norm": 0.6693213582038879,
+      "learning_rate": 8.244546930743678e-06,
+      "loss": 0.6368,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3481789199111197,
+      "grad_norm": 0.644589900970459,
+      "learning_rate": 8.241980359926704e-06,
+      "loss": 0.5972,
+      "step": 1802
+    },
+    {
+      "epoch": 0.34837213795768524,
+      "grad_norm": 0.6316778063774109,
+      "learning_rate": 8.239412314384397e-06,
+      "loss": 0.6079,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3485653560042508,
+      "grad_norm": 0.7224228978157043,
+      "learning_rate": 8.236842795284923e-06,
+      "loss": 0.6037,
+      "step": 1804
+    },
+    {
+      "epoch": 0.34875857405081634,
+      "grad_norm": 0.6215246319770813,
+      "learning_rate": 8.234271803797113e-06,
+      "loss": 0.634,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3489517920973819,
+      "grad_norm": 0.616243839263916,
+      "learning_rate": 8.231699341090474e-06,
+      "loss": 0.5666,
+      "step": 1806
+    },
+    {
+      "epoch": 0.34914501014394744,
+      "grad_norm": 0.681548535823822,
+      "learning_rate": 8.229125408335176e-06,
+      "loss": 0.6033,
+      "step": 1807
+    },
+    {
+      "epoch": 0.34933822819051297,
+      "grad_norm": 0.7282174825668335,
+      "learning_rate": 8.226550006702058e-06,
+      "loss": 0.6687,
+      "step": 1808
+    },
+    {
+      "epoch": 0.34953144623707855,
+      "grad_norm": 0.625798225402832,
+      "learning_rate": 8.223973137362632e-06,
+      "loss": 0.5941,
+      "step": 1809
+    },
+    {
+      "epoch": 0.34972466428364407,
+      "grad_norm": 0.6687496304512024,
+      "learning_rate": 8.221394801489076e-06,
+      "loss": 0.645,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34991788233020965,
+      "grad_norm": 0.6932955980300903,
+      "learning_rate": 8.218815000254233e-06,
+      "loss": 0.5947,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3501111003767752,
+      "grad_norm": 0.7315152883529663,
+      "learning_rate": 8.216233734831612e-06,
+      "loss": 0.6439,
+      "step": 1812
+    },
+    {
+      "epoch": 0.35030431842334075,
+      "grad_norm": 0.6480579376220703,
+      "learning_rate": 8.213651006395388e-06,
+      "loss": 0.5651,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3504975364699063,
+      "grad_norm": 0.5891026258468628,
+      "learning_rate": 8.211066816120407e-06,
+      "loss": 0.568,
+      "step": 1814
+    },
+    {
+      "epoch": 0.35069075451647186,
+      "grad_norm": 0.6964541077613831,
+      "learning_rate": 8.208481165182176e-06,
+      "loss": 0.6583,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3508839725630374,
+      "grad_norm": 0.7279807925224304,
+      "learning_rate": 8.205894054756864e-06,
+      "loss": 0.6463,
+      "step": 1816
+    },
+    {
+      "epoch": 0.35107719060960296,
+      "grad_norm": 0.6698747277259827,
+      "learning_rate": 8.203305486021308e-06,
+      "loss": 0.6485,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3512704086561685,
+      "grad_norm": 0.6235692501068115,
+      "learning_rate": 8.200715460153009e-06,
+      "loss": 0.6377,
+      "step": 1818
+    },
+    {
+      "epoch": 0.351463626702734,
+      "grad_norm": 0.5943323373794556,
+      "learning_rate": 8.198123978330123e-06,
+      "loss": 0.5974,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3516568447492996,
+      "grad_norm": 0.6290449500083923,
+      "learning_rate": 8.195531041731476e-06,
+      "loss": 0.6256,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3518500627958651,
+      "grad_norm": 0.6505613327026367,
+      "learning_rate": 8.192936651536556e-06,
+      "loss": 0.5937,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3520432808424307,
+      "grad_norm": 0.6949810981750488,
+      "learning_rate": 8.190340808925509e-06,
+      "loss": 0.6503,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3522364988889962,
+      "grad_norm": 0.6922764182090759,
+      "learning_rate": 8.18774351507914e-06,
+      "loss": 0.558,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3524297169355618,
+      "grad_norm": 0.7102041244506836,
+      "learning_rate": 8.185144771178922e-06,
+      "loss": 0.6285,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3526229349821273,
+      "grad_norm": 0.6893911957740784,
+      "learning_rate": 8.182544578406977e-06,
+      "loss": 0.6522,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3528161530286929,
+      "grad_norm": 0.61988365650177,
+      "learning_rate": 8.179942937946092e-06,
+      "loss": 0.5655,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3530093710752584,
+      "grad_norm": 0.7407771348953247,
+      "learning_rate": 8.177339850979715e-06,
+      "loss": 0.6676,
+      "step": 1827
+    },
+    {
+      "epoch": 0.35320258912182395,
+      "grad_norm": 0.7094574570655823,
+      "learning_rate": 8.174735318691946e-06,
+      "loss": 0.6385,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35339580716838953,
+      "grad_norm": 0.6454529166221619,
+      "learning_rate": 8.172129342267546e-06,
+      "loss": 0.6157,
+      "step": 1829
+    },
+    {
+      "epoch": 0.35358902521495505,
+      "grad_norm": 0.6487497091293335,
+      "learning_rate": 8.169521922891933e-06,
+      "loss": 0.5921,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35378224326152063,
+      "grad_norm": 0.6693918704986572,
+      "learning_rate": 8.16691306175118e-06,
+      "loss": 0.6606,
+      "step": 1831
+    },
+    {
+      "epoch": 0.35397546130808616,
+      "grad_norm": 0.6209411025047302,
+      "learning_rate": 8.164302760032018e-06,
+      "loss": 0.6361,
+      "step": 1832
+    },
+    {
+      "epoch": 0.35416867935465174,
+      "grad_norm": 0.6884275674819946,
+      "learning_rate": 8.16169101892183e-06,
+      "loss": 0.6921,
+      "step": 1833
+    },
+    {
+      "epoch": 0.35436189740121726,
+      "grad_norm": 0.6304588913917542,
+      "learning_rate": 8.159077839608656e-06,
+      "loss": 0.5476,
+      "step": 1834
+    },
+    {
+      "epoch": 0.35455511544778284,
+      "grad_norm": 0.7084200978279114,
+      "learning_rate": 8.156463223281188e-06,
+      "loss": 0.6483,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35474833349434837,
+      "grad_norm": 0.7147865295410156,
+      "learning_rate": 8.153847171128777e-06,
+      "loss": 0.7166,
+      "step": 1836
+    },
+    {
+      "epoch": 0.35494155154091395,
+      "grad_norm": 0.6397827863693237,
+      "learning_rate": 8.151229684341422e-06,
+      "loss": 0.6391,
+      "step": 1837
+    },
+    {
+      "epoch": 0.35513476958747947,
+      "grad_norm": 0.6771610975265503,
+      "learning_rate": 8.148610764109774e-06,
+      "loss": 0.6354,
+      "step": 1838
+    },
+    {
+      "epoch": 0.355327987634045,
+      "grad_norm": 0.6359531879425049,
+      "learning_rate": 8.145990411625141e-06,
+      "loss": 0.6317,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3555212056806106,
+      "grad_norm": 0.6459310054779053,
+      "learning_rate": 8.143368628079477e-06,
+      "loss": 0.5613,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3557144237271761,
+      "grad_norm": 0.694820761680603,
+      "learning_rate": 8.140745414665394e-06,
+      "loss": 0.6274,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3559076417737417,
+      "grad_norm": 0.6848176717758179,
+      "learning_rate": 8.138120772576146e-06,
+      "loss": 0.636,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3561008598203072,
+      "grad_norm": 0.6527798771858215,
+      "learning_rate": 8.135494703005642e-06,
+      "loss": 0.6022,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3562940778668728,
+      "grad_norm": 0.6567660570144653,
+      "learning_rate": 8.132867207148442e-06,
+      "loss": 0.6355,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3564872959134383,
+      "grad_norm": 0.6557737588882446,
+      "learning_rate": 8.130238286199747e-06,
+      "loss": 0.6401,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3566805139600039,
+      "grad_norm": 0.5915987491607666,
+      "learning_rate": 8.127607941355415e-06,
+      "loss": 0.6202,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3568737320065694,
+      "grad_norm": 0.6808528304100037,
+      "learning_rate": 8.124976173811947e-06,
+      "loss": 0.5926,
+      "step": 1847
+    },
+    {
+      "epoch": 0.357066950053135,
+      "grad_norm": 0.7072309255599976,
+      "learning_rate": 8.122342984766493e-06,
+      "loss": 0.6468,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3572601680997005,
+      "grad_norm": 0.7351856827735901,
+      "learning_rate": 8.119708375416849e-06,
+      "loss": 0.6219,
+      "step": 1849
+    },
+    {
+      "epoch": 0.35745338614626604,
+      "grad_norm": 0.642025351524353,
+      "learning_rate": 8.117072346961454e-06,
+      "loss": 0.6535,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3576466041928316,
+      "grad_norm": 0.7112191915512085,
+      "learning_rate": 8.1144349005994e-06,
+      "loss": 0.6683,
+      "step": 1851
+    },
+    {
+      "epoch": 0.35783982223939714,
+      "grad_norm": 0.7107006907463074,
+      "learning_rate": 8.111796037530415e-06,
+      "loss": 0.6011,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3580330402859627,
+      "grad_norm": 0.6819424033164978,
+      "learning_rate": 8.109155758954879e-06,
+      "loss": 0.6871,
+      "step": 1853
+    },
+    {
+      "epoch": 0.35822625833252825,
+      "grad_norm": 0.6940486431121826,
+      "learning_rate": 8.106514066073812e-06,
+      "loss": 0.6566,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3584194763790938,
+      "grad_norm": 0.7076148390769958,
+      "learning_rate": 8.103870960088876e-06,
+      "loss": 0.5608,
+      "step": 1855
+    },
+    {
+      "epoch": 0.35861269442565935,
+      "grad_norm": 0.6860665678977966,
+      "learning_rate": 8.10122644220238e-06,
+      "loss": 0.6015,
+      "step": 1856
+    },
+    {
+      "epoch": 0.35880591247222493,
+      "grad_norm": 0.61956787109375,
+      "learning_rate": 8.098580513617275e-06,
+      "loss": 0.5555,
+      "step": 1857
+    },
+    {
+      "epoch": 0.35899913051879045,
+      "grad_norm": 0.6909151673316956,
+      "learning_rate": 8.095933175537148e-06,
+      "loss": 0.5764,
+      "step": 1858
+    },
+    {
+      "epoch": 0.359192348565356,
+      "grad_norm": 0.6746104955673218,
+      "learning_rate": 8.093284429166231e-06,
+      "loss": 0.5982,
+      "step": 1859
+    },
+    {
+      "epoch": 0.35938556661192156,
+      "grad_norm": 0.660260796546936,
+      "learning_rate": 8.090634275709402e-06,
+      "loss": 0.6037,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3595787846584871,
+      "grad_norm": 0.6550877690315247,
+      "learning_rate": 8.087982716372168e-06,
+      "loss": 0.6424,
+      "step": 1861
+    },
+    {
+      "epoch": 0.35977200270505266,
+      "grad_norm": 0.6771214008331299,
+      "learning_rate": 8.085329752360683e-06,
+      "loss": 0.5807,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3599652207516182,
+      "grad_norm": 0.6334057450294495,
+      "learning_rate": 8.082675384881737e-06,
+      "loss": 0.5538,
+      "step": 1863
+    },
+    {
+      "epoch": 0.36015843879818377,
+      "grad_norm": 0.6008335947990417,
+      "learning_rate": 8.080019615142763e-06,
+      "loss": 0.5536,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3603516568447493,
+      "grad_norm": 0.6921840906143188,
+      "learning_rate": 8.077362444351825e-06,
+      "loss": 0.694,
+      "step": 1865
+    },
+    {
+      "epoch": 0.36054487489131487,
+      "grad_norm": 0.7158055901527405,
+      "learning_rate": 8.07470387371763e-06,
+      "loss": 0.592,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3607380929378804,
+      "grad_norm": 0.6430540084838867,
+      "learning_rate": 8.072043904449517e-06,
+      "loss": 0.6194,
+      "step": 1867
+    },
+    {
+      "epoch": 0.360931310984446,
+      "grad_norm": 0.6790100932121277,
+      "learning_rate": 8.069382537757466e-06,
+      "loss": 0.6262,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3611245290310115,
+      "grad_norm": 0.6959518790245056,
+      "learning_rate": 8.066719774852092e-06,
+      "loss": 0.6951,
+      "step": 1869
+    },
+    {
+      "epoch": 0.361317747077577,
+      "grad_norm": 0.6456522345542908,
+      "learning_rate": 8.06405561694464e-06,
+      "loss": 0.6459,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3615109651241426,
+      "grad_norm": 0.6799988150596619,
+      "learning_rate": 8.061390065246992e-06,
+      "loss": 0.5946,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3617041831707081,
+      "grad_norm": 0.6817371845245361,
+      "learning_rate": 8.05872312097167e-06,
+      "loss": 0.6774,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3618974012172737,
+      "grad_norm": 0.6734983921051025,
+      "learning_rate": 8.05605478533182e-06,
+      "loss": 0.6808,
+      "step": 1873
+    },
+    {
+      "epoch": 0.36209061926383923,
+      "grad_norm": 0.6641425490379333,
+      "learning_rate": 8.05338505954123e-06,
+      "loss": 0.6225,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3622838373104048,
+      "grad_norm": 0.6056984663009644,
+      "learning_rate": 8.050713944814312e-06,
+      "loss": 0.5533,
+      "step": 1875
+    },
+    {
+      "epoch": 0.36247705535697033,
+      "grad_norm": 0.7348392009735107,
+      "learning_rate": 8.048041442366113e-06,
+      "loss": 0.7389,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3626702734035359,
+      "grad_norm": 0.6982197165489197,
+      "learning_rate": 8.045367553412316e-06,
+      "loss": 0.6107,
+      "step": 1877
+    },
+    {
+      "epoch": 0.36286349145010144,
+      "grad_norm": 0.72201007604599,
+      "learning_rate": 8.042692279169228e-06,
+      "loss": 0.6592,
+      "step": 1878
+    },
+    {
+      "epoch": 0.36305670949666696,
+      "grad_norm": 0.6350598931312561,
+      "learning_rate": 8.04001562085379e-06,
+      "loss": 0.5375,
+      "step": 1879
+    },
+    {
+      "epoch": 0.36324992754323254,
+      "grad_norm": 0.6758071184158325,
+      "learning_rate": 8.037337579683568e-06,
+      "loss": 0.5641,
+      "step": 1880
+    },
+    {
+      "epoch": 0.36344314558979807,
+      "grad_norm": 0.7005548477172852,
+      "learning_rate": 8.034658156876764e-06,
+      "loss": 0.587,
+      "step": 1881
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.6339564323425293,
+      "learning_rate": 8.031977353652204e-06,
+      "loss": 0.628,
+      "step": 1882
+    },
+    {
+      "epoch": 0.36382958168292917,
+      "grad_norm": 0.6859617233276367,
+      "learning_rate": 8.029295171229343e-06,
+      "loss": 0.6253,
+      "step": 1883
+    },
+    {
+      "epoch": 0.36402279972949475,
+      "grad_norm": 0.6832352876663208,
+      "learning_rate": 8.026611610828261e-06,
+      "loss": 0.6218,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3642160177760603,
+      "grad_norm": 0.6661505699157715,
+      "learning_rate": 8.02392667366967e-06,
+      "loss": 0.6859,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36440923582262585,
+      "grad_norm": 0.689653217792511,
+      "learning_rate": 8.021240360974903e-06,
+      "loss": 0.6535,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3646024538691914,
+      "grad_norm": 0.6515517234802246,
+      "learning_rate": 8.01855267396592e-06,
+      "loss": 0.6254,
+      "step": 1887
+    },
+    {
+      "epoch": 0.36479567191575696,
+      "grad_norm": 0.7004567384719849,
+      "learning_rate": 8.01586361386531e-06,
+      "loss": 0.6574,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3649888899623225,
+      "grad_norm": 0.611571729183197,
+      "learning_rate": 8.013173181896283e-06,
+      "loss": 0.5466,
+      "step": 1889
+    },
+    {
+      "epoch": 0.365182108008888,
+      "grad_norm": 0.664970874786377,
+      "learning_rate": 8.010481379282674e-06,
+      "loss": 0.585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3653753260554536,
+      "grad_norm": 0.6796653270721436,
+      "learning_rate": 8.007788207248937e-06,
+      "loss": 0.6439,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3655685441020191,
+      "grad_norm": 0.6473020315170288,
+      "learning_rate": 8.00509366702016e-06,
+      "loss": 0.5605,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3657617621485847,
+      "grad_norm": 0.6851828098297119,
+      "learning_rate": 8.002397759822042e-06,
+      "loss": 0.6195,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3659549801951502,
+      "grad_norm": 0.7153826951980591,
+      "learning_rate": 7.999700486880908e-06,
+      "loss": 0.6527,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3661481982417158,
+      "grad_norm": 0.678263783454895,
+      "learning_rate": 7.997001849423709e-06,
+      "loss": 0.6452,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3663414162882813,
+      "grad_norm": 0.7078882455825806,
+      "learning_rate": 7.994301848678006e-06,
+      "loss": 0.5962,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3665346343348469,
+      "grad_norm": 0.5946508646011353,
+      "learning_rate": 7.991600485871991e-06,
+      "loss": 0.6995,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3667278523814124,
+      "grad_norm": 0.6663452386856079,
+      "learning_rate": 7.98889776223447e-06,
+      "loss": 0.6376,
+      "step": 1898
+    },
+    {
+      "epoch": 0.36692107042797795,
+      "grad_norm": 0.6516513228416443,
+      "learning_rate": 7.986193678994869e-06,
+      "loss": 0.5676,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3671142884745435,
+      "grad_norm": 0.6839028000831604,
+      "learning_rate": 7.983488237383233e-06,
+      "loss": 0.6542,
+      "step": 1900
+    },
+    {
+      "epoch": 0.36730750652110905,
+      "grad_norm": 0.619810163974762,
+      "learning_rate": 7.980781438630221e-06,
+      "loss": 0.575,
+      "step": 1901
+    },
+    {
+      "epoch": 0.36750072456767463,
+      "grad_norm": 0.6423595547676086,
+      "learning_rate": 7.978073283967118e-06,
+      "loss": 0.6368,
+      "step": 1902
+    },
+    {
+      "epoch": 0.36769394261424015,
+      "grad_norm": 0.5996134281158447,
+      "learning_rate": 7.975363774625817e-06,
+      "loss": 0.5777,
+      "step": 1903
+    },
+    {
+      "epoch": 0.36788716066080573,
+      "grad_norm": 0.7073270082473755,
+      "learning_rate": 7.972652911838831e-06,
+      "loss": 0.6269,
+      "step": 1904
+    },
+    {
+      "epoch": 0.36808037870737126,
+      "grad_norm": 0.6661731600761414,
+      "learning_rate": 7.969940696839288e-06,
+      "loss": 0.5946,
+      "step": 1905
+    },
+    {
+      "epoch": 0.36827359675393684,
+      "grad_norm": 0.6470728516578674,
+      "learning_rate": 7.967227130860933e-06,
+      "loss": 0.6295,
+      "step": 1906
+    },
+    {
+      "epoch": 0.36846681480050236,
+      "grad_norm": 0.657211184501648,
+      "learning_rate": 7.964512215138124e-06,
+      "loss": 0.5941,
+      "step": 1907
+    },
+    {
+      "epoch": 0.36866003284706794,
+      "grad_norm": 0.7161780595779419,
+      "learning_rate": 7.961795950905833e-06,
+      "loss": 0.6123,
+      "step": 1908
+    },
+    {
+      "epoch": 0.36885325089363347,
+      "grad_norm": 0.638896644115448,
+      "learning_rate": 7.959078339399644e-06,
+      "loss": 0.6174,
+      "step": 1909
+    },
+    {
+      "epoch": 0.369046468940199,
+      "grad_norm": 0.6952905654907227,
+      "learning_rate": 7.956359381855756e-06,
+      "loss": 0.7016,
+      "step": 1910
+    },
+    {
+      "epoch": 0.36923968698676457,
+      "grad_norm": 0.6939736604690552,
+      "learning_rate": 7.95363907951098e-06,
+      "loss": 0.5998,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3694329050333301,
+      "grad_norm": 0.6497328877449036,
+      "learning_rate": 7.950917433602739e-06,
+      "loss": 0.6245,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3696261230798957,
+      "grad_norm": 0.670664370059967,
+      "learning_rate": 7.948194445369065e-06,
+      "loss": 0.6232,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3698193411264612,
+      "grad_norm": 0.7401769161224365,
+      "learning_rate": 7.945470116048601e-06,
+      "loss": 0.6689,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3700125591730268,
+      "grad_norm": 0.6523237824440002,
+      "learning_rate": 7.942744446880603e-06,
+      "loss": 0.6386,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3702057772195923,
+      "grad_norm": 0.695050060749054,
+      "learning_rate": 7.940017439104936e-06,
+      "loss": 0.6521,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3703989952661579,
+      "grad_norm": 0.7418084144592285,
+      "learning_rate": 7.937289093962069e-06,
+      "loss": 0.6565,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3705922133127234,
+      "grad_norm": 0.6626476645469666,
+      "learning_rate": 7.934559412693086e-06,
+      "loss": 0.6124,
+      "step": 1918
+    },
+    {
+      "epoch": 0.37078543135928893,
+      "grad_norm": 0.6450240015983582,
+      "learning_rate": 7.931828396539674e-06,
+      "loss": 0.6167,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3709786494058545,
+      "grad_norm": 0.724885880947113,
+      "learning_rate": 7.929096046744127e-06,
+      "loss": 0.6462,
+      "step": 1920
+    },
+    {
+      "epoch": 0.37117186745242003,
+      "grad_norm": 0.6323454976081848,
+      "learning_rate": 7.926362364549352e-06,
+      "loss": 0.59,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3713650854989856,
+      "grad_norm": 0.717215359210968,
+      "learning_rate": 7.923627351198852e-06,
+      "loss": 0.5784,
+      "step": 1922
+    },
+    {
+      "epoch": 0.37155830354555114,
+      "grad_norm": 0.6209149360656738,
+      "learning_rate": 7.920891007936746e-06,
+      "loss": 0.6034,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3717515215921167,
+      "grad_norm": 0.707101047039032,
+      "learning_rate": 7.918153336007753e-06,
+      "loss": 0.701,
+      "step": 1924
+    },
+    {
+      "epoch": 0.37194473963868224,
+      "grad_norm": 0.65291428565979,
+      "learning_rate": 7.915414336657193e-06,
+      "loss": 0.6431,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3721379576852478,
+      "grad_norm": 0.6423499584197998,
+      "learning_rate": 7.912674011130994e-06,
+      "loss": 0.6121,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37233117573181335,
+      "grad_norm": 0.6539754867553711,
+      "learning_rate": 7.909932360675687e-06,
+      "loss": 0.6393,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3725243937783789,
+      "grad_norm": 0.6901910901069641,
+      "learning_rate": 7.907189386538408e-06,
+      "loss": 0.6274,
+      "step": 1928
+    },
+    {
+      "epoch": 0.37271761182494445,
+      "grad_norm": 0.7075971364974976,
+      "learning_rate": 7.90444508996689e-06,
+      "loss": 0.6706,
+      "step": 1929
+    },
+    {
+      "epoch": 0.37291082987151,
+      "grad_norm": 0.6405119895935059,
+      "learning_rate": 7.901699472209467e-06,
+      "loss": 0.6576,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37310404791807555,
+      "grad_norm": 0.6750436425209045,
+      "learning_rate": 7.898952534515082e-06,
+      "loss": 0.5881,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3732972659646411,
+      "grad_norm": 0.6898776888847351,
+      "learning_rate": 7.89620427813327e-06,
+      "loss": 0.5832,
+      "step": 1932
+    },
+    {
+      "epoch": 0.37349048401120666,
+      "grad_norm": 0.6212689280509949,
+      "learning_rate": 7.893454704314171e-06,
+      "loss": 0.5694,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3736837020577722,
+      "grad_norm": 0.6858654618263245,
+      "learning_rate": 7.890703814308521e-06,
+      "loss": 0.6483,
+      "step": 1934
+    },
+    {
+      "epoch": 0.37387692010433776,
+      "grad_norm": 0.6993388533592224,
+      "learning_rate": 7.887951609367656e-06,
+      "loss": 0.6039,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3740701381509033,
+      "grad_norm": 0.6259585618972778,
+      "learning_rate": 7.885198090743514e-06,
+      "loss": 0.6044,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37426335619746887,
+      "grad_norm": 0.6783968806266785,
+      "learning_rate": 7.88244325968862e-06,
+      "loss": 0.6391,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3744565742440344,
+      "grad_norm": 0.7273280620574951,
+      "learning_rate": 7.879687117456107e-06,
+      "loss": 0.651,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3746497922905999,
+      "grad_norm": 0.665178120136261,
+      "learning_rate": 7.876929665299703e-06,
+      "loss": 0.5442,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3748430103371655,
+      "grad_norm": 0.651997983455658,
+      "learning_rate": 7.874170904473725e-06,
+      "loss": 0.6143,
+      "step": 1940
+    },
+    {
+      "epoch": 0.375036228383731,
+      "grad_norm": 0.7210155725479126,
+      "learning_rate": 7.87141083623309e-06,
+      "loss": 0.6746,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3752294464302966,
+      "grad_norm": 0.7050049304962158,
+      "learning_rate": 7.868649461833309e-06,
+      "loss": 0.7263,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3754226644768621,
+      "grad_norm": 0.7215242385864258,
+      "learning_rate": 7.86588678253049e-06,
+      "loss": 0.6136,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3756158825234277,
+      "grad_norm": 0.5917205810546875,
+      "learning_rate": 7.86312279958133e-06,
+      "loss": 0.5932,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3758091005699932,
+      "grad_norm": 0.7026675939559937,
+      "learning_rate": 7.860357514243119e-06,
+      "loss": 0.649,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3760023186165588,
+      "grad_norm": 0.6768333315849304,
+      "learning_rate": 7.857590927773744e-06,
+      "loss": 0.6267,
+      "step": 1946
+    },
+    {
+      "epoch": 0.37619553666312433,
+      "grad_norm": 0.6696262359619141,
+      "learning_rate": 7.85482304143168e-06,
+      "loss": 0.6127,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3763887547096899,
+      "grad_norm": 0.7130764126777649,
+      "learning_rate": 7.852053856475996e-06,
+      "loss": 0.6727,
+      "step": 1948
+    },
+    {
+      "epoch": 0.37658197275625543,
+      "grad_norm": 0.651828408241272,
+      "learning_rate": 7.84928337416635e-06,
+      "loss": 0.6123,
+      "step": 1949
+    },
+    {
+      "epoch": 0.37677519080282096,
+      "grad_norm": 0.6409557461738586,
+      "learning_rate": 7.846511595762989e-06,
+      "loss": 0.6339,
+      "step": 1950
+    },
+    {
+      "epoch": 0.37696840884938654,
+      "grad_norm": 0.6870968341827393,
+      "learning_rate": 7.84373852252675e-06,
+      "loss": 0.648,
+      "step": 1951
+    },
+    {
+      "epoch": 0.37716162689595206,
+      "grad_norm": 0.6684063673019409,
+      "learning_rate": 7.84096415571906e-06,
+      "loss": 0.6655,
+      "step": 1952
+    },
+    {
+      "epoch": 0.37735484494251764,
+      "grad_norm": 0.6545553803443909,
+      "learning_rate": 7.83818849660194e-06,
+      "loss": 0.5811,
+      "step": 1953
+    },
+    {
+      "epoch": 0.37754806298908317,
+      "grad_norm": 0.644171416759491,
+      "learning_rate": 7.835411546437986e-06,
+      "loss": 0.5844,
+      "step": 1954
+    },
+    {
+      "epoch": 0.37774128103564875,
+      "grad_norm": 0.66285240650177,
+      "learning_rate": 7.832633306490392e-06,
+      "loss": 0.5846,
+      "step": 1955
+    },
+    {
+      "epoch": 0.37793449908221427,
+      "grad_norm": 0.6913794875144958,
+      "learning_rate": 7.829853778022934e-06,
+      "loss": 0.6562,
+      "step": 1956
+    },
+    {
+      "epoch": 0.37812771712877985,
+      "grad_norm": 0.7105599641799927,
+      "learning_rate": 7.827072962299978e-06,
+      "loss": 0.7038,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3783209351753454,
+      "grad_norm": 0.6511922478675842,
+      "learning_rate": 7.82429086058647e-06,
+      "loss": 0.5754,
+      "step": 1958
+    },
+    {
+      "epoch": 0.37851415322191095,
+      "grad_norm": 0.6237979531288147,
+      "learning_rate": 7.821507474147943e-06,
+      "loss": 0.6423,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3787073712684765,
+      "grad_norm": 0.6769509315490723,
+      "learning_rate": 7.818722804250518e-06,
+      "loss": 0.5897,
+      "step": 1960
+    },
+    {
+      "epoch": 0.378900589315042,
+      "grad_norm": 0.6772351861000061,
+      "learning_rate": 7.815936852160898e-06,
+      "loss": 0.6397,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3790938073616076,
+      "grad_norm": 0.6936584711074829,
+      "learning_rate": 7.813149619146363e-06,
+      "loss": 0.6552,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3792870254081731,
+      "grad_norm": 0.6388738751411438,
+      "learning_rate": 7.810361106474786e-06,
+      "loss": 0.6212,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3794802434547387,
+      "grad_norm": 0.7098832130432129,
+      "learning_rate": 7.807571315414616e-06,
+      "loss": 0.7486,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3796734615013042,
+      "grad_norm": 0.6388980746269226,
+      "learning_rate": 7.804780247234882e-06,
+      "loss": 0.6026,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3798666795478698,
+      "grad_norm": 0.6529965400695801,
+      "learning_rate": 7.801987903205201e-06,
+      "loss": 0.6082,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3800598975944353,
+      "grad_norm": 0.646763801574707,
+      "learning_rate": 7.799194284595762e-06,
+      "loss": 0.5924,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3802531156410009,
+      "grad_norm": 0.6837568879127502,
+      "learning_rate": 7.796399392677338e-06,
+      "loss": 0.6397,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3804463336875664,
+      "grad_norm": 0.7214167714118958,
+      "learning_rate": 7.793603228721285e-06,
+      "loss": 0.5762,
+      "step": 1969
+    },
+    {
+      "epoch": 0.38063955173413194,
+      "grad_norm": 0.71705162525177,
+      "learning_rate": 7.79080579399953e-06,
+      "loss": 0.6816,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3808327697806975,
+      "grad_norm": 0.5831835269927979,
+      "learning_rate": 7.788007089784579e-06,
+      "loss": 0.5752,
+      "step": 1971
+    },
+    {
+      "epoch": 0.38102598782726305,
+      "grad_norm": 0.7381232976913452,
+      "learning_rate": 7.785207117349525e-06,
+      "loss": 0.5797,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3812192058738286,
+      "grad_norm": 0.6481449007987976,
+      "learning_rate": 7.782405877968024e-06,
+      "loss": 0.6945,
+      "step": 1973
+    },
+    {
+      "epoch": 0.38141242392039415,
+      "grad_norm": 0.6488431096076965,
+      "learning_rate": 7.779603372914321e-06,
+      "loss": 0.6213,
+      "step": 1974
+    },
+    {
+      "epoch": 0.38160564196695973,
+      "grad_norm": 0.6575244069099426,
+      "learning_rate": 7.776799603463228e-06,
+      "loss": 0.6506,
+      "step": 1975
+    },
+    {
+      "epoch": 0.38179886001352525,
+      "grad_norm": 0.7259254455566406,
+      "learning_rate": 7.773994570890134e-06,
+      "loss": 0.643,
+      "step": 1976
+    },
+    {
+      "epoch": 0.38199207806009083,
+      "grad_norm": 0.6621873378753662,
+      "learning_rate": 7.771188276471007e-06,
+      "loss": 0.6478,
+      "step": 1977
+    },
+    {
+      "epoch": 0.38218529610665636,
+      "grad_norm": 0.6408529281616211,
+      "learning_rate": 7.768380721482383e-06,
+      "loss": 0.5724,
+      "step": 1978
+    },
+    {
+      "epoch": 0.38237851415322194,
+      "grad_norm": 0.643339991569519,
+      "learning_rate": 7.765571907201377e-06,
+      "loss": 0.6023,
+      "step": 1979
+    },
+    {
+      "epoch": 0.38257173219978746,
+      "grad_norm": 0.6658744812011719,
+      "learning_rate": 7.76276183490567e-06,
+      "loss": 0.6726,
+      "step": 1980
+    },
+    {
+      "epoch": 0.382764950246353,
+      "grad_norm": 0.7783738970756531,
+      "learning_rate": 7.759950505873523e-06,
+      "loss": 0.6877,
+      "step": 1981
+    },
+    {
+      "epoch": 0.38295816829291857,
+      "grad_norm": 0.706911027431488,
+      "learning_rate": 7.757137921383761e-06,
+      "loss": 0.7314,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3831513863394841,
+      "grad_norm": 0.6127927303314209,
+      "learning_rate": 7.75432408271579e-06,
+      "loss": 0.5789,
+      "step": 1983
+    },
+    {
+      "epoch": 0.38334460438604967,
+      "grad_norm": 0.6393814086914062,
+      "learning_rate": 7.751508991149573e-06,
+      "loss": 0.6171,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3835378224326152,
+      "grad_norm": 0.6182101964950562,
+      "learning_rate": 7.748692647965657e-06,
+      "loss": 0.5891,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3837310404791808,
+      "grad_norm": 0.7281214594841003,
+      "learning_rate": 7.745875054445146e-06,
+      "loss": 0.6885,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3839242585257463,
+      "grad_norm": 0.6789116859436035,
+      "learning_rate": 7.743056211869723e-06,
+      "loss": 0.6607,
+      "step": 1987
+    },
+    {
+      "epoch": 0.3841174765723119,
+      "grad_norm": 0.6873614192008972,
+      "learning_rate": 7.740236121521634e-06,
+      "loss": 0.6113,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3843106946188774,
+      "grad_norm": 0.6913609504699707,
+      "learning_rate": 7.73741478468369e-06,
+      "loss": 0.6687,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3845039126654429,
+      "grad_norm": 0.6552664041519165,
+      "learning_rate": 7.734592202639277e-06,
+      "loss": 0.6345,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3846971307120085,
+      "grad_norm": 0.7414533495903015,
+      "learning_rate": 7.73176837667234e-06,
+      "loss": 0.6347,
+      "step": 1991
+    },
+    {
+      "epoch": 0.38489034875857403,
+      "grad_norm": 0.6337794065475464,
+      "learning_rate": 7.72894330806739e-06,
+      "loss": 0.5368,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3850835668051396,
+      "grad_norm": 0.7334190011024475,
+      "learning_rate": 7.72611699810951e-06,
+      "loss": 0.6345,
+      "step": 1993
+    },
+    {
+      "epoch": 0.38527678485170513,
+      "grad_norm": 0.6909582614898682,
+      "learning_rate": 7.723289448084343e-06,
+      "loss": 0.673,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3854700028982707,
+      "grad_norm": 0.7199613451957703,
+      "learning_rate": 7.720460659278097e-06,
+      "loss": 0.6172,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38566322094483624,
+      "grad_norm": 0.6208944320678711,
+      "learning_rate": 7.71763063297754e-06,
+      "loss": 0.5624,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3858564389914018,
+      "grad_norm": 0.7461687326431274,
+      "learning_rate": 7.714799370470007e-06,
+      "loss": 0.6062,
+      "step": 1997
+    },
+    {
+      "epoch": 0.38604965703796734,
+      "grad_norm": 0.65488600730896,
+      "learning_rate": 7.711966873043396e-06,
+      "loss": 0.5572,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3862428750845329,
+      "grad_norm": 0.6622769236564636,
+      "learning_rate": 7.709133141986165e-06,
+      "loss": 0.6207,
+      "step": 1999
+    },
+    {
+      "epoch": 0.38643609313109845,
+      "grad_norm": 0.6557113528251648,
+      "learning_rate": 7.706298178587331e-06,
+      "loss": 0.6306,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38662931117766397,
+      "grad_norm": 0.6891779899597168,
+      "learning_rate": 7.703461984136475e-06,
+      "loss": 0.6449,
+      "step": 2001
+    },
+    {
+      "epoch": 0.38682252922422955,
+      "grad_norm": 0.665860652923584,
+      "learning_rate": 7.700624559923738e-06,
+      "loss": 0.6083,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3870157472707951,
+      "grad_norm": 0.6983426809310913,
+      "learning_rate": 7.697785907239819e-06,
+      "loss": 0.6173,
+      "step": 2003
+    },
+    {
+      "epoch": 0.38720896531736065,
+      "grad_norm": 0.6503394246101379,
+      "learning_rate": 7.694946027375972e-06,
+      "loss": 0.6361,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3874021833639262,
+      "grad_norm": 0.6754993200302124,
+      "learning_rate": 7.692104921624017e-06,
+      "loss": 0.6579,
+      "step": 2005
+    },
+    {
+      "epoch": 0.38759540141049176,
+      "grad_norm": 0.6708557605743408,
+      "learning_rate": 7.689262591276328e-06,
+      "loss": 0.5994,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3877886194570573,
+      "grad_norm": 0.7063451409339905,
+      "learning_rate": 7.686419037625834e-06,
+      "loss": 0.59,
+      "step": 2007
+    },
+    {
+      "epoch": 0.38798183750362286,
+      "grad_norm": 0.6672752499580383,
+      "learning_rate": 7.683574261966024e-06,
+      "loss": 0.6435,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3881750555501884,
+      "grad_norm": 0.6641849875450134,
+      "learning_rate": 7.68072826559094e-06,
+      "loss": 0.723,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3883682735967539,
+      "grad_norm": 0.7090732455253601,
+      "learning_rate": 7.677881049795183e-06,
+      "loss": 0.5878,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3885614916433195,
+      "grad_norm": 0.6408359408378601,
+      "learning_rate": 7.675032615873903e-06,
+      "loss": 0.6223,
+      "step": 2011
+    },
+    {
+      "epoch": 0.388754709689885,
+      "grad_norm": 0.6255666613578796,
+      "learning_rate": 7.672182965122808e-06,
+      "loss": 0.5995,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3889479277364506,
+      "grad_norm": 0.6584436893463135,
+      "learning_rate": 7.669332098838161e-06,
+      "loss": 0.6291,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3891411457830161,
+      "grad_norm": 0.6589295864105225,
+      "learning_rate": 7.666480018316771e-06,
+      "loss": 0.5799,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3893343638295817,
+      "grad_norm": 0.6782766580581665,
+      "learning_rate": 7.66362672485601e-06,
+      "loss": 0.6762,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3895275818761472,
+      "grad_norm": 0.7227813601493835,
+      "learning_rate": 7.660772219753792e-06,
+      "loss": 0.6368,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3897207999227128,
+      "grad_norm": 0.632710337638855,
+      "learning_rate": 7.657916504308587e-06,
+      "loss": 0.5735,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3899140179692783,
+      "grad_norm": 0.7233976125717163,
+      "learning_rate": 7.655059579819414e-06,
+      "loss": 0.6483,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3901072360158439,
+      "grad_norm": 0.6342703104019165,
+      "learning_rate": 7.652201447585845e-06,
+      "loss": 0.6141,
+      "step": 2019
+    },
+    {
+      "epoch": 0.39030045406240943,
+      "grad_norm": 0.6752959489822388,
+      "learning_rate": 7.649342108907998e-06,
+      "loss": 0.5941,
+      "step": 2020
+    },
+    {
+      "epoch": 0.39049367210897495,
+      "grad_norm": 0.7302737236022949,
+      "learning_rate": 7.646481565086537e-06,
+      "loss": 0.6808,
+      "step": 2021
+    },
+    {
+      "epoch": 0.39068689015554053,
+      "grad_norm": 0.7574686408042908,
+      "learning_rate": 7.64361981742268e-06,
+      "loss": 0.6812,
+      "step": 2022
+    },
+    {
+      "epoch": 0.39088010820210606,
+      "grad_norm": 0.6549816131591797,
+      "learning_rate": 7.640756867218194e-06,
+      "loss": 0.5862,
+      "step": 2023
+    },
+    {
+      "epoch": 0.39107332624867164,
+      "grad_norm": 0.7136653661727905,
+      "learning_rate": 7.637892715775385e-06,
+      "loss": 0.6538,
+      "step": 2024
+    },
+    {
+      "epoch": 0.39126654429523716,
+      "grad_norm": 0.656946063041687,
+      "learning_rate": 7.635027364397114e-06,
+      "loss": 0.6236,
+      "step": 2025
+    },
+    {
+      "epoch": 0.39145976234180274,
+      "grad_norm": 0.6660289764404297,
+      "learning_rate": 7.63216081438678e-06,
+      "loss": 0.5958,
+      "step": 2026
+    },
+    {
+      "epoch": 0.39165298038836827,
+      "grad_norm": 0.6436180472373962,
+      "learning_rate": 7.629293067048334e-06,
+      "loss": 0.6232,
+      "step": 2027
+    },
+    {
+      "epoch": 0.39184619843493385,
+      "grad_norm": 0.7067444324493408,
+      "learning_rate": 7.626424123686264e-06,
+      "loss": 0.6126,
+      "step": 2028
+    },
+    {
+      "epoch": 0.39203941648149937,
+      "grad_norm": 0.7418837547302246,
+      "learning_rate": 7.623553985605612e-06,
+      "loss": 0.5886,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3922326345280649,
+      "grad_norm": 0.6752511262893677,
+      "learning_rate": 7.620682654111955e-06,
+      "loss": 0.6088,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3924258525746305,
+      "grad_norm": 0.7069725394248962,
+      "learning_rate": 7.617810130511419e-06,
+      "loss": 0.6621,
+      "step": 2031
+    },
+    {
+      "epoch": 0.392619070621196,
+      "grad_norm": 0.7117714881896973,
+      "learning_rate": 7.614936416110668e-06,
+      "loss": 0.6296,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3928122886677616,
+      "grad_norm": 0.6504923701286316,
+      "learning_rate": 7.612061512216907e-06,
+      "loss": 0.6783,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3930055067143271,
+      "grad_norm": 0.6737610697746277,
+      "learning_rate": 7.609185420137886e-06,
+      "loss": 0.6346,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3931987247608927,
+      "grad_norm": 0.6762468814849854,
+      "learning_rate": 7.606308141181893e-06,
+      "loss": 0.6501,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3933919428074582,
+      "grad_norm": 0.6071451902389526,
+      "learning_rate": 7.603429676657756e-06,
+      "loss": 0.592,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3935851608540238,
+      "grad_norm": 0.7296298742294312,
+      "learning_rate": 7.600550027874844e-06,
+      "loss": 0.6564,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3937783789005893,
+      "grad_norm": 0.6787495613098145,
+      "learning_rate": 7.597669196143063e-06,
+      "loss": 0.6019,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3939715969471549,
+      "grad_norm": 0.70437091588974,
+      "learning_rate": 7.594787182772858e-06,
+      "loss": 0.604,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3941648149937204,
+      "grad_norm": 0.6922838091850281,
+      "learning_rate": 7.591903989075211e-06,
+      "loss": 0.5971,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39435803304028594,
+      "grad_norm": 0.6532790660858154,
+      "learning_rate": 7.5890196163616396e-06,
+      "loss": 0.6286,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3945512510868515,
+      "grad_norm": 0.6912763118743896,
+      "learning_rate": 7.5861340659442e-06,
+      "loss": 0.6203,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39474446913341704,
+      "grad_norm": 0.673835039138794,
+      "learning_rate": 7.5832473391354825e-06,
+      "loss": 0.6208,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3949376871799826,
+      "grad_norm": 0.661154568195343,
+      "learning_rate": 7.580359437248613e-06,
+      "loss": 0.6285,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39513090522654815,
+      "grad_norm": 0.6069462299346924,
+      "learning_rate": 7.577470361597256e-06,
+      "loss": 0.5444,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3953241232731137,
+      "grad_norm": 0.6433176398277283,
+      "learning_rate": 7.5745801134955994e-06,
+      "loss": 0.6163,
+      "step": 2046
+    },
+    {
+      "epoch": 0.39551734131967925,
+      "grad_norm": 0.7406687140464783,
+      "learning_rate": 7.571688694258379e-06,
+      "loss": 0.6742,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39571055936624483,
+      "grad_norm": 0.6301264762878418,
+      "learning_rate": 7.568796105200849e-06,
+      "loss": 0.6182,
+      "step": 2048
+    },
+    {
+      "epoch": 0.39590377741281035,
+      "grad_norm": 0.659673810005188,
+      "learning_rate": 7.565902347638806e-06,
+      "loss": 0.5901,
+      "step": 2049
+    },
+    {
+      "epoch": 0.39609699545937593,
+      "grad_norm": 0.5929459929466248,
+      "learning_rate": 7.563007422888571e-06,
+      "loss": 0.5244,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39629021350594146,
+      "grad_norm": 0.7307798266410828,
+      "learning_rate": 7.560111332267002e-06,
+      "loss": 0.6759,
+      "step": 2051
+    },
+    {
+      "epoch": 0.396483431552507,
+      "grad_norm": 0.6598896384239197,
+      "learning_rate": 7.557214077091483e-06,
+      "loss": 0.607,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39667664959907256,
+      "grad_norm": 0.6687085032463074,
+      "learning_rate": 7.554315658679932e-06,
+      "loss": 0.6132,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3968698676456381,
+      "grad_norm": 0.6298744082450867,
+      "learning_rate": 7.5514160783507906e-06,
+      "loss": 0.5151,
+      "step": 2054
+    },
+    {
+      "epoch": 0.39706308569220367,
+      "grad_norm": 0.6075497269630432,
+      "learning_rate": 7.548515337423033e-06,
+      "loss": 0.5784,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3972563037387692,
+      "grad_norm": 0.6306989192962646,
+      "learning_rate": 7.545613437216158e-06,
+      "loss": 0.5268,
+      "step": 2056
+    },
+    {
+      "epoch": 0.39744952178533477,
+      "grad_norm": 0.6692643761634827,
+      "learning_rate": 7.542710379050198e-06,
+      "loss": 0.6534,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3976427398319003,
+      "grad_norm": 0.8413420915603638,
+      "learning_rate": 7.539806164245703e-06,
+      "loss": 0.6051,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3978359578784659,
+      "grad_norm": 0.6725656390190125,
+      "learning_rate": 7.536900794123758e-06,
+      "loss": 0.6058,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3980291759250314,
+      "grad_norm": 0.7201004028320312,
+      "learning_rate": 7.533994270005969e-06,
+      "loss": 0.7435,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3982223939715969,
+      "grad_norm": 0.7333308458328247,
+      "learning_rate": 7.531086593214468e-06,
+      "loss": 0.6682,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3984156120181625,
+      "grad_norm": 0.7050367593765259,
+      "learning_rate": 7.528177765071909e-06,
+      "loss": 0.7065,
+      "step": 2062
+    },
+    {
+      "epoch": 0.398608830064728,
+      "grad_norm": 0.6816328763961792,
+      "learning_rate": 7.525267786901472e-06,
+      "loss": 0.6309,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3988020481112936,
+      "grad_norm": 0.661410927772522,
+      "learning_rate": 7.522356660026861e-06,
+      "loss": 0.6319,
+      "step": 2064
+    },
+    {
+      "epoch": 0.39899526615785913,
+      "grad_norm": 0.6452182531356812,
+      "learning_rate": 7.519444385772303e-06,
+      "loss": 0.5638,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3991884842044247,
+      "grad_norm": 0.6219295263290405,
+      "learning_rate": 7.5165309654625405e-06,
+      "loss": 0.5846,
+      "step": 2066
+    },
+    {
+      "epoch": 0.39938170225099023,
+      "grad_norm": 0.6982712149620056,
+      "learning_rate": 7.513616400422845e-06,
+      "loss": 0.6444,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3995749202975558,
+      "grad_norm": 0.6853764653205872,
+      "learning_rate": 7.5107006919790075e-06,
+      "loss": 0.6749,
+      "step": 2068
+    },
+    {
+      "epoch": 0.39976813834412134,
+      "grad_norm": 0.6444025039672852,
+      "learning_rate": 7.507783841457335e-06,
+      "loss": 0.6405,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3999613563906869,
+      "grad_norm": 0.615207850933075,
+      "learning_rate": 7.504865850184657e-06,
+      "loss": 0.6001,
+      "step": 2070
+    },
+    {
+      "epoch": 0.40015457443725244,
+      "grad_norm": 0.6246384978294373,
+      "learning_rate": 7.50194671948832e-06,
+      "loss": 0.5639,
+      "step": 2071
+    },
+    {
+      "epoch": 0.40034779248381797,
+      "grad_norm": 0.7192694544792175,
+      "learning_rate": 7.499026450696192e-06,
+      "loss": 0.706,
+      "step": 2072
+    },
+    {
+      "epoch": 0.40054101053038355,
+      "grad_norm": 0.6590126752853394,
+      "learning_rate": 7.496105045136656e-06,
+      "loss": 0.6357,
+      "step": 2073
+    },
+    {
+      "epoch": 0.40073422857694907,
+      "grad_norm": 0.6767734289169312,
+      "learning_rate": 7.493182504138614e-06,
+      "loss": 0.6217,
+      "step": 2074
+    },
+    {
+      "epoch": 0.40092744662351465,
+      "grad_norm": 0.6955440640449524,
+      "learning_rate": 7.490258829031482e-06,
+      "loss": 0.6008,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4011206646700802,
+      "grad_norm": 0.6890498995780945,
+      "learning_rate": 7.487334021145194e-06,
+      "loss": 0.5593,
+      "step": 2076
+    },
+    {
+      "epoch": 0.40131388271664575,
+      "grad_norm": 0.625720202922821,
+      "learning_rate": 7.4844080818101996e-06,
+      "loss": 0.6734,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4015071007632113,
+      "grad_norm": 0.7054499387741089,
+      "learning_rate": 7.481481012357461e-06,
+      "loss": 0.6504,
+      "step": 2078
+    },
+    {
+      "epoch": 0.40170031880977686,
+      "grad_norm": 0.6577969789505005,
+      "learning_rate": 7.478552814118456e-06,
+      "loss": 0.6645,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4018935368563424,
+      "grad_norm": 0.6737346649169922,
+      "learning_rate": 7.475623488425175e-06,
+      "loss": 0.5657,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4020867549029079,
+      "grad_norm": 0.6806143522262573,
+      "learning_rate": 7.472693036610119e-06,
+      "loss": 0.6359,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4022799729494735,
+      "grad_norm": 0.6800273656845093,
+      "learning_rate": 7.469761460006309e-06,
+      "loss": 0.6423,
+      "step": 2082
+    },
+    {
+      "epoch": 0.402473190996039,
+      "grad_norm": 0.6726073622703552,
+      "learning_rate": 7.466828759947271e-06,
+      "loss": 0.5743,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4026664090426046,
+      "grad_norm": 0.695620596408844,
+      "learning_rate": 7.463894937767042e-06,
+      "loss": 0.6694,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4028596270891701,
+      "grad_norm": 0.6810131669044495,
+      "learning_rate": 7.460959994800173e-06,
+      "loss": 0.5886,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4030528451357357,
+      "grad_norm": 0.7302511930465698,
+      "learning_rate": 7.458023932381722e-06,
+      "loss": 0.6538,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4032460631823012,
+      "grad_norm": 0.6560137867927551,
+      "learning_rate": 7.455086751847257e-06,
+      "loss": 0.6041,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4034392812288668,
+      "grad_norm": 0.5975316762924194,
+      "learning_rate": 7.452148454532852e-06,
+      "loss": 0.5772,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4036324992754323,
+      "grad_norm": 0.6206765174865723,
+      "learning_rate": 7.449209041775098e-06,
+      "loss": 0.5721,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4038257173219979,
+      "grad_norm": 0.7213665843009949,
+      "learning_rate": 7.446268514911082e-06,
+      "loss": 0.6264,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4040189353685634,
+      "grad_norm": 0.6876228451728821,
+      "learning_rate": 7.443326875278406e-06,
+      "loss": 0.6748,
+      "step": 2091
+    },
+    {
+      "epoch": 0.40421215341512895,
+      "grad_norm": 0.696732759475708,
+      "learning_rate": 7.440384124215173e-06,
+      "loss": 0.583,
+      "step": 2092
+    },
+    {
+      "epoch": 0.40440537146169453,
+      "grad_norm": 0.6708269715309143,
+      "learning_rate": 7.437440263059994e-06,
+      "loss": 0.6035,
+      "step": 2093
+    },
+    {
+      "epoch": 0.40459858950826005,
+      "grad_norm": 0.7025800943374634,
+      "learning_rate": 7.434495293151987e-06,
+      "loss": 0.6764,
+      "step": 2094
+    },
+    {
+      "epoch": 0.40479180755482563,
+      "grad_norm": 0.6457666158676147,
+      "learning_rate": 7.431549215830768e-06,
+      "loss": 0.6381,
+      "step": 2095
+    },
+    {
+      "epoch": 0.40498502560139116,
+      "grad_norm": 0.628460168838501,
+      "learning_rate": 7.428602032436463e-06,
+      "loss": 0.5602,
+      "step": 2096
+    },
+    {
+      "epoch": 0.40517824364795674,
+      "grad_norm": 0.6788326501846313,
+      "learning_rate": 7.425653744309699e-06,
+      "loss": 0.6099,
+      "step": 2097
+    },
+    {
+      "epoch": 0.40537146169452226,
+      "grad_norm": 0.6846693754196167,
+      "learning_rate": 7.4227043527916054e-06,
+      "loss": 0.5605,
+      "step": 2098
+    },
+    {
+      "epoch": 0.40556467974108784,
+      "grad_norm": 0.6556950211524963,
+      "learning_rate": 7.4197538592238125e-06,
+      "loss": 0.6112,
+      "step": 2099
+    },
+    {
+      "epoch": 0.40575789778765337,
+      "grad_norm": 0.7406560182571411,
+      "learning_rate": 7.416802264948455e-06,
+      "loss": 0.6007,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4059511158342189,
+      "grad_norm": 0.6819958090782166,
+      "learning_rate": 7.413849571308164e-06,
+      "loss": 0.6012,
+      "step": 2101
+    },
+    {
+      "epoch": 0.40614433388078447,
+      "grad_norm": 0.6772670745849609,
+      "learning_rate": 7.410895779646075e-06,
+      "loss": 0.5971,
+      "step": 2102
+    },
+    {
+      "epoch": 0.40633755192735,
+      "grad_norm": 0.706831693649292,
+      "learning_rate": 7.407940891305817e-06,
+      "loss": 0.6449,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4065307699739156,
+      "grad_norm": 0.664724588394165,
+      "learning_rate": 7.404984907631526e-06,
+      "loss": 0.5569,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4067239880204811,
+      "grad_norm": 0.6596086621284485,
+      "learning_rate": 7.40202782996783e-06,
+      "loss": 0.574,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4069172060670467,
+      "grad_norm": 0.7035719752311707,
+      "learning_rate": 7.399069659659857e-06,
+      "loss": 0.6416,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4071104241136122,
+      "grad_norm": 0.7647302746772766,
+      "learning_rate": 7.396110398053231e-06,
+      "loss": 0.637,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4073036421601778,
+      "grad_norm": 0.6788449883460999,
+      "learning_rate": 7.393150046494075e-06,
+      "loss": 0.6179,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4074968602067433,
+      "grad_norm": 0.6259075999259949,
+      "learning_rate": 7.390188606329004e-06,
+      "loss": 0.6098,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4076900782533089,
+      "grad_norm": 0.6721020936965942,
+      "learning_rate": 7.387226078905131e-06,
+      "loss": 0.6292,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4078832962998744,
+      "grad_norm": 0.6701378226280212,
+      "learning_rate": 7.384262465570061e-06,
+      "loss": 0.5985,
+      "step": 2111
+    },
+    {
+      "epoch": 0.40807651434643993,
+      "grad_norm": 0.5993145108222961,
+      "learning_rate": 7.381297767671897e-06,
+      "loss": 0.5884,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4082697323930055,
+      "grad_norm": 0.7082571387290955,
+      "learning_rate": 7.3783319865592306e-06,
+      "loss": 0.6798,
+      "step": 2113
+    },
+    {
+      "epoch": 0.40846295043957104,
+      "grad_norm": 0.6963121294975281,
+      "learning_rate": 7.3753651235811505e-06,
+      "loss": 0.5743,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4086561684861366,
+      "grad_norm": 0.6942605376243591,
+      "learning_rate": 7.372397180087235e-06,
+      "loss": 0.6176,
+      "step": 2115
+    },
+    {
+      "epoch": 0.40884938653270214,
+      "grad_norm": 0.6287813186645508,
+      "learning_rate": 7.369428157427554e-06,
+      "loss": 0.6407,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4090426045792677,
+      "grad_norm": 0.6543014049530029,
+      "learning_rate": 7.366458056952668e-06,
+      "loss": 0.6287,
+      "step": 2117
+    },
+    {
+      "epoch": 0.40923582262583325,
+      "grad_norm": 0.6604689359664917,
+      "learning_rate": 7.36348688001363e-06,
+      "loss": 0.5803,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4094290406723988,
+      "grad_norm": 0.6374219655990601,
+      "learning_rate": 7.36051462796198e-06,
+      "loss": 0.6448,
+      "step": 2119
+    },
+    {
+      "epoch": 0.40962225871896435,
+      "grad_norm": 0.6406296491622925,
+      "learning_rate": 7.35754130214975e-06,
+      "loss": 0.6017,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4098154767655299,
+      "grad_norm": 0.7046404480934143,
+      "learning_rate": 7.354566903929456e-06,
+      "loss": 0.6403,
+      "step": 2121
+    },
+    {
+      "epoch": 0.41000869481209545,
+      "grad_norm": 0.723128080368042,
+      "learning_rate": 7.351591434654106e-06,
+      "loss": 0.6684,
+      "step": 2122
+    },
+    {
+      "epoch": 0.410201912858661,
+      "grad_norm": 0.7090283632278442,
+      "learning_rate": 7.348614895677193e-06,
+      "loss": 0.6384,
+      "step": 2123
+    },
+    {
+      "epoch": 0.41039513090522656,
+      "grad_norm": 0.6532183289527893,
+      "learning_rate": 7.3456372883527e-06,
+      "loss": 0.5953,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4105883489517921,
+      "grad_norm": 0.7115582823753357,
+      "learning_rate": 7.3426586140350875e-06,
+      "loss": 0.5752,
+      "step": 2125
+    },
+    {
+      "epoch": 0.41078156699835766,
+      "grad_norm": 0.7227739095687866,
+      "learning_rate": 7.339678874079312e-06,
+      "loss": 0.713,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4109747850449232,
+      "grad_norm": 0.6985675096511841,
+      "learning_rate": 7.336698069840808e-06,
+      "loss": 0.6054,
+      "step": 2127
+    },
+    {
+      "epoch": 0.41116800309148877,
+      "grad_norm": 0.6137675642967224,
+      "learning_rate": 7.3337162026754985e-06,
+      "loss": 0.5707,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4113612211380543,
+      "grad_norm": 0.6266011595726013,
+      "learning_rate": 7.330733273939784e-06,
+      "loss": 0.6154,
+      "step": 2129
+    },
+    {
+      "epoch": 0.41155443918461987,
+      "grad_norm": 0.6761579513549805,
+      "learning_rate": 7.3277492849905515e-06,
+      "loss": 0.5988,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4117476572311854,
+      "grad_norm": 0.6265912055969238,
+      "learning_rate": 7.324764237185174e-06,
+      "loss": 0.6343,
+      "step": 2131
+    },
+    {
+      "epoch": 0.4119408752777509,
+      "grad_norm": 0.6212235689163208,
+      "learning_rate": 7.321778131881497e-06,
+      "loss": 0.5788,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4121340933243165,
+      "grad_norm": 0.7112824320793152,
+      "learning_rate": 7.318790970437856e-06,
+      "loss": 0.5652,
+      "step": 2133
+    },
+    {
+      "epoch": 0.412327311370882,
+      "grad_norm": 0.6436079740524292,
+      "learning_rate": 7.315802754213062e-06,
+      "loss": 0.6184,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4125205294174476,
+      "grad_norm": 0.6424702405929565,
+      "learning_rate": 7.3128134845664074e-06,
+      "loss": 0.6204,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4127137474640131,
+      "grad_norm": 0.6909423470497131,
+      "learning_rate": 7.309823162857663e-06,
+      "loss": 0.6646,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4129069655105787,
+      "grad_norm": 0.6223532557487488,
+      "learning_rate": 7.30683179044708e-06,
+      "loss": 0.6103,
+      "step": 2137
+    },
+    {
+      "epoch": 0.41310018355714423,
+      "grad_norm": 0.6614201664924622,
+      "learning_rate": 7.303839368695384e-06,
+      "loss": 0.6482,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4132934016037098,
+      "grad_norm": 0.713736355304718,
+      "learning_rate": 7.300845898963782e-06,
+      "loss": 0.6388,
+      "step": 2139
+    },
+    {
+      "epoch": 0.41348661965027533,
+      "grad_norm": 0.621274471282959,
+      "learning_rate": 7.297851382613955e-06,
+      "loss": 0.5502,
+      "step": 2140
+    },
+    {
+      "epoch": 0.41367983769684086,
+      "grad_norm": 0.587506115436554,
+      "learning_rate": 7.29485582100806e-06,
+      "loss": 0.6522,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41387305574340644,
+      "grad_norm": 0.6261395215988159,
+      "learning_rate": 7.291859215508732e-06,
+      "loss": 0.6032,
+      "step": 2142
+    },
+    {
+      "epoch": 0.41406627378997196,
+      "grad_norm": 0.7300223708152771,
+      "learning_rate": 7.288861567479078e-06,
+      "loss": 0.5844,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41425949183653754,
+      "grad_norm": 0.6551735401153564,
+      "learning_rate": 7.285862878282681e-06,
+      "loss": 0.6172,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41445270988310307,
+      "grad_norm": 0.6083824634552002,
+      "learning_rate": 7.282863149283598e-06,
+      "loss": 0.5443,
+      "step": 2145
+    },
+    {
+      "epoch": 0.41464592792966865,
+      "grad_norm": 0.7016834616661072,
+      "learning_rate": 7.2798623818463555e-06,
+      "loss": 0.6392,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41483914597623417,
+      "grad_norm": 0.7261713147163391,
+      "learning_rate": 7.276860577335957e-06,
+      "loss": 0.689,
+      "step": 2147
+    },
+    {
+      "epoch": 0.41503236402279975,
+      "grad_norm": 0.6647863984107971,
+      "learning_rate": 7.273857737117874e-06,
+      "loss": 0.6363,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4152255820693653,
+      "grad_norm": 0.7028880715370178,
+      "learning_rate": 7.270853862558054e-06,
+      "loss": 0.6223,
+      "step": 2149
+    },
+    {
+      "epoch": 0.41541880011593085,
+      "grad_norm": 0.6972260475158691,
+      "learning_rate": 7.267848955022908e-06,
+      "loss": 0.684,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4156120181624964,
+      "grad_norm": 0.6885837912559509,
+      "learning_rate": 7.264843015879321e-06,
+      "loss": 0.658,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4158052362090619,
+      "grad_norm": 0.6620647311210632,
+      "learning_rate": 7.261836046494651e-06,
+      "loss": 0.5675,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4159984542556275,
+      "grad_norm": 0.6109854578971863,
+      "learning_rate": 7.258828048236717e-06,
+      "loss": 0.539,
+      "step": 2153
+    },
+    {
+      "epoch": 0.416191672302193,
+      "grad_norm": 0.620019257068634,
+      "learning_rate": 7.25581902247381e-06,
+      "loss": 0.5593,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4163848903487586,
+      "grad_norm": 0.6232174038887024,
+      "learning_rate": 7.252808970574688e-06,
+      "loss": 0.622,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4165781083953241,
+      "grad_norm": 0.6551336646080017,
+      "learning_rate": 7.249797893908575e-06,
+      "loss": 0.6308,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4167713264418897,
+      "grad_norm": 0.6861372590065002,
+      "learning_rate": 7.2467857938451655e-06,
+      "loss": 0.6208,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4169645444884552,
+      "grad_norm": 0.6400995850563049,
+      "learning_rate": 7.243772671754612e-06,
+      "loss": 0.6581,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4171577625350208,
+      "grad_norm": 0.5969094038009644,
+      "learning_rate": 7.240758529007538e-06,
+      "loss": 0.5577,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4173509805815863,
+      "grad_norm": 0.6593167781829834,
+      "learning_rate": 7.237743366975027e-06,
+      "loss": 0.6412,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4175441986281519,
+      "grad_norm": 0.7100246548652649,
+      "learning_rate": 7.234727187028631e-06,
+      "loss": 0.7148,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4177374166747174,
+      "grad_norm": 0.6505737900733948,
+      "learning_rate": 7.231709990540362e-06,
+      "loss": 0.59,
+      "step": 2162
+    },
+    {
+      "epoch": 0.41793063472128295,
+      "grad_norm": 0.6370893120765686,
+      "learning_rate": 7.2286917788826926e-06,
+      "loss": 0.5779,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4181238527678485,
+      "grad_norm": 0.6972851753234863,
+      "learning_rate": 7.22567255342856e-06,
+      "loss": 0.6748,
+      "step": 2164
+    },
+    {
+      "epoch": 0.41831707081441405,
+      "grad_norm": 0.7022752165794373,
+      "learning_rate": 7.222652315551363e-06,
+      "loss": 0.6013,
+      "step": 2165
+    },
+    {
+      "epoch": 0.41851028886097963,
+      "grad_norm": 0.6806923151016235,
+      "learning_rate": 7.219631066624959e-06,
+      "loss": 0.6588,
+      "step": 2166
+    },
+    {
+      "epoch": 0.41870350690754515,
+      "grad_norm": 0.6888837814331055,
+      "learning_rate": 7.2166088080236655e-06,
+      "loss": 0.6176,
+      "step": 2167
+    },
+    {
+      "epoch": 0.41889672495411073,
+      "grad_norm": 0.6800544261932373,
+      "learning_rate": 7.213585541122261e-06,
+      "loss": 0.6188,
+      "step": 2168
+    },
+    {
+      "epoch": 0.41908994300067626,
+      "grad_norm": 0.659703254699707,
+      "learning_rate": 7.2105612672959815e-06,
+      "loss": 0.588,
+      "step": 2169
+    },
+    {
+      "epoch": 0.41928316104724184,
+      "grad_norm": 0.628434419631958,
+      "learning_rate": 7.20753598792052e-06,
+      "loss": 0.6503,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41947637909380736,
+      "grad_norm": 0.6626314520835876,
+      "learning_rate": 7.20450970437203e-06,
+      "loss": 0.6079,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4196695971403729,
+      "grad_norm": 0.6524759531021118,
+      "learning_rate": 7.201482418027117e-06,
+      "loss": 0.534,
+      "step": 2172
+    },
+    {
+      "epoch": 0.41986281518693847,
+      "grad_norm": 0.6622713804244995,
+      "learning_rate": 7.198454130262849e-06,
+      "loss": 0.6368,
+      "step": 2173
+    },
+    {
+      "epoch": 0.420056033233504,
+      "grad_norm": 0.6918185949325562,
+      "learning_rate": 7.195424842456742e-06,
+      "loss": 0.6105,
+      "step": 2174
+    },
+    {
+      "epoch": 0.42024925128006957,
+      "grad_norm": 0.6403723955154419,
+      "learning_rate": 7.1923945559867745e-06,
+      "loss": 0.6633,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4204424693266351,
+      "grad_norm": 0.7196930646896362,
+      "learning_rate": 7.189363272231372e-06,
+      "loss": 0.6753,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4206356873732007,
+      "grad_norm": 0.6928418278694153,
+      "learning_rate": 7.186330992569421e-06,
+      "loss": 0.5721,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4208289054197662,
+      "grad_norm": 0.6016371846199036,
+      "learning_rate": 7.183297718380254e-06,
+      "loss": 0.526,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4210221234663318,
+      "grad_norm": 0.6343263983726501,
+      "learning_rate": 7.180263451043662e-06,
+      "loss": 0.5798,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4212153415128973,
+      "grad_norm": 0.6860288977622986,
+      "learning_rate": 7.177228191939884e-06,
+      "loss": 0.6443,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4214085595594629,
+      "grad_norm": 0.628670871257782,
+      "learning_rate": 7.17419194244961e-06,
+      "loss": 0.5688,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4216017776060284,
+      "grad_norm": 0.7133058905601501,
+      "learning_rate": 7.171154703953983e-06,
+      "loss": 0.6235,
+      "step": 2182
+    },
+    {
+      "epoch": 0.42179499565259393,
+      "grad_norm": 0.6948515176773071,
+      "learning_rate": 7.168116477834595e-06,
+      "loss": 0.5937,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4219882136991595,
+      "grad_norm": 0.7261781096458435,
+      "learning_rate": 7.165077265473483e-06,
+      "loss": 0.5912,
+      "step": 2184
+    },
+    {
+      "epoch": 0.42218143174572503,
+      "grad_norm": 0.6915099620819092,
+      "learning_rate": 7.162037068253141e-06,
+      "loss": 0.6649,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4223746497922906,
+      "grad_norm": 0.6243175864219666,
+      "learning_rate": 7.158995887556504e-06,
+      "loss": 0.5765,
+      "step": 2186
+    },
+    {
+      "epoch": 0.42256786783885614,
+      "grad_norm": 0.6435629725456238,
+      "learning_rate": 7.155953724766958e-06,
+      "loss": 0.6424,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4227610858854217,
+      "grad_norm": 0.6248393058776855,
+      "learning_rate": 7.152910581268333e-06,
+      "loss": 0.6487,
+      "step": 2188
+    },
+    {
+      "epoch": 0.42295430393198724,
+      "grad_norm": 0.6671506762504578,
+      "learning_rate": 7.149866458444908e-06,
+      "loss": 0.648,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4231475219785528,
+      "grad_norm": 0.6694079041481018,
+      "learning_rate": 7.146821357681406e-06,
+      "loss": 0.5912,
+      "step": 2190
+    },
+    {
+      "epoch": 0.42334074002511835,
+      "grad_norm": 0.7419486045837402,
+      "learning_rate": 7.1437752803629945e-06,
+      "loss": 0.6161,
+      "step": 2191
+    },
+    {
+      "epoch": 0.42353395807168387,
+      "grad_norm": 0.7070165872573853,
+      "learning_rate": 7.140728227875284e-06,
+      "loss": 0.682,
+      "step": 2192
+    },
+    {
+      "epoch": 0.42372717611824945,
+      "grad_norm": 0.6788286566734314,
+      "learning_rate": 7.137680201604333e-06,
+      "loss": 0.622,
+      "step": 2193
+    },
+    {
+      "epoch": 0.423920394164815,
+      "grad_norm": 0.645447850227356,
+      "learning_rate": 7.134631202936639e-06,
+      "loss": 0.622,
+      "step": 2194
+    },
+    {
+      "epoch": 0.42411361221138055,
+      "grad_norm": 0.6706958413124084,
+      "learning_rate": 7.131581233259142e-06,
+      "loss": 0.6348,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4243068302579461,
+      "grad_norm": 0.6812925338745117,
+      "learning_rate": 7.128530293959226e-06,
+      "loss": 0.6568,
+      "step": 2196
+    },
+    {
+      "epoch": 0.42450004830451166,
+      "grad_norm": 0.6441834568977356,
+      "learning_rate": 7.125478386424716e-06,
+      "loss": 0.6031,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4246932663510772,
+      "grad_norm": 0.7349604368209839,
+      "learning_rate": 7.122425512043874e-06,
+      "loss": 0.6334,
+      "step": 2198
+    },
+    {
+      "epoch": 0.42488648439764276,
+      "grad_norm": 0.6573999524116516,
+      "learning_rate": 7.119371672205405e-06,
+      "loss": 0.5997,
+      "step": 2199
+    },
+    {
+      "epoch": 0.4250797024442083,
+      "grad_norm": 0.6682370901107788,
+      "learning_rate": 7.116316868298451e-06,
+      "loss": 0.6486,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42527292049077386,
+      "grad_norm": 0.7391337156295776,
+      "learning_rate": 7.113261101712595e-06,
+      "loss": 0.6612,
+      "step": 2201
+    },
+    {
+      "epoch": 0.4254661385373394,
+      "grad_norm": 0.7007299065589905,
+      "learning_rate": 7.110204373837857e-06,
+      "loss": 0.6259,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4256593565839049,
+      "grad_norm": 0.6342196464538574,
+      "learning_rate": 7.107146686064694e-06,
+      "loss": 0.6518,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4258525746304705,
+      "grad_norm": 0.6857762932777405,
+      "learning_rate": 7.104088039783998e-06,
+      "loss": 0.6469,
+      "step": 2204
+    },
+    {
+      "epoch": 0.426045792677036,
+      "grad_norm": 0.6536127328872681,
+      "learning_rate": 7.101028436387099e-06,
+      "loss": 0.6085,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4262390107236016,
+      "grad_norm": 0.6482356786727905,
+      "learning_rate": 7.097967877265763e-06,
+      "loss": 0.5978,
+      "step": 2206
+    },
+    {
+      "epoch": 0.4264322287701671,
+      "grad_norm": 0.6724984645843506,
+      "learning_rate": 7.094906363812189e-06,
+      "loss": 0.6417,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4266254468167327,
+      "grad_norm": 0.683216392993927,
+      "learning_rate": 7.091843897419011e-06,
+      "loss": 0.5976,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4268186648632982,
+      "grad_norm": 0.650254487991333,
+      "learning_rate": 7.088780479479296e-06,
+      "loss": 0.6086,
+      "step": 2209
+    },
+    {
+      "epoch": 0.4270118829098638,
+      "grad_norm": 0.6273249983787537,
+      "learning_rate": 7.085716111386545e-06,
+      "loss": 0.7024,
+      "step": 2210
+    },
+    {
+      "epoch": 0.42720510095642933,
+      "grad_norm": 0.5989320874214172,
+      "learning_rate": 7.082650794534689e-06,
+      "loss": 0.523,
+      "step": 2211
+    },
+    {
+      "epoch": 0.42739831900299485,
+      "grad_norm": 0.7161542773246765,
+      "learning_rate": 7.0795845303180934e-06,
+      "loss": 0.6386,
+      "step": 2212
+    },
+    {
+      "epoch": 0.42759153704956043,
+      "grad_norm": 0.6225477457046509,
+      "learning_rate": 7.076517320131549e-06,
+      "loss": 0.5693,
+      "step": 2213
+    },
+    {
+      "epoch": 0.42778475509612596,
+      "grad_norm": 0.6379281282424927,
+      "learning_rate": 7.0734491653702875e-06,
+      "loss": 0.6376,
+      "step": 2214
+    },
+    {
+      "epoch": 0.42797797314269154,
+      "grad_norm": 0.7030832171440125,
+      "learning_rate": 7.070380067429961e-06,
+      "loss": 0.681,
+      "step": 2215
+    },
+    {
+      "epoch": 0.42817119118925706,
+      "grad_norm": 0.6540177464485168,
+      "learning_rate": 7.06731002770665e-06,
+      "loss": 0.52,
+      "step": 2216
+    },
+    {
+      "epoch": 0.42836440923582264,
+      "grad_norm": 0.6766851544380188,
+      "learning_rate": 7.06423904759687e-06,
+      "loss": 0.6057,
+      "step": 2217
+    },
+    {
+      "epoch": 0.42855762728238816,
+      "grad_norm": 0.6413094997406006,
+      "learning_rate": 7.061167128497561e-06,
+      "loss": 0.5766,
+      "step": 2218
+    },
+    {
+      "epoch": 0.42875084532895374,
+      "grad_norm": 0.6523410081863403,
+      "learning_rate": 7.058094271806091e-06,
+      "loss": 0.6313,
+      "step": 2219
+    },
+    {
+      "epoch": 0.42894406337551927,
+      "grad_norm": 0.642035186290741,
+      "learning_rate": 7.0550204789202515e-06,
+      "loss": 0.58,
+      "step": 2220
+    },
+    {
+      "epoch": 0.42913728142208485,
+      "grad_norm": 0.6962005496025085,
+      "learning_rate": 7.051945751238262e-06,
+      "loss": 0.6287,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4293304994686504,
+      "grad_norm": 0.7182521224021912,
+      "learning_rate": 7.048870090158769e-06,
+      "loss": 0.6367,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4295237175152159,
+      "grad_norm": 0.6890539526939392,
+      "learning_rate": 7.045793497080841e-06,
+      "loss": 0.6169,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4297169355617815,
+      "grad_norm": 0.6147426962852478,
+      "learning_rate": 7.042715973403971e-06,
+      "loss": 0.6367,
+      "step": 2224
+    },
+    {
+      "epoch": 0.429910153608347,
+      "grad_norm": 0.656883716583252,
+      "learning_rate": 7.039637520528077e-06,
+      "loss": 0.5812,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4301033716549126,
+      "grad_norm": 0.7737193703651428,
+      "learning_rate": 7.036558139853495e-06,
+      "loss": 0.6514,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4302965897014781,
+      "grad_norm": 0.6874740123748779,
+      "learning_rate": 7.03347783278099e-06,
+      "loss": 0.6231,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4304898077480437,
+      "grad_norm": 0.6342445015907288,
+      "learning_rate": 7.030396600711743e-06,
+      "loss": 0.5642,
+      "step": 2228
+    },
+    {
+      "epoch": 0.4306830257946092,
+      "grad_norm": 0.6689804792404175,
+      "learning_rate": 7.027314445047359e-06,
+      "loss": 0.614,
+      "step": 2229
+    },
+    {
+      "epoch": 0.4308762438411748,
+      "grad_norm": 0.6269736289978027,
+      "learning_rate": 7.024231367189858e-06,
+      "loss": 0.5662,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4310694618877403,
+      "grad_norm": 0.6717842817306519,
+      "learning_rate": 7.0211473685416855e-06,
+      "loss": 0.6267,
+      "step": 2231
+    },
+    {
+      "epoch": 0.43126267993430584,
+      "grad_norm": 0.6750964522361755,
+      "learning_rate": 7.018062450505704e-06,
+      "loss": 0.6323,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4314558979808714,
+      "grad_norm": 0.6598057150840759,
+      "learning_rate": 7.014976614485194e-06,
+      "loss": 0.5457,
+      "step": 2233
+    },
+    {
+      "epoch": 0.43164911602743694,
+      "grad_norm": 0.6753153800964355,
+      "learning_rate": 7.01188986188385e-06,
+      "loss": 0.5725,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4318423340740025,
+      "grad_norm": 0.6653080582618713,
+      "learning_rate": 7.00880219410579e-06,
+      "loss": 0.5592,
+      "step": 2235
+    },
+    {
+      "epoch": 0.43203555212056804,
+      "grad_norm": 0.6242656111717224,
+      "learning_rate": 7.0057136125555456e-06,
+      "loss": 0.6146,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4322287701671336,
+      "grad_norm": 0.6419099569320679,
+      "learning_rate": 7.0026241186380616e-06,
+      "loss": 0.6235,
+      "step": 2237
+    },
+    {
+      "epoch": 0.43242198821369915,
+      "grad_norm": 0.6492735147476196,
+      "learning_rate": 6.999533713758697e-06,
+      "loss": 0.6413,
+      "step": 2238
+    },
+    {
+      "epoch": 0.43261520626026473,
+      "grad_norm": 0.6392232775688171,
+      "learning_rate": 6.9964423993232335e-06,
+      "loss": 0.6004,
+      "step": 2239
+    },
+    {
+      "epoch": 0.43280842430683025,
+      "grad_norm": 0.7386446595191956,
+      "learning_rate": 6.993350176737857e-06,
+      "loss": 0.6264,
+      "step": 2240
+    },
+    {
+      "epoch": 0.43300164235339583,
+      "grad_norm": 0.6641836166381836,
+      "learning_rate": 6.990257047409172e-06,
+      "loss": 0.6227,
+      "step": 2241
+    },
+    {
+      "epoch": 0.43319486039996136,
+      "grad_norm": 0.6948634386062622,
+      "learning_rate": 6.987163012744193e-06,
+      "loss": 0.5927,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4333880784465269,
+      "grad_norm": 0.6931523680686951,
+      "learning_rate": 6.984068074150347e-06,
+      "loss": 0.6642,
+      "step": 2243
+    },
+    {
+      "epoch": 0.43358129649309246,
+      "grad_norm": 0.5653287768363953,
+      "learning_rate": 6.9809722330354756e-06,
+      "loss": 0.5876,
+      "step": 2244
+    },
+    {
+      "epoch": 0.433774514539658,
+      "grad_norm": 0.6146456599235535,
+      "learning_rate": 6.9778754908078225e-06,
+      "loss": 0.6132,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43396773258622356,
+      "grad_norm": 0.7078841924667358,
+      "learning_rate": 6.974777848876049e-06,
+      "loss": 0.6206,
+      "step": 2246
+    },
+    {
+      "epoch": 0.4341609506327891,
+      "grad_norm": 0.5965237021446228,
+      "learning_rate": 6.971679308649225e-06,
+      "loss": 0.5914,
+      "step": 2247
+    },
+    {
+      "epoch": 0.43435416867935467,
+      "grad_norm": 0.6646762490272522,
+      "learning_rate": 6.968579871536825e-06,
+      "loss": 0.6205,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4345473867259202,
+      "grad_norm": 0.6399896144866943,
+      "learning_rate": 6.965479538948734e-06,
+      "loss": 0.6073,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4347406047724858,
+      "grad_norm": 0.6456260681152344,
+      "learning_rate": 6.962378312295244e-06,
+      "loss": 0.5441,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4349338228190513,
+      "grad_norm": 0.5874357223510742,
+      "learning_rate": 6.959276192987053e-06,
+      "loss": 0.6129,
+      "step": 2251
+    },
+    {
+      "epoch": 0.4351270408656168,
+      "grad_norm": 0.6217551231384277,
+      "learning_rate": 6.956173182435268e-06,
+      "loss": 0.6275,
+      "step": 2252
+    },
+    {
+      "epoch": 0.4353202589121824,
+      "grad_norm": 0.6732254028320312,
+      "learning_rate": 6.953069282051397e-06,
+      "loss": 0.6435,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4355134769587479,
+      "grad_norm": 0.6581429243087769,
+      "learning_rate": 6.949964493247357e-06,
+      "loss": 0.5913,
+      "step": 2254
+    },
+    {
+      "epoch": 0.4357066950053135,
+      "grad_norm": 0.6331517696380615,
+      "learning_rate": 6.946858817435463e-06,
+      "loss": 0.6241,
+      "step": 2255
+    },
+    {
+      "epoch": 0.43589991305187903,
+      "grad_norm": 0.727761447429657,
+      "learning_rate": 6.943752256028443e-06,
+      "loss": 0.6544,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4360931310984446,
+      "grad_norm": 0.6671453714370728,
+      "learning_rate": 6.9406448104394185e-06,
+      "loss": 0.6468,
+      "step": 2257
+    },
+    {
+      "epoch": 0.43628634914501013,
+      "grad_norm": 0.6961790919303894,
+      "learning_rate": 6.93753648208192e-06,
+      "loss": 0.6337,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4364795671915757,
+      "grad_norm": 0.6253305077552795,
+      "learning_rate": 6.934427272369872e-06,
+      "loss": 0.5488,
+      "step": 2259
+    },
+    {
+      "epoch": 0.43667278523814124,
+      "grad_norm": 0.6961461305618286,
+      "learning_rate": 6.9313171827176104e-06,
+      "loss": 0.5885,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4368660032847068,
+      "grad_norm": 0.7538652420043945,
+      "learning_rate": 6.928206214539862e-06,
+      "loss": 0.654,
+      "step": 2261
+    },
+    {
+      "epoch": 0.43705922133127234,
+      "grad_norm": 0.6897897124290466,
+      "learning_rate": 6.925094369251757e-06,
+      "loss": 0.6145,
+      "step": 2262
+    },
+    {
+      "epoch": 0.43725243937783786,
+      "grad_norm": 0.669102132320404,
+      "learning_rate": 6.921981648268825e-06,
+      "loss": 0.6138,
+      "step": 2263
+    },
+    {
+      "epoch": 0.43744565742440344,
+      "grad_norm": 0.6945213079452515,
+      "learning_rate": 6.918868053006993e-06,
+      "loss": 0.6608,
+      "step": 2264
+    },
+    {
+      "epoch": 0.43763887547096897,
+      "grad_norm": 0.648155689239502,
+      "learning_rate": 6.915753584882587e-06,
+      "loss": 0.5497,
+      "step": 2265
+    },
+    {
+      "epoch": 0.43783209351753455,
+      "grad_norm": 0.7134128212928772,
+      "learning_rate": 6.912638245312328e-06,
+      "loss": 0.6703,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4380253115641001,
+      "grad_norm": 0.6517565250396729,
+      "learning_rate": 6.909522035713334e-06,
+      "loss": 0.6295,
+      "step": 2267
+    },
+    {
+      "epoch": 0.43821852961066565,
+      "grad_norm": 0.6422836780548096,
+      "learning_rate": 6.90640495750312e-06,
+      "loss": 0.6468,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4384117476572312,
+      "grad_norm": 0.7131401300430298,
+      "learning_rate": 6.903287012099595e-06,
+      "loss": 0.6238,
+      "step": 2269
+    },
+    {
+      "epoch": 0.43860496570379676,
+      "grad_norm": 0.6540867686271667,
+      "learning_rate": 6.900168200921065e-06,
+      "loss": 0.568,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4387981837503623,
+      "grad_norm": 0.6452621221542358,
+      "learning_rate": 6.897048525386223e-06,
+      "loss": 0.6204,
+      "step": 2271
+    },
+    {
+      "epoch": 0.43899140179692786,
+      "grad_norm": 0.7092915177345276,
+      "learning_rate": 6.893927986914164e-06,
+      "loss": 0.6115,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4391846198434934,
+      "grad_norm": 0.685431182384491,
+      "learning_rate": 6.89080658692437e-06,
+      "loss": 0.6334,
+      "step": 2273
+    },
+    {
+      "epoch": 0.4393778378900589,
+      "grad_norm": 0.7854279279708862,
+      "learning_rate": 6.8876843268367155e-06,
+      "loss": 0.66,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4395710559366245,
+      "grad_norm": 0.6293447613716125,
+      "learning_rate": 6.884561208071467e-06,
+      "loss": 0.6188,
+      "step": 2275
+    },
+    {
+      "epoch": 0.43976427398319,
+      "grad_norm": 0.6196038126945496,
+      "learning_rate": 6.8814372320492815e-06,
+      "loss": 0.5709,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4399574920297556,
+      "grad_norm": 0.666381299495697,
+      "learning_rate": 6.8783124001912075e-06,
+      "loss": 0.6584,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4401507100763211,
+      "grad_norm": 0.6302186250686646,
+      "learning_rate": 6.87518671391868e-06,
+      "loss": 0.6068,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4403439281228867,
+      "grad_norm": 0.7469252347946167,
+      "learning_rate": 6.872060174653525e-06,
+      "loss": 0.6522,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4405371461694522,
+      "grad_norm": 0.6278390288352966,
+      "learning_rate": 6.868932783817956e-06,
+      "loss": 0.5758,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4407303642160178,
+      "grad_norm": 0.6261811852455139,
+      "learning_rate": 6.865804542834571e-06,
+      "loss": 0.6337,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4409235822625833,
+      "grad_norm": 0.6248775124549866,
+      "learning_rate": 6.862675453126358e-06,
+      "loss": 0.5038,
+      "step": 2282
+    },
+    {
+      "epoch": 0.44111680030914885,
+      "grad_norm": 0.6978963017463684,
+      "learning_rate": 6.859545516116691e-06,
+      "loss": 0.639,
+      "step": 2283
+    },
+    {
+      "epoch": 0.44131001835571443,
+      "grad_norm": 0.6354128122329712,
+      "learning_rate": 6.856414733229328e-06,
+      "loss": 0.6036,
+      "step": 2284
+    },
+    {
+      "epoch": 0.44150323640227995,
+      "grad_norm": 0.6418664455413818,
+      "learning_rate": 6.853283105888413e-06,
+      "loss": 0.5885,
+      "step": 2285
+    },
+    {
+      "epoch": 0.44169645444884553,
+      "grad_norm": 0.6855936050415039,
+      "learning_rate": 6.850150635518475e-06,
+      "loss": 0.6589,
+      "step": 2286
+    },
+    {
+      "epoch": 0.44188967249541106,
+      "grad_norm": 0.686258852481842,
+      "learning_rate": 6.84701732354442e-06,
+      "loss": 0.6583,
+      "step": 2287
+    },
+    {
+      "epoch": 0.44208289054197664,
+      "grad_norm": 0.6340892314910889,
+      "learning_rate": 6.843883171391545e-06,
+      "loss": 0.5903,
+      "step": 2288
+    },
+    {
+      "epoch": 0.44227610858854216,
+      "grad_norm": 0.7037873864173889,
+      "learning_rate": 6.840748180485528e-06,
+      "loss": 0.57,
+      "step": 2289
+    },
+    {
+      "epoch": 0.44246932663510774,
+      "grad_norm": 0.7208712697029114,
+      "learning_rate": 6.837612352252423e-06,
+      "loss": 0.6593,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44266254468167326,
+      "grad_norm": 0.6642042398452759,
+      "learning_rate": 6.8344756881186704e-06,
+      "loss": 0.6687,
+      "step": 2291
+    },
+    {
+      "epoch": 0.44285576272823884,
+      "grad_norm": 0.6903440356254578,
+      "learning_rate": 6.831338189511087e-06,
+      "loss": 0.679,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44304898077480437,
+      "grad_norm": 0.6874251961708069,
+      "learning_rate": 6.828199857856873e-06,
+      "loss": 0.6067,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4432421988213699,
+      "grad_norm": 0.6496092081069946,
+      "learning_rate": 6.825060694583605e-06,
+      "loss": 0.6022,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4434354168679355,
+      "grad_norm": 0.64189612865448,
+      "learning_rate": 6.821920701119239e-06,
+      "loss": 0.633,
+      "step": 2295
+    },
+    {
+      "epoch": 0.443628634914501,
+      "grad_norm": 0.6302836537361145,
+      "learning_rate": 6.818779878892107e-06,
+      "loss": 0.653,
+      "step": 2296
+    },
+    {
+      "epoch": 0.4438218529610666,
+      "grad_norm": 0.5973377823829651,
+      "learning_rate": 6.815638229330919e-06,
+      "loss": 0.6055,
+      "step": 2297
+    },
+    {
+      "epoch": 0.4440150710076321,
+      "grad_norm": 0.6109885573387146,
+      "learning_rate": 6.812495753864762e-06,
+      "loss": 0.568,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4442082890541977,
+      "grad_norm": 0.6674631834030151,
+      "learning_rate": 6.8093524539231e-06,
+      "loss": 0.6312,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4444015071007632,
+      "grad_norm": 0.7032104730606079,
+      "learning_rate": 6.806208330935766e-06,
+      "loss": 0.6507,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4445947251473288,
+      "grad_norm": 0.6101316213607788,
+      "learning_rate": 6.803063386332975e-06,
+      "loss": 0.5469,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4447879431938943,
+      "grad_norm": 0.6730210781097412,
+      "learning_rate": 6.79991762154531e-06,
+      "loss": 0.5919,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44498116124045983,
+      "grad_norm": 0.6411260962486267,
+      "learning_rate": 6.796771038003732e-06,
+      "loss": 0.6264,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4451743792870254,
+      "grad_norm": 0.6522818803787231,
+      "learning_rate": 6.79362363713957e-06,
+      "loss": 0.6558,
+      "step": 2304
+    },
+    {
+      "epoch": 0.44536759733359094,
+      "grad_norm": 0.7033416032791138,
+      "learning_rate": 6.7904754203845265e-06,
+      "loss": 0.6875,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4455608153801565,
+      "grad_norm": 0.7068544626235962,
+      "learning_rate": 6.7873263891706756e-06,
+      "loss": 0.7005,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44575403342672204,
+      "grad_norm": 0.7388836741447449,
+      "learning_rate": 6.784176544930463e-06,
+      "loss": 0.6619,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4459472514732876,
+      "grad_norm": 0.6538330316543579,
+      "learning_rate": 6.781025889096702e-06,
+      "loss": 0.6267,
+      "step": 2308
+    },
+    {
+      "epoch": 0.44614046951985314,
+      "grad_norm": 0.6498190760612488,
+      "learning_rate": 6.777874423102573e-06,
+      "loss": 0.6132,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4463336875664187,
+      "grad_norm": 0.6766545176506042,
+      "learning_rate": 6.7747221483816315e-06,
+      "loss": 0.6236,
+      "step": 2310
+    },
+    {
+      "epoch": 0.44652690561298425,
+      "grad_norm": 0.6418600678443909,
+      "learning_rate": 6.771569066367796e-06,
+      "loss": 0.5556,
+      "step": 2311
+    },
+    {
+      "epoch": 0.44672012365954983,
+      "grad_norm": 0.6854467988014221,
+      "learning_rate": 6.7684151784953526e-06,
+      "loss": 0.6163,
+      "step": 2312
+    },
+    {
+      "epoch": 0.44691334170611535,
+      "grad_norm": 0.6886604428291321,
+      "learning_rate": 6.765260486198956e-06,
+      "loss": 0.6753,
+      "step": 2313
+    },
+    {
+      "epoch": 0.4471065597526809,
+      "grad_norm": 0.662898600101471,
+      "learning_rate": 6.762104990913624e-06,
+      "loss": 0.7081,
+      "step": 2314
+    },
+    {
+      "epoch": 0.44729977779924646,
+      "grad_norm": 0.7003542184829712,
+      "learning_rate": 6.758948694074743e-06,
+      "loss": 0.6468,
+      "step": 2315
+    },
+    {
+      "epoch": 0.447492995845812,
+      "grad_norm": 0.6622884273529053,
+      "learning_rate": 6.7557915971180625e-06,
+      "loss": 0.5965,
+      "step": 2316
+    },
+    {
+      "epoch": 0.44768621389237756,
+      "grad_norm": 0.6607614159584045,
+      "learning_rate": 6.752633701479693e-06,
+      "loss": 0.6227,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4478794319389431,
+      "grad_norm": 0.6702031493186951,
+      "learning_rate": 6.749475008596113e-06,
+      "loss": 0.6246,
+      "step": 2318
+    },
+    {
+      "epoch": 0.44807264998550866,
+      "grad_norm": 0.740099310874939,
+      "learning_rate": 6.746315519904162e-06,
+      "loss": 0.617,
+      "step": 2319
+    },
+    {
+      "epoch": 0.4482658680320742,
+      "grad_norm": 0.6476364731788635,
+      "learning_rate": 6.743155236841039e-06,
+      "loss": 0.5624,
+      "step": 2320
+    },
+    {
+      "epoch": 0.44845908607863977,
+      "grad_norm": 0.6241509318351746,
+      "learning_rate": 6.7399941608443096e-06,
+      "loss": 0.5792,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4486523041252053,
+      "grad_norm": 0.6858493685722351,
+      "learning_rate": 6.7368322933518945e-06,
+      "loss": 0.5862,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4488455221717708,
+      "grad_norm": 0.6926106810569763,
+      "learning_rate": 6.7336696358020806e-06,
+      "loss": 0.558,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4490387402183364,
+      "grad_norm": 0.6500473618507385,
+      "learning_rate": 6.730506189633506e-06,
+      "loss": 0.6336,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4492319582649019,
+      "grad_norm": 0.6627086400985718,
+      "learning_rate": 6.727341956285174e-06,
+      "loss": 0.6384,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4494251763114675,
+      "grad_norm": 0.592478334903717,
+      "learning_rate": 6.724176937196443e-06,
+      "loss": 0.5707,
+      "step": 2326
+    },
+    {
+      "epoch": 0.449618394358033,
+      "grad_norm": 0.655165433883667,
+      "learning_rate": 6.721011133807033e-06,
+      "loss": 0.63,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4498116124045986,
+      "grad_norm": 0.6117173433303833,
+      "learning_rate": 6.717844547557013e-06,
+      "loss": 0.6217,
+      "step": 2328
+    },
+    {
+      "epoch": 0.45000483045116413,
+      "grad_norm": 0.7115458846092224,
+      "learning_rate": 6.714677179886817e-06,
+      "loss": 0.6577,
+      "step": 2329
+    },
+    {
+      "epoch": 0.4501980484977297,
+      "grad_norm": 0.6163074374198914,
+      "learning_rate": 6.7115090322372274e-06,
+      "loss": 0.6119,
+      "step": 2330
+    },
+    {
+      "epoch": 0.45039126654429523,
+      "grad_norm": 0.6869255900382996,
+      "learning_rate": 6.708340106049387e-06,
+      "loss": 0.5915,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4505844845908608,
+      "grad_norm": 0.6662833094596863,
+      "learning_rate": 6.7051704027647875e-06,
+      "loss": 0.6723,
+      "step": 2332
+    },
+    {
+      "epoch": 0.45077770263742634,
+      "grad_norm": 0.7145954966545105,
+      "learning_rate": 6.701999923825279e-06,
+      "loss": 0.6419,
+      "step": 2333
+    },
+    {
+      "epoch": 0.45097092068399186,
+      "grad_norm": 0.6738686561584473,
+      "learning_rate": 6.698828670673058e-06,
+      "loss": 0.6563,
+      "step": 2334
+    },
+    {
+      "epoch": 0.45116413873055744,
+      "grad_norm": 0.61698317527771,
+      "learning_rate": 6.6956566447506795e-06,
+      "loss": 0.5583,
+      "step": 2335
+    },
+    {
+      "epoch": 0.45135735677712296,
+      "grad_norm": 0.6211065053939819,
+      "learning_rate": 6.692483847501048e-06,
+      "loss": 0.5416,
+      "step": 2336
+    },
+    {
+      "epoch": 0.45155057482368854,
+      "grad_norm": 0.7066130638122559,
+      "learning_rate": 6.689310280367418e-06,
+      "loss": 0.6128,
+      "step": 2337
+    },
+    {
+      "epoch": 0.45174379287025407,
+      "grad_norm": 0.7013624906539917,
+      "learning_rate": 6.686135944793395e-06,
+      "loss": 0.5734,
+      "step": 2338
+    },
+    {
+      "epoch": 0.45193701091681965,
+      "grad_norm": 0.6968424320220947,
+      "learning_rate": 6.682960842222934e-06,
+      "loss": 0.7019,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4521302289633852,
+      "grad_norm": 0.6719240546226501,
+      "learning_rate": 6.6797849741003386e-06,
+      "loss": 0.6666,
+      "step": 2340
+    },
+    {
+      "epoch": 0.45232344700995075,
+      "grad_norm": 0.673333466053009,
+      "learning_rate": 6.67660834187026e-06,
+      "loss": 0.5956,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4525166650565163,
+      "grad_norm": 0.6222085952758789,
+      "learning_rate": 6.673430946977697e-06,
+      "loss": 0.5648,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4527098831030818,
+      "grad_norm": 0.6951294541358948,
+      "learning_rate": 6.670252790867998e-06,
+      "loss": 0.6223,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4529031011496474,
+      "grad_norm": 0.6692245006561279,
+      "learning_rate": 6.667073874986856e-06,
+      "loss": 0.642,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4530963191962129,
+      "grad_norm": 0.6813640594482422,
+      "learning_rate": 6.663894200780308e-06,
+      "loss": 0.6567,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4532895372427785,
+      "grad_norm": 0.7091935276985168,
+      "learning_rate": 6.6607137696947355e-06,
+      "loss": 0.6386,
+      "step": 2346
+    },
+    {
+      "epoch": 0.453482755289344,
+      "grad_norm": 0.7109883427619934,
+      "learning_rate": 6.657532583176871e-06,
+      "loss": 0.6677,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4536759733359096,
+      "grad_norm": 0.6363698840141296,
+      "learning_rate": 6.654350642673781e-06,
+      "loss": 0.5767,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4538691913824751,
+      "grad_norm": 0.6721639633178711,
+      "learning_rate": 6.6511679496328815e-06,
+      "loss": 0.6895,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4540624094290407,
+      "grad_norm": 0.7455520033836365,
+      "learning_rate": 6.6479845055019286e-06,
+      "loss": 0.6948,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4542556274756062,
+      "grad_norm": 0.6837520599365234,
+      "learning_rate": 6.64480031172902e-06,
+      "loss": 0.5709,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4544488455221718,
+      "grad_norm": 0.6305144429206848,
+      "learning_rate": 6.641615369762596e-06,
+      "loss": 0.5837,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4546420635687373,
+      "grad_norm": 0.6131630539894104,
+      "learning_rate": 6.638429681051437e-06,
+      "loss": 0.5386,
+      "step": 2353
+    },
+    {
+      "epoch": 0.45483528161530284,
+      "grad_norm": 0.6883224248886108,
+      "learning_rate": 6.635243247044659e-06,
+      "loss": 0.651,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4550284996618684,
+      "grad_norm": 0.6923902034759521,
+      "learning_rate": 6.632056069191723e-06,
+      "loss": 0.6076,
+      "step": 2355
+    },
+    {
+      "epoch": 0.45522171770843395,
+      "grad_norm": 0.6675470471382141,
+      "learning_rate": 6.6288681489424235e-06,
+      "loss": 0.6424,
+      "step": 2356
+    },
+    {
+      "epoch": 0.45541493575499953,
+      "grad_norm": 0.8398420810699463,
+      "learning_rate": 6.6256794877468965e-06,
+      "loss": 0.6108,
+      "step": 2357
+    },
+    {
+      "epoch": 0.45560815380156505,
+      "grad_norm": 0.6172304153442383,
+      "learning_rate": 6.622490087055611e-06,
+      "loss": 0.7147,
+      "step": 2358
+    },
+    {
+      "epoch": 0.45580137184813063,
+      "grad_norm": 0.6278538703918457,
+      "learning_rate": 6.619299948319377e-06,
+      "loss": 0.6278,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45599458989469616,
+      "grad_norm": 0.6173030734062195,
+      "learning_rate": 6.616109072989337e-06,
+      "loss": 0.6553,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45618780794126174,
+      "grad_norm": 0.6741093993186951,
+      "learning_rate": 6.61291746251697e-06,
+      "loss": 0.6066,
+      "step": 2361
+    },
+    {
+      "epoch": 0.45638102598782726,
+      "grad_norm": 0.6110511422157288,
+      "learning_rate": 6.609725118354088e-06,
+      "loss": 0.5896,
+      "step": 2362
+    },
+    {
+      "epoch": 0.45657424403439284,
+      "grad_norm": 0.6776458621025085,
+      "learning_rate": 6.606532041952838e-06,
+      "loss": 0.5853,
+      "step": 2363
+    },
+    {
+      "epoch": 0.45676746208095836,
+      "grad_norm": 0.6435579657554626,
+      "learning_rate": 6.6033382347656995e-06,
+      "loss": 0.5896,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4569606801275239,
+      "grad_norm": 0.6974932551383972,
+      "learning_rate": 6.600143698245485e-06,
+      "loss": 0.6352,
+      "step": 2365
+    },
+    {
+      "epoch": 0.45715389817408947,
+      "grad_norm": 0.6639747023582458,
+      "learning_rate": 6.596948433845338e-06,
+      "loss": 0.6284,
+      "step": 2366
+    },
+    {
+      "epoch": 0.457347116220655,
+      "grad_norm": 0.6298837065696716,
+      "learning_rate": 6.593752443018733e-06,
+      "loss": 0.5955,
+      "step": 2367
+    },
+    {
+      "epoch": 0.45754033426722057,
+      "grad_norm": 0.6423676013946533,
+      "learning_rate": 6.5905557272194765e-06,
+      "loss": 0.6263,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4577335523137861,
+      "grad_norm": 0.6426789164543152,
+      "learning_rate": 6.587358287901704e-06,
+      "loss": 0.657,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4579267703603517,
+      "grad_norm": 0.6405795216560364,
+      "learning_rate": 6.584160126519879e-06,
+      "loss": 0.5999,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4581199884069172,
+      "grad_norm": 0.6356592774391174,
+      "learning_rate": 6.580961244528794e-06,
+      "loss": 0.5807,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4583132064534828,
+      "grad_norm": 0.6279131174087524,
+      "learning_rate": 6.57776164338357e-06,
+      "loss": 0.5971,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4585064245000483,
+      "grad_norm": 0.6761428713798523,
+      "learning_rate": 6.5745613245396544e-06,
+      "loss": 0.6392,
+      "step": 2373
+    },
+    {
+      "epoch": 0.45869964254661383,
+      "grad_norm": 0.7186028957366943,
+      "learning_rate": 6.571360289452822e-06,
+      "loss": 0.6332,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4588928605931794,
+      "grad_norm": 0.6575113534927368,
+      "learning_rate": 6.568158539579173e-06,
+      "loss": 0.6486,
+      "step": 2375
+    },
+    {
+      "epoch": 0.45908607863974493,
+      "grad_norm": 0.6023368835449219,
+      "learning_rate": 6.564956076375131e-06,
+      "loss": 0.5687,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4592792966863105,
+      "grad_norm": 0.6503453254699707,
+      "learning_rate": 6.561752901297447e-06,
+      "loss": 0.6096,
+      "step": 2377
+    },
+    {
+      "epoch": 0.45947251473287604,
+      "grad_norm": 0.6727077960968018,
+      "learning_rate": 6.558549015803193e-06,
+      "loss": 0.6027,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4596657327794416,
+      "grad_norm": 0.6469513177871704,
+      "learning_rate": 6.555344421349768e-06,
+      "loss": 0.6402,
+      "step": 2379
+    },
+    {
+      "epoch": 0.45985895082600714,
+      "grad_norm": 0.6620474457740784,
+      "learning_rate": 6.552139119394888e-06,
+      "loss": 0.6082,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4600521688725727,
+      "grad_norm": 0.6391301155090332,
+      "learning_rate": 6.548933111396596e-06,
+      "loss": 0.6282,
+      "step": 2381
+    },
+    {
+      "epoch": 0.46024538691913824,
+      "grad_norm": 0.6321401000022888,
+      "learning_rate": 6.545726398813253e-06,
+      "loss": 0.6891,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4604386049657038,
+      "grad_norm": 0.7132486701011658,
+      "learning_rate": 6.542518983103544e-06,
+      "loss": 0.6252,
+      "step": 2383
+    },
+    {
+      "epoch": 0.46063182301226935,
+      "grad_norm": 0.655623197555542,
+      "learning_rate": 6.539310865726468e-06,
+      "loss": 0.5903,
+      "step": 2384
+    },
+    {
+      "epoch": 0.46082504105883487,
+      "grad_norm": 0.639884352684021,
+      "learning_rate": 6.536102048141349e-06,
+      "loss": 0.6029,
+      "step": 2385
+    },
+    {
+      "epoch": 0.46101825910540045,
+      "grad_norm": 0.6373783946037292,
+      "learning_rate": 6.532892531807827e-06,
+      "loss": 0.6414,
+      "step": 2386
+    },
+    {
+      "epoch": 0.461211477151966,
+      "grad_norm": 0.7245307564735413,
+      "learning_rate": 6.529682318185857e-06,
+      "loss": 0.5538,
+      "step": 2387
+    },
+    {
+      "epoch": 0.46140469519853156,
+      "grad_norm": 0.7143988013267517,
+      "learning_rate": 6.52647140873572e-06,
+      "loss": 0.6142,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4615979132450971,
+      "grad_norm": 0.6598471999168396,
+      "learning_rate": 6.523259804918001e-06,
+      "loss": 0.6187,
+      "step": 2389
+    },
+    {
+      "epoch": 0.46179113129166266,
+      "grad_norm": 0.6982096433639526,
+      "learning_rate": 6.520047508193611e-06,
+      "loss": 0.6425,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4619843493382282,
+      "grad_norm": 0.6761695742607117,
+      "learning_rate": 6.5168345200237735e-06,
+      "loss": 0.5555,
+      "step": 2391
+    },
+    {
+      "epoch": 0.46217756738479376,
+      "grad_norm": 0.6155459880828857,
+      "learning_rate": 6.513620841870023e-06,
+      "loss": 0.5941,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4623707854313593,
+      "grad_norm": 0.6917835474014282,
+      "learning_rate": 6.510406475194213e-06,
+      "loss": 0.6758,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4625640034779248,
+      "grad_norm": 0.6930914521217346,
+      "learning_rate": 6.507191421458508e-06,
+      "loss": 0.6344,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4627572215244904,
+      "grad_norm": 0.5480935573577881,
+      "learning_rate": 6.503975682125382e-06,
+      "loss": 0.5291,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4629504395710559,
+      "grad_norm": 0.728971540927887,
+      "learning_rate": 6.500759258657625e-06,
+      "loss": 0.6601,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4631436576176215,
+      "grad_norm": 1.6462327241897583,
+      "learning_rate": 6.497542152518339e-06,
+      "loss": 0.6336,
+      "step": 2397
+    },
+    {
+      "epoch": 0.463336875664187,
+      "grad_norm": 0.6249005198478699,
+      "learning_rate": 6.494324365170933e-06,
+      "loss": 0.6218,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4635300937107526,
+      "grad_norm": 0.6366087794303894,
+      "learning_rate": 6.491105898079128e-06,
+      "loss": 0.5863,
+      "step": 2399
+    },
+    {
+      "epoch": 0.4637233117573181,
+      "grad_norm": 0.645389199256897,
+      "learning_rate": 6.487886752706951e-06,
+      "loss": 0.6429,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4639165298038837,
+      "grad_norm": 0.6958004832267761,
+      "learning_rate": 6.484666930518742e-06,
+      "loss": 0.5804,
+      "step": 2401
+    },
+    {
+      "epoch": 0.46410974785044923,
+      "grad_norm": 0.7100715041160583,
+      "learning_rate": 6.481446432979147e-06,
+      "loss": 0.5929,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4643029658970148,
+      "grad_norm": 0.6242349743843079,
+      "learning_rate": 6.4782252615531186e-06,
+      "loss": 0.6074,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46449618394358033,
+      "grad_norm": 0.7127718925476074,
+      "learning_rate": 6.475003417705917e-06,
+      "loss": 0.7016,
+      "step": 2404
+    },
+    {
+      "epoch": 0.46468940199014586,
+      "grad_norm": 0.7256423234939575,
+      "learning_rate": 6.4717809029031065e-06,
+      "loss": 0.6548,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46488262003671144,
+      "grad_norm": 0.7284982204437256,
+      "learning_rate": 6.4685577186105595e-06,
+      "loss": 0.6488,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46507583808327696,
+      "grad_norm": 0.6454406380653381,
+      "learning_rate": 6.465333866294452e-06,
+      "loss": 0.5643,
+      "step": 2407
+    },
+    {
+      "epoch": 0.46526905612984254,
+      "grad_norm": 0.7150859832763672,
+      "learning_rate": 6.462109347421261e-06,
+      "loss": 0.6502,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46546227417640806,
+      "grad_norm": 0.6539508700370789,
+      "learning_rate": 6.458884163457768e-06,
+      "loss": 0.5969,
+      "step": 2409
+    },
+    {
+      "epoch": 0.46565549222297364,
+      "grad_norm": 0.656039834022522,
+      "learning_rate": 6.455658315871062e-06,
+      "loss": 0.5519,
+      "step": 2410
+    },
+    {
+      "epoch": 0.46584871026953917,
+      "grad_norm": 0.675858199596405,
+      "learning_rate": 6.452431806128525e-06,
+      "loss": 0.6369,
+      "step": 2411
+    },
+    {
+      "epoch": 0.46604192831610475,
+      "grad_norm": 0.7317296266555786,
+      "learning_rate": 6.449204635697848e-06,
+      "loss": 0.6454,
+      "step": 2412
+    },
+    {
+      "epoch": 0.46623514636267027,
+      "grad_norm": 0.7116714715957642,
+      "learning_rate": 6.445976806047019e-06,
+      "loss": 0.6043,
+      "step": 2413
+    },
+    {
+      "epoch": 0.4664283644092358,
+      "grad_norm": 0.6627038717269897,
+      "learning_rate": 6.442748318644327e-06,
+      "loss": 0.6505,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4666215824558014,
+      "grad_norm": 0.6766278743743896,
+      "learning_rate": 6.4395191749583596e-06,
+      "loss": 0.6711,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4668148005023669,
+      "grad_norm": 0.6602400541305542,
+      "learning_rate": 6.436289376458002e-06,
+      "loss": 0.6558,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4670080185489325,
+      "grad_norm": 0.6745182871818542,
+      "learning_rate": 6.433058924612436e-06,
+      "loss": 0.5498,
+      "step": 2417
+    },
+    {
+      "epoch": 0.467201236595498,
+      "grad_norm": 0.7361905574798584,
+      "learning_rate": 6.429827820891147e-06,
+      "loss": 0.6276,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4673944546420636,
+      "grad_norm": 0.7023398876190186,
+      "learning_rate": 6.4265960667639106e-06,
+      "loss": 0.6709,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4675876726886291,
+      "grad_norm": 0.7385457754135132,
+      "learning_rate": 6.423363663700801e-06,
+      "loss": 0.6636,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4677808907351947,
+      "grad_norm": 0.6799448728561401,
+      "learning_rate": 6.420130613172186e-06,
+      "loss": 0.5998,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4679741087817602,
+      "grad_norm": 0.752800464630127,
+      "learning_rate": 6.416896916648731e-06,
+      "loss": 0.5537,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4681673268283258,
+      "grad_norm": 0.602005124092102,
+      "learning_rate": 6.413662575601391e-06,
+      "loss": 0.5491,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4683605448748913,
+      "grad_norm": 0.6634283661842346,
+      "learning_rate": 6.410427591501418e-06,
+      "loss": 0.6655,
+      "step": 2424
+    },
+    {
+      "epoch": 0.46855376292145684,
+      "grad_norm": 0.6522394418716431,
+      "learning_rate": 6.407191965820352e-06,
+      "loss": 0.5905,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4687469809680224,
+      "grad_norm": 0.635780394077301,
+      "learning_rate": 6.403955700030029e-06,
+      "loss": 0.5703,
+      "step": 2426
+    },
+    {
+      "epoch": 0.46894019901458794,
+      "grad_norm": 0.6499520540237427,
+      "learning_rate": 6.400718795602578e-06,
+      "loss": 0.6324,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4691334170611535,
+      "grad_norm": 0.663664698600769,
+      "learning_rate": 6.397481254010411e-06,
+      "loss": 0.5639,
+      "step": 2428
+    },
+    {
+      "epoch": 0.46932663510771905,
+      "grad_norm": 0.7018650770187378,
+      "learning_rate": 6.394243076726237e-06,
+      "loss": 0.6354,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4695198531542846,
+      "grad_norm": 0.7016417980194092,
+      "learning_rate": 6.39100426522305e-06,
+      "loss": 0.608,
+      "step": 2430
+    },
+    {
+      "epoch": 0.46971307120085015,
+      "grad_norm": 0.7435153722763062,
+      "learning_rate": 6.387764820974135e-06,
+      "loss": 0.6294,
+      "step": 2431
+    },
+    {
+      "epoch": 0.46990628924741573,
+      "grad_norm": 0.7134484648704529,
+      "learning_rate": 6.3845247454530626e-06,
+      "loss": 0.6097,
+      "step": 2432
+    },
+    {
+      "epoch": 0.47009950729398126,
+      "grad_norm": 0.6787815690040588,
+      "learning_rate": 6.3812840401336905e-06,
+      "loss": 0.584,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4702927253405468,
+      "grad_norm": 0.6415289640426636,
+      "learning_rate": 6.378042706490166e-06,
+      "loss": 0.6443,
+      "step": 2434
+    },
+    {
+      "epoch": 0.47048594338711236,
+      "grad_norm": 0.7101773023605347,
+      "learning_rate": 6.3748007459969205e-06,
+      "loss": 0.6613,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4706791614336779,
+      "grad_norm": 0.7126517295837402,
+      "learning_rate": 6.37155816012867e-06,
+      "loss": 0.6619,
+      "step": 2436
+    },
+    {
+      "epoch": 0.47087237948024346,
+      "grad_norm": 0.6302406191825867,
+      "learning_rate": 6.368314950360416e-06,
+      "loss": 0.5747,
+      "step": 2437
+    },
+    {
+      "epoch": 0.471065597526809,
+      "grad_norm": 0.7152489423751831,
+      "learning_rate": 6.36507111816744e-06,
+      "loss": 0.5715,
+      "step": 2438
+    },
+    {
+      "epoch": 0.47125881557337457,
+      "grad_norm": 0.6821061968803406,
+      "learning_rate": 6.361826665025311e-06,
+      "loss": 0.595,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4714520336199401,
+      "grad_norm": 0.6517532467842102,
+      "learning_rate": 6.358581592409881e-06,
+      "loss": 0.6513,
+      "step": 2440
+    },
+    {
+      "epoch": 0.47164525166650567,
+      "grad_norm": 0.6567256450653076,
+      "learning_rate": 6.35533590179728e-06,
+      "loss": 0.6492,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4718384697130712,
+      "grad_norm": 0.6566718816757202,
+      "learning_rate": 6.352089594663923e-06,
+      "loss": 0.6234,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4720316877596368,
+      "grad_norm": 0.6848618388175964,
+      "learning_rate": 6.348842672486502e-06,
+      "loss": 0.6092,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4722249058062023,
+      "grad_norm": 0.6409578323364258,
+      "learning_rate": 6.34559513674199e-06,
+      "loss": 0.5593,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4724181238527678,
+      "grad_norm": 0.7228472828865051,
+      "learning_rate": 6.342346988907641e-06,
+      "loss": 0.6764,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4726113418993334,
+      "grad_norm": 0.6219584345817566,
+      "learning_rate": 6.339098230460983e-06,
+      "loss": 0.6268,
+      "step": 2446
+    },
+    {
+      "epoch": 0.47280455994589893,
+      "grad_norm": 0.5764530301094055,
+      "learning_rate": 6.3358488628798275e-06,
+      "loss": 0.5552,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4729977779924645,
+      "grad_norm": 0.6661320924758911,
+      "learning_rate": 6.3325988876422585e-06,
+      "loss": 0.6346,
+      "step": 2448
+    },
+    {
+      "epoch": 0.47319099603903003,
+      "grad_norm": 0.6808934807777405,
+      "learning_rate": 6.329348306226639e-06,
+      "loss": 0.6339,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4733842140855956,
+      "grad_norm": 0.6244999766349792,
+      "learning_rate": 6.326097120111608e-06,
+      "loss": 0.6469,
+      "step": 2450
+    },
+    {
+      "epoch": 0.47357743213216114,
+      "grad_norm": 0.6576670408248901,
+      "learning_rate": 6.322845330776074e-06,
+      "loss": 0.6134,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4737706501787267,
+      "grad_norm": 0.647864818572998,
+      "learning_rate": 6.319592939699229e-06,
+      "loss": 0.6261,
+      "step": 2452
+    },
+    {
+      "epoch": 0.47396386822529224,
+      "grad_norm": 0.7111817002296448,
+      "learning_rate": 6.316339948360532e-06,
+      "loss": 0.5725,
+      "step": 2453
+    },
+    {
+      "epoch": 0.47415708627185776,
+      "grad_norm": 0.7058026194572449,
+      "learning_rate": 6.313086358239716e-06,
+      "loss": 0.611,
+      "step": 2454
+    },
+    {
+      "epoch": 0.47435030431842334,
+      "grad_norm": 0.7051585912704468,
+      "learning_rate": 6.3098321708167885e-06,
+      "loss": 0.6025,
+      "step": 2455
+    },
+    {
+      "epoch": 0.47454352236498887,
+      "grad_norm": 0.6382871866226196,
+      "learning_rate": 6.306577387572029e-06,
+      "loss": 0.6546,
+      "step": 2456
+    },
+    {
+      "epoch": 0.47473674041155445,
+      "grad_norm": 0.6691299676895142,
+      "learning_rate": 6.303322009985984e-06,
+      "loss": 0.6508,
+      "step": 2457
+    },
+    {
+      "epoch": 0.47492995845811997,
+      "grad_norm": 0.6618803143501282,
+      "learning_rate": 6.3000660395394755e-06,
+      "loss": 0.6067,
+      "step": 2458
+    },
+    {
+      "epoch": 0.47512317650468555,
+      "grad_norm": 0.6986579298973083,
+      "learning_rate": 6.2968094777135905e-06,
+      "loss": 0.6059,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4753163945512511,
+      "grad_norm": 0.7025730013847351,
+      "learning_rate": 6.293552325989689e-06,
+      "loss": 0.6168,
+      "step": 2460
+    },
+    {
+      "epoch": 0.47550961259781666,
+      "grad_norm": 0.6719696521759033,
+      "learning_rate": 6.290294585849397e-06,
+      "loss": 0.6412,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4757028306443822,
+      "grad_norm": 0.6434871554374695,
+      "learning_rate": 6.287036258774607e-06,
+      "loss": 0.5824,
+      "step": 2462
+    },
+    {
+      "epoch": 0.47589604869094776,
+      "grad_norm": 0.6317356824874878,
+      "learning_rate": 6.283777346247482e-06,
+      "loss": 0.6039,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4760892667375133,
+      "grad_norm": 0.6656129360198975,
+      "learning_rate": 6.280517849750452e-06,
+      "loss": 0.6083,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4762824847840788,
+      "grad_norm": 0.6845144629478455,
+      "learning_rate": 6.277257770766206e-06,
+      "loss": 0.6502,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4764757028306444,
+      "grad_norm": 0.6336469054222107,
+      "learning_rate": 6.273997110777702e-06,
+      "loss": 0.6024,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4766689208772099,
+      "grad_norm": 0.6557308435440063,
+      "learning_rate": 6.270735871268164e-06,
+      "loss": 0.6193,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4768621389237755,
+      "grad_norm": 0.6701752543449402,
+      "learning_rate": 6.267474053721079e-06,
+      "loss": 0.5871,
+      "step": 2468
+    },
+    {
+      "epoch": 0.477055356970341,
+      "grad_norm": 0.7117101550102234,
+      "learning_rate": 6.264211659620195e-06,
+      "loss": 0.5918,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4772485750169066,
+      "grad_norm": 0.6677953004837036,
+      "learning_rate": 6.260948690449522e-06,
+      "loss": 0.6156,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4774417930634721,
+      "grad_norm": 0.7027629613876343,
+      "learning_rate": 6.257685147693333e-06,
+      "loss": 0.628,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4776350111100377,
+      "grad_norm": 0.6838759183883667,
+      "learning_rate": 6.254421032836165e-06,
+      "loss": 0.5896,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4778282291566032,
+      "grad_norm": 0.7434727549552917,
+      "learning_rate": 6.251156347362809e-06,
+      "loss": 0.6584,
+      "step": 2473
+    },
+    {
+      "epoch": 0.4780214472031688,
+      "grad_norm": 0.6320463418960571,
+      "learning_rate": 6.247891092758319e-06,
+      "loss": 0.6311,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4782146652497343,
+      "grad_norm": 0.6417965292930603,
+      "learning_rate": 6.244625270508008e-06,
+      "loss": 0.6435,
+      "step": 2475
+    },
+    {
+      "epoch": 0.47840788329629985,
+      "grad_norm": 0.7013752460479736,
+      "learning_rate": 6.241358882097443e-06,
+      "loss": 0.6708,
+      "step": 2476
+    },
+    {
+      "epoch": 0.47860110134286543,
+      "grad_norm": 0.6276121139526367,
+      "learning_rate": 6.2380919290124585e-06,
+      "loss": 0.5912,
+      "step": 2477
+    },
+    {
+      "epoch": 0.47879431938943096,
+      "grad_norm": 0.675273060798645,
+      "learning_rate": 6.234824412739135e-06,
+      "loss": 0.6302,
+      "step": 2478
+    },
+    {
+      "epoch": 0.47898753743599654,
+      "grad_norm": 0.6559112668037415,
+      "learning_rate": 6.231556334763815e-06,
+      "loss": 0.6668,
+      "step": 2479
+    },
+    {
+      "epoch": 0.47918075548256206,
+      "grad_norm": 0.6012775897979736,
+      "learning_rate": 6.228287696573095e-06,
+      "loss": 0.5312,
+      "step": 2480
+    },
+    {
+      "epoch": 0.47937397352912764,
+      "grad_norm": 0.6608500480651855,
+      "learning_rate": 6.225018499653827e-06,
+      "loss": 0.6252,
+      "step": 2481
+    },
+    {
+      "epoch": 0.47956719157569316,
+      "grad_norm": 0.5881716012954712,
+      "learning_rate": 6.221748745493118e-06,
+      "loss": 0.5794,
+      "step": 2482
+    },
+    {
+      "epoch": 0.47976040962225874,
+      "grad_norm": 0.5924418568611145,
+      "learning_rate": 6.218478435578326e-06,
+      "loss": 0.5994,
+      "step": 2483
+    },
+    {
+      "epoch": 0.47995362766882427,
+      "grad_norm": 0.7298254370689392,
+      "learning_rate": 6.215207571397059e-06,
+      "loss": 0.6376,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4801468457153898,
+      "grad_norm": 0.7024441957473755,
+      "learning_rate": 6.211936154437187e-06,
+      "loss": 0.6192,
+      "step": 2485
+    },
+    {
+      "epoch": 0.48034006376195537,
+      "grad_norm": 0.7182833552360535,
+      "learning_rate": 6.208664186186823e-06,
+      "loss": 0.6501,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4805332818085209,
+      "grad_norm": 0.7058293223381042,
+      "learning_rate": 6.205391668134335e-06,
+      "loss": 0.636,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4807264998550865,
+      "grad_norm": 0.6544630527496338,
+      "learning_rate": 6.202118601768336e-06,
+      "loss": 0.6447,
+      "step": 2488
+    },
+    {
+      "epoch": 0.480919717901652,
+      "grad_norm": 0.6418452262878418,
+      "learning_rate": 6.198844988577696e-06,
+      "loss": 0.5934,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4811129359482176,
+      "grad_norm": 0.6905584335327148,
+      "learning_rate": 6.1955708300515274e-06,
+      "loss": 0.668,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4813061539947831,
+      "grad_norm": 0.6149024963378906,
+      "learning_rate": 6.1922961276791925e-06,
+      "loss": 0.5375,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4814993720413487,
+      "grad_norm": 0.6630429029464722,
+      "learning_rate": 6.189020882950302e-06,
+      "loss": 0.5852,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4816925900879142,
+      "grad_norm": 0.6637009978294373,
+      "learning_rate": 6.1857450973547115e-06,
+      "loss": 0.6092,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4818858081344798,
+      "grad_norm": 0.6476129293441772,
+      "learning_rate": 6.182468772382526e-06,
+      "loss": 0.6105,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4820790261810453,
+      "grad_norm": 0.661055862903595,
+      "learning_rate": 6.179191909524093e-06,
+      "loss": 0.5492,
+      "step": 2495
+    },
+    {
+      "epoch": 0.48227224422761084,
+      "grad_norm": 0.6576955914497375,
+      "learning_rate": 6.175914510270004e-06,
+      "loss": 0.5663,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4824654622741764,
+      "grad_norm": 0.7016202211380005,
+      "learning_rate": 6.172636576111098e-06,
+      "loss": 0.6669,
+      "step": 2497
+    },
+    {
+      "epoch": 0.48265868032074194,
+      "grad_norm": 0.6496583819389343,
+      "learning_rate": 6.169358108538454e-06,
+      "loss": 0.6022,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4828518983673075,
+      "grad_norm": 0.6798077821731567,
+      "learning_rate": 6.166079109043394e-06,
+      "loss": 0.6413,
+      "step": 2499
+    },
+    {
+      "epoch": 0.48304511641387304,
+      "grad_norm": 0.6625571250915527,
+      "learning_rate": 6.1627995791174826e-06,
+      "loss": 0.6311,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4832383344604386,
+      "grad_norm": 0.642199695110321,
+      "learning_rate": 6.159519520252526e-06,
+      "loss": 0.6675,
+      "step": 2501
+    },
+    {
+      "epoch": 0.48343155250700415,
+      "grad_norm": 0.6318427324295044,
+      "learning_rate": 6.156238933940573e-06,
+      "loss": 0.638,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4836247705535697,
+      "grad_norm": 0.6998022198677063,
+      "learning_rate": 6.152957821673908e-06,
+      "loss": 0.5813,
+      "step": 2503
+    },
+    {
+      "epoch": 0.48381798860013525,
+      "grad_norm": 0.658295750617981,
+      "learning_rate": 6.149676184945058e-06,
+      "loss": 0.6085,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4840112066467008,
+      "grad_norm": 0.5887330770492554,
+      "learning_rate": 6.1463940252467855e-06,
+      "loss": 0.5644,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48420442469326636,
+      "grad_norm": 0.6817413568496704,
+      "learning_rate": 6.143111344072095e-06,
+      "loss": 0.6543,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4843976427398319,
+      "grad_norm": 0.660243570804596,
+      "learning_rate": 6.139828142914226e-06,
+      "loss": 0.662,
+      "step": 2507
+    },
+    {
+      "epoch": 0.48459086078639746,
+      "grad_norm": 0.6131651997566223,
+      "learning_rate": 6.136544423266651e-06,
+      "loss": 0.6422,
+      "step": 2508
+    },
+    {
+      "epoch": 0.484784078832963,
+      "grad_norm": 0.5654097199440002,
+      "learning_rate": 6.133260186623087e-06,
+      "loss": 0.5773,
+      "step": 2509
+    },
+    {
+      "epoch": 0.48497729687952856,
+      "grad_norm": 0.6165230870246887,
+      "learning_rate": 6.129975434477478e-06,
+      "loss": 0.5698,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4851705149260941,
+      "grad_norm": 0.6954243779182434,
+      "learning_rate": 6.126690168324005e-06,
+      "loss": 0.6294,
+      "step": 2511
+    },
+    {
+      "epoch": 0.48536373297265967,
+      "grad_norm": 0.6481770873069763,
+      "learning_rate": 6.123404389657086e-06,
+      "loss": 0.6052,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4855569510192252,
+      "grad_norm": 0.6404730081558228,
+      "learning_rate": 6.120118099971366e-06,
+      "loss": 0.5798,
+      "step": 2513
+    },
+    {
+      "epoch": 0.48575016906579077,
+      "grad_norm": 0.7041105031967163,
+      "learning_rate": 6.116831300761728e-06,
+      "loss": 0.6257,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4859433871123563,
+      "grad_norm": 0.6712058782577515,
+      "learning_rate": 6.113543993523286e-06,
+      "loss": 0.6134,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4861366051589218,
+      "grad_norm": 0.6402435898780823,
+      "learning_rate": 6.110256179751382e-06,
+      "loss": 0.6159,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4863298232054874,
+      "grad_norm": 0.6374814510345459,
+      "learning_rate": 6.106967860941588e-06,
+      "loss": 0.5629,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4865230412520529,
+      "grad_norm": 0.6441965103149414,
+      "learning_rate": 6.103679038589714e-06,
+      "loss": 0.5877,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4867162592986185,
+      "grad_norm": 0.5908647775650024,
+      "learning_rate": 6.1003897141917876e-06,
+      "loss": 0.6366,
+      "step": 2519
+    },
+    {
+      "epoch": 0.486909477345184,
+      "grad_norm": 0.6850588321685791,
+      "learning_rate": 6.097099889244074e-06,
+      "loss": 0.6209,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4871026953917496,
+      "grad_norm": 0.6806885004043579,
+      "learning_rate": 6.0938095652430585e-06,
+      "loss": 0.6489,
+      "step": 2521
+    },
+    {
+      "epoch": 0.48729591343831513,
+      "grad_norm": 0.7047280073165894,
+      "learning_rate": 6.090518743685459e-06,
+      "loss": 0.6434,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4874891314848807,
+      "grad_norm": 0.5826668739318848,
+      "learning_rate": 6.0872274260682184e-06,
+      "loss": 0.5098,
+      "step": 2523
+    },
+    {
+      "epoch": 0.48768234953144624,
+      "grad_norm": 0.6922494769096375,
+      "learning_rate": 6.083935613888504e-06,
+      "loss": 0.6258,
+      "step": 2524
+    },
+    {
+      "epoch": 0.48787556757801176,
+      "grad_norm": 0.6513851284980774,
+      "learning_rate": 6.08064330864371e-06,
+      "loss": 0.6245,
+      "step": 2525
+    },
+    {
+      "epoch": 0.48806878562457734,
+      "grad_norm": 0.6038588285446167,
+      "learning_rate": 6.077350511831453e-06,
+      "loss": 0.6269,
+      "step": 2526
+    },
+    {
+      "epoch": 0.48826200367114286,
+      "grad_norm": 0.6524699926376343,
+      "learning_rate": 6.074057224949573e-06,
+      "loss": 0.621,
+      "step": 2527
+    },
+    {
+      "epoch": 0.48845522171770844,
+      "grad_norm": 0.7391641139984131,
+      "learning_rate": 6.070763449496135e-06,
+      "loss": 0.6553,
+      "step": 2528
+    },
+    {
+      "epoch": 0.48864843976427397,
+      "grad_norm": 0.6063517928123474,
+      "learning_rate": 6.0674691869694256e-06,
+      "loss": 0.6025,
+      "step": 2529
+    },
+    {
+      "epoch": 0.48884165781083955,
+      "grad_norm": 0.6999491453170776,
+      "learning_rate": 6.0641744388679506e-06,
+      "loss": 0.6354,
+      "step": 2530
+    },
+    {
+      "epoch": 0.48903487585740507,
+      "grad_norm": 0.6263997554779053,
+      "learning_rate": 6.060879206690438e-06,
+      "loss": 0.6073,
+      "step": 2531
+    },
+    {
+      "epoch": 0.48922809390397065,
+      "grad_norm": 0.6499011516571045,
+      "learning_rate": 6.05758349193584e-06,
+      "loss": 0.6163,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4894213119505362,
+      "grad_norm": 0.6485540270805359,
+      "learning_rate": 6.054287296103324e-06,
+      "loss": 0.6109,
+      "step": 2533
+    },
+    {
+      "epoch": 0.48961452999710176,
+      "grad_norm": 0.7666997313499451,
+      "learning_rate": 6.0509906206922745e-06,
+      "loss": 0.6272,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4898077480436673,
+      "grad_norm": 0.6203537583351135,
+      "learning_rate": 6.047693467202298e-06,
+      "loss": 0.5941,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4900009660902328,
+      "grad_norm": 0.6242226958274841,
+      "learning_rate": 6.044395837133219e-06,
+      "loss": 0.6334,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4901941841367984,
+      "grad_norm": 0.6836203336715698,
+      "learning_rate": 6.041097731985073e-06,
+      "loss": 0.586,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4903874021833639,
+      "grad_norm": 0.6860162019729614,
+      "learning_rate": 6.0377991532581196e-06,
+      "loss": 0.6334,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4905806202299295,
+      "grad_norm": 0.6926743388175964,
+      "learning_rate": 6.034500102452828e-06,
+      "loss": 0.6683,
+      "step": 2539
+    },
+    {
+      "epoch": 0.490773838276495,
+      "grad_norm": 0.6811209917068481,
+      "learning_rate": 6.031200581069885e-06,
+      "loss": 0.6111,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4909670563230606,
+      "grad_norm": 0.6656458973884583,
+      "learning_rate": 6.027900590610189e-06,
+      "loss": 0.7111,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4911602743696261,
+      "grad_norm": 0.632133960723877,
+      "learning_rate": 6.024600132574855e-06,
+      "loss": 0.6007,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4913534924161917,
+      "grad_norm": 0.6962891221046448,
+      "learning_rate": 6.0212992084652065e-06,
+      "loss": 0.6409,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4915467104627572,
+      "grad_norm": 0.6130524277687073,
+      "learning_rate": 6.0179978197827834e-06,
+      "loss": 0.5699,
+      "step": 2544
+    },
+    {
+      "epoch": 0.49173992850932274,
+      "grad_norm": 0.6345669031143188,
+      "learning_rate": 6.0146959680293336e-06,
+      "loss": 0.5825,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4919331465558883,
+      "grad_norm": 0.7199712991714478,
+      "learning_rate": 6.011393654706818e-06,
+      "loss": 0.6497,
+      "step": 2546
+    },
+    {
+      "epoch": 0.49212636460245385,
+      "grad_norm": 0.6979694366455078,
+      "learning_rate": 6.0080908813174065e-06,
+      "loss": 0.6785,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4923195826490194,
+      "grad_norm": 0.6557851433753967,
+      "learning_rate": 6.004787649363478e-06,
+      "loss": 0.6335,
+      "step": 2548
+    },
+    {
+      "epoch": 0.49251280069558495,
+      "grad_norm": 0.6547532677650452,
+      "learning_rate": 6.00148396034762e-06,
+      "loss": 0.5737,
+      "step": 2549
+    },
+    {
+      "epoch": 0.49270601874215053,
+      "grad_norm": 0.5670865774154663,
+      "learning_rate": 5.998179815772628e-06,
+      "loss": 0.5864,
+      "step": 2550
+    },
+    {
+      "epoch": 0.49289923678871606,
+      "grad_norm": 0.6839258670806885,
+      "learning_rate": 5.994875217141504e-06,
+      "loss": 0.6067,
+      "step": 2551
+    },
+    {
+      "epoch": 0.49309245483528164,
+      "grad_norm": 0.6927772760391235,
+      "learning_rate": 5.99157016595746e-06,
+      "loss": 0.6094,
+      "step": 2552
+    },
+    {
+      "epoch": 0.49328567288184716,
+      "grad_norm": 0.6989185214042664,
+      "learning_rate": 5.9882646637239075e-06,
+      "loss": 0.6468,
+      "step": 2553
+    },
+    {
+      "epoch": 0.49347889092841274,
+      "grad_norm": 0.7274642586708069,
+      "learning_rate": 5.98495871194447e-06,
+      "loss": 0.6729,
+      "step": 2554
+    },
+    {
+      "epoch": 0.49367210897497826,
+      "grad_norm": 0.6526702642440796,
+      "learning_rate": 5.981652312122971e-06,
+      "loss": 0.6363,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4938653270215438,
+      "grad_norm": 0.6925713419914246,
+      "learning_rate": 5.97834546576344e-06,
+      "loss": 0.6354,
+      "step": 2556
+    },
+    {
+      "epoch": 0.49405854506810937,
+      "grad_norm": 0.6756467819213867,
+      "learning_rate": 5.9750381743701095e-06,
+      "loss": 0.5946,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4942517631146749,
+      "grad_norm": 0.6518468260765076,
+      "learning_rate": 5.971730439447412e-06,
+      "loss": 0.5985,
+      "step": 2558
+    },
+    {
+      "epoch": 0.49444498116124047,
+      "grad_norm": 0.6253699660301208,
+      "learning_rate": 5.968422262499983e-06,
+      "loss": 0.5591,
+      "step": 2559
+    },
+    {
+      "epoch": 0.494638199207806,
+      "grad_norm": 0.6223752498626709,
+      "learning_rate": 5.965113645032662e-06,
+      "loss": 0.6273,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4948314172543716,
+      "grad_norm": 0.6594724655151367,
+      "learning_rate": 5.9618045885504845e-06,
+      "loss": 0.6521,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4950246353009371,
+      "grad_norm": 0.6084594130516052,
+      "learning_rate": 5.95849509455869e-06,
+      "loss": 0.6372,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4952178533475027,
+      "grad_norm": 0.6258931756019592,
+      "learning_rate": 5.955185164562712e-06,
+      "loss": 0.6113,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4954110713940682,
+      "grad_norm": 0.6560694575309753,
+      "learning_rate": 5.9518748000681885e-06,
+      "loss": 0.6183,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4956042894406338,
+      "grad_norm": 0.633563756942749,
+      "learning_rate": 5.948564002580949e-06,
+      "loss": 0.6245,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4957975074871993,
+      "grad_norm": 0.729521632194519,
+      "learning_rate": 5.945252773607022e-06,
+      "loss": 0.6871,
+      "step": 2566
+    },
+    {
+      "epoch": 0.49599072553376483,
+      "grad_norm": 0.6851238012313843,
+      "learning_rate": 5.941941114652634e-06,
+      "loss": 0.6736,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4961839435803304,
+      "grad_norm": 0.6673080325126648,
+      "learning_rate": 5.938629027224205e-06,
+      "loss": 0.5877,
+      "step": 2568
+    },
+    {
+      "epoch": 0.49637716162689594,
+      "grad_norm": 0.672149121761322,
+      "learning_rate": 5.935316512828354e-06,
+      "loss": 0.6575,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4965703796734615,
+      "grad_norm": 0.6841381788253784,
+      "learning_rate": 5.93200357297189e-06,
+      "loss": 0.6849,
+      "step": 2570
+    },
+    {
+      "epoch": 0.49676359772002704,
+      "grad_norm": 0.6196581125259399,
+      "learning_rate": 5.928690209161814e-06,
+      "loss": 0.657,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4969568157665926,
+      "grad_norm": 0.6392834782600403,
+      "learning_rate": 5.925376422905324e-06,
+      "loss": 0.6461,
+      "step": 2572
+    },
+    {
+      "epoch": 0.49715003381315814,
+      "grad_norm": 0.6133564710617065,
+      "learning_rate": 5.922062215709811e-06,
+      "loss": 0.5773,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4973432518597237,
+      "grad_norm": 0.6533375978469849,
+      "learning_rate": 5.918747589082853e-06,
+      "loss": 0.6327,
+      "step": 2574
+    },
+    {
+      "epoch": 0.49753646990628925,
+      "grad_norm": 0.678657054901123,
+      "learning_rate": 5.91543254453222e-06,
+      "loss": 0.7011,
+      "step": 2575
+    },
+    {
+      "epoch": 0.49772968795285477,
+      "grad_norm": 0.6999790668487549,
+      "learning_rate": 5.912117083565874e-06,
+      "loss": 0.5726,
+      "step": 2576
+    },
+    {
+      "epoch": 0.49792290599942035,
+      "grad_norm": 0.6411318778991699,
+      "learning_rate": 5.9088012076919665e-06,
+      "loss": 0.5863,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4981161240459859,
+      "grad_norm": 0.6458742022514343,
+      "learning_rate": 5.905484918418837e-06,
+      "loss": 0.6239,
+      "step": 2578
+    },
+    {
+      "epoch": 0.49830934209255145,
+      "grad_norm": 0.6810184717178345,
+      "learning_rate": 5.902168217255011e-06,
+      "loss": 0.675,
+      "step": 2579
+    },
+    {
+      "epoch": 0.498502560139117,
+      "grad_norm": 0.6272901892662048,
+      "learning_rate": 5.898851105709203e-06,
+      "loss": 0.5734,
+      "step": 2580
+    },
+    {
+      "epoch": 0.49869577818568256,
+      "grad_norm": 0.6340354084968567,
+      "learning_rate": 5.895533585290316e-06,
+      "loss": 0.5786,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4988889962322481,
+      "grad_norm": 0.7004643678665161,
+      "learning_rate": 5.892215657507436e-06,
+      "loss": 0.5704,
+      "step": 2582
+    },
+    {
+      "epoch": 0.49908221427881366,
+      "grad_norm": 0.6663402318954468,
+      "learning_rate": 5.888897323869838e-06,
+      "loss": 0.6389,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4992754323253792,
+      "grad_norm": 0.621563196182251,
+      "learning_rate": 5.885578585886975e-06,
+      "loss": 0.5607,
+      "step": 2584
+    },
+    {
+      "epoch": 0.49946865037194477,
+      "grad_norm": 0.6984502673149109,
+      "learning_rate": 5.882259445068491e-06,
+      "loss": 0.6496,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4996618684185103,
+      "grad_norm": 0.6837053298950195,
+      "learning_rate": 5.878939902924211e-06,
+      "loss": 0.6303,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4998550864650758,
+      "grad_norm": 0.6393314599990845,
+      "learning_rate": 5.875619960964141e-06,
+      "loss": 0.581,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5000483045116414,
+      "grad_norm": 0.6873040795326233,
+      "learning_rate": 5.872299620698468e-06,
+      "loss": 0.6405,
+      "step": 2588
+    },
+    {
+      "epoch": 0.500241522558207,
+      "grad_norm": 0.6256605982780457,
+      "learning_rate": 5.868978883637565e-06,
+      "loss": 0.5888,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5004347406047724,
+      "grad_norm": 0.6599069833755493,
+      "learning_rate": 5.86565775129198e-06,
+      "loss": 0.6079,
+      "step": 2590
+    },
+    {
+      "epoch": 0.500627958651338,
+      "grad_norm": 0.7463533282279968,
+      "learning_rate": 5.862336225172446e-06,
+      "loss": 0.6044,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5008211766979036,
+      "grad_norm": 0.6739301681518555,
+      "learning_rate": 5.859014306789869e-06,
+      "loss": 0.6035,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5010143947444692,
+      "grad_norm": 0.6515234708786011,
+      "learning_rate": 5.85569199765534e-06,
+      "loss": 0.6206,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5012076127910347,
+      "grad_norm": 0.6015059351921082,
+      "learning_rate": 5.8523692992801254e-06,
+      "loss": 0.6069,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5014008308376002,
+      "grad_norm": 0.739020824432373,
+      "learning_rate": 5.8490462131756645e-06,
+      "loss": 0.6839,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5015940488841658,
+      "grad_norm": 0.6538762450218201,
+      "learning_rate": 5.845722740853575e-06,
+      "loss": 0.668,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5017872669307313,
+      "grad_norm": 0.6097992062568665,
+      "learning_rate": 5.842398883825658e-06,
+      "loss": 0.5381,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5019804849772969,
+      "grad_norm": 0.7102888226509094,
+      "learning_rate": 5.839074643603877e-06,
+      "loss": 0.6197,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5021737030238624,
+      "grad_norm": 0.6422329545021057,
+      "learning_rate": 5.83575002170038e-06,
+      "loss": 0.6067,
+      "step": 2599
+    },
+    {
+      "epoch": 0.502366921070428,
+      "grad_norm": 0.6353461146354675,
+      "learning_rate": 5.832425019627483e-06,
+      "loss": 0.5966,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5025601391169935,
+      "grad_norm": 0.6211331486701965,
+      "learning_rate": 5.829099638897676e-06,
+      "loss": 0.5965,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5027533571635591,
+      "grad_norm": 0.7014438509941101,
+      "learning_rate": 5.825773881023624e-06,
+      "loss": 0.6112,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5029465752101246,
+      "grad_norm": 0.6234802007675171,
+      "learning_rate": 5.8224477475181606e-06,
+      "loss": 0.5607,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5031397932566902,
+      "grad_norm": 0.6403101682662964,
+      "learning_rate": 5.8191212398942885e-06,
+      "loss": 0.6134,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5033330113032557,
+      "grad_norm": 0.7103291749954224,
+      "learning_rate": 5.815794359665187e-06,
+      "loss": 0.6016,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5035262293498213,
+      "grad_norm": 0.754841148853302,
+      "learning_rate": 5.812467108344197e-06,
+      "loss": 0.6555,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5037194473963869,
+      "grad_norm": 0.6588846445083618,
+      "learning_rate": 5.809139487444837e-06,
+      "loss": 0.5941,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5039126654429523,
+      "grad_norm": 0.6401468515396118,
+      "learning_rate": 5.805811498480786e-06,
+      "loss": 0.6422,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5041058834895179,
+      "grad_norm": 0.6863776445388794,
+      "learning_rate": 5.802483142965892e-06,
+      "loss": 0.6234,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5042991015360835,
+      "grad_norm": 0.7024611234664917,
+      "learning_rate": 5.799154422414174e-06,
+      "loss": 0.6674,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5044923195826491,
+      "grad_norm": 0.6465156674385071,
+      "learning_rate": 5.7958253383398125e-06,
+      "loss": 0.5931,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5046855376292145,
+      "grad_norm": 0.5989453196525574,
+      "learning_rate": 5.792495892257154e-06,
+      "loss": 0.5861,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5048787556757801,
+      "grad_norm": 0.7412034869194031,
+      "learning_rate": 5.789166085680712e-06,
+      "loss": 0.6097,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5050719737223457,
+      "grad_norm": 0.7053552865982056,
+      "learning_rate": 5.785835920125162e-06,
+      "loss": 0.6359,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5052651917689113,
+      "grad_norm": 0.6897547841072083,
+      "learning_rate": 5.782505397105346e-06,
+      "loss": 0.623,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5054584098154767,
+      "grad_norm": 0.7389828562736511,
+      "learning_rate": 5.779174518136261e-06,
+      "loss": 0.5987,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5056516278620423,
+      "grad_norm": 0.6698645353317261,
+      "learning_rate": 5.7758432847330745e-06,
+      "loss": 0.625,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5058448459086079,
+      "grad_norm": 0.6617449522018433,
+      "learning_rate": 5.772511698411111e-06,
+      "loss": 0.6432,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5060380639551734,
+      "grad_norm": 0.663697361946106,
+      "learning_rate": 5.769179760685855e-06,
+      "loss": 0.6292,
+      "step": 2619
+    },
+    {
+      "epoch": 0.506231282001739,
+      "grad_norm": 0.649132251739502,
+      "learning_rate": 5.765847473072953e-06,
+      "loss": 0.6078,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5064245000483045,
+      "grad_norm": 0.7572246193885803,
+      "learning_rate": 5.762514837088209e-06,
+      "loss": 0.6675,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5066177180948701,
+      "grad_norm": 0.6414758563041687,
+      "learning_rate": 5.759181854247587e-06,
+      "loss": 0.6047,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5068109361414356,
+      "grad_norm": 0.6166477799415588,
+      "learning_rate": 5.75584852606721e-06,
+      "loss": 0.6066,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5070041541880012,
+      "grad_norm": 0.6314778327941895,
+      "learning_rate": 5.752514854063352e-06,
+      "loss": 0.5499,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5071973722345667,
+      "grad_norm": 0.6790992021560669,
+      "learning_rate": 5.749180839752448e-06,
+      "loss": 0.612,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5073905902811322,
+      "grad_norm": 0.7170957326889038,
+      "learning_rate": 5.745846484651088e-06,
+      "loss": 0.6107,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5075838083276978,
+      "grad_norm": 0.6408445835113525,
+      "learning_rate": 5.7425117902760195e-06,
+      "loss": 0.5826,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5077770263742634,
+      "grad_norm": 0.6264194846153259,
+      "learning_rate": 5.73917675814414e-06,
+      "loss": 0.6407,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5079702444208289,
+      "grad_norm": 0.7073822617530823,
+      "learning_rate": 5.735841389772504e-06,
+      "loss": 0.6776,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5081634624673944,
+      "grad_norm": 0.6584728956222534,
+      "learning_rate": 5.732505686678316e-06,
+      "loss": 0.575,
+      "step": 2630
+    },
+    {
+      "epoch": 0.50835668051396,
+      "grad_norm": 0.6637831926345825,
+      "learning_rate": 5.729169650378936e-06,
+      "loss": 0.6709,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5085498985605256,
+      "grad_norm": 0.6386693716049194,
+      "learning_rate": 5.7258332823918736e-06,
+      "loss": 0.5998,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5087431166070912,
+      "grad_norm": 0.6222007870674133,
+      "learning_rate": 5.7224965842347875e-06,
+      "loss": 0.5152,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5089363346536566,
+      "grad_norm": 0.658407986164093,
+      "learning_rate": 5.719159557425491e-06,
+      "loss": 0.578,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5091295527002222,
+      "grad_norm": 0.6780976057052612,
+      "learning_rate": 5.715822203481942e-06,
+      "loss": 0.6117,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5093227707467878,
+      "grad_norm": 0.6413382887840271,
+      "learning_rate": 5.712484523922253e-06,
+      "loss": 0.5245,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5095159887933532,
+      "grad_norm": 0.6029859185218811,
+      "learning_rate": 5.70914652026468e-06,
+      "loss": 0.6411,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5097092068399188,
+      "grad_norm": 0.6806251406669617,
+      "learning_rate": 5.705808194027626e-06,
+      "loss": 0.6677,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5099024248864844,
+      "grad_norm": 0.6593809723854065,
+      "learning_rate": 5.7024695467296455e-06,
+      "loss": 0.6188,
+      "step": 2639
+    },
+    {
+      "epoch": 0.51009564293305,
+      "grad_norm": 0.6814280152320862,
+      "learning_rate": 5.699130579889435e-06,
+      "loss": 0.5853,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5102888609796155,
+      "grad_norm": 0.6602916717529297,
+      "learning_rate": 5.695791295025836e-06,
+      "loss": 0.6366,
+      "step": 2641
+    },
+    {
+      "epoch": 0.510482079026181,
+      "grad_norm": 0.6823318600654602,
+      "learning_rate": 5.692451693657838e-06,
+      "loss": 0.5741,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5106752970727466,
+      "grad_norm": 0.6351958513259888,
+      "learning_rate": 5.68911177730457e-06,
+      "loss": 0.6037,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5108685151193122,
+      "grad_norm": 0.6258712410926819,
+      "learning_rate": 5.685771547485312e-06,
+      "loss": 0.5968,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5110617331658777,
+      "grad_norm": 0.6743547916412354,
+      "learning_rate": 5.682431005719476e-06,
+      "loss": 0.6168,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5112549512124432,
+      "grad_norm": 0.7343285083770752,
+      "learning_rate": 5.679090153526623e-06,
+      "loss": 0.6399,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5114481692590088,
+      "grad_norm": 0.6592523455619812,
+      "learning_rate": 5.675748992426455e-06,
+      "loss": 0.595,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5116413873055743,
+      "grad_norm": 0.686687707901001,
+      "learning_rate": 5.672407523938812e-06,
+      "loss": 0.6079,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5118346053521399,
+      "grad_norm": 0.6283683776855469,
+      "learning_rate": 5.669065749583677e-06,
+      "loss": 0.5847,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5120278233987055,
+      "grad_norm": 0.7083540558815002,
+      "learning_rate": 5.6657236708811645e-06,
+      "loss": 0.6633,
+      "step": 2650
+    },
+    {
+      "epoch": 0.512221041445271,
+      "grad_norm": 0.6702708005905151,
+      "learning_rate": 5.662381289351538e-06,
+      "loss": 0.5958,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5124142594918365,
+      "grad_norm": 0.6189550161361694,
+      "learning_rate": 5.659038606515192e-06,
+      "loss": 0.5709,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5126074775384021,
+      "grad_norm": 0.6320112943649292,
+      "learning_rate": 5.655695623892661e-06,
+      "loss": 0.5237,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5128006955849677,
+      "grad_norm": 0.7134533524513245,
+      "learning_rate": 5.652352343004612e-06,
+      "loss": 0.6416,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5129939136315332,
+      "grad_norm": 0.7378610372543335,
+      "learning_rate": 5.6490087653718525e-06,
+      "loss": 0.6233,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5131871316780987,
+      "grad_norm": 0.6779961585998535,
+      "learning_rate": 5.645664892515323e-06,
+      "loss": 0.6153,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5133803497246643,
+      "grad_norm": 0.6983788013458252,
+      "learning_rate": 5.642320725956098e-06,
+      "loss": 0.5593,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5135735677712299,
+      "grad_norm": 0.6985644102096558,
+      "learning_rate": 5.638976267215384e-06,
+      "loss": 0.7121,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5137667858177953,
+      "grad_norm": 0.6762123107910156,
+      "learning_rate": 5.635631517814523e-06,
+      "loss": 0.6634,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5139600038643609,
+      "grad_norm": 0.7402012944221497,
+      "learning_rate": 5.63228647927499e-06,
+      "loss": 0.6032,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5141532219109265,
+      "grad_norm": 0.707537055015564,
+      "learning_rate": 5.628941153118388e-06,
+      "loss": 0.6243,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5143464399574921,
+      "grad_norm": 0.6698459982872009,
+      "learning_rate": 5.625595540866453e-06,
+      "loss": 0.612,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5145396580040575,
+      "grad_norm": 0.7033206820487976,
+      "learning_rate": 5.62224964404105e-06,
+      "loss": 0.64,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5147328760506231,
+      "grad_norm": 0.688809335231781,
+      "learning_rate": 5.618903464164176e-06,
+      "loss": 0.6102,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5149260940971887,
+      "grad_norm": 0.6451379060745239,
+      "learning_rate": 5.6155570027579545e-06,
+      "loss": 0.5785,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5151193121437543,
+      "grad_norm": 0.6548169255256653,
+      "learning_rate": 5.6122102613446375e-06,
+      "loss": 0.5899,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5153125301903198,
+      "grad_norm": 0.6319230198860168,
+      "learning_rate": 5.608863241446603e-06,
+      "loss": 0.5992,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5155057482368853,
+      "grad_norm": 0.6549973487854004,
+      "learning_rate": 5.60551594458636e-06,
+      "loss": 0.6364,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5156989662834509,
+      "grad_norm": 0.7596017718315125,
+      "learning_rate": 5.602168372286538e-06,
+      "loss": 0.6941,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5158921843300164,
+      "grad_norm": 0.6469964385032654,
+      "learning_rate": 5.598820526069898e-06,
+      "loss": 0.6573,
+      "step": 2670
+    },
+    {
+      "epoch": 0.516085402376582,
+      "grad_norm": 0.6029120683670044,
+      "learning_rate": 5.595472407459317e-06,
+      "loss": 0.6557,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5162786204231475,
+      "grad_norm": 0.6519042253494263,
+      "learning_rate": 5.592124017977805e-06,
+      "loss": 0.5871,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5164718384697131,
+      "grad_norm": 0.6030751466751099,
+      "learning_rate": 5.588775359148488e-06,
+      "loss": 0.5646,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5166650565162786,
+      "grad_norm": 0.6155280470848083,
+      "learning_rate": 5.585426432494621e-06,
+      "loss": 0.6457,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5168582745628442,
+      "grad_norm": 0.6021338105201721,
+      "learning_rate": 5.582077239539575e-06,
+      "loss": 0.5621,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5170514926094097,
+      "grad_norm": 0.6500157713890076,
+      "learning_rate": 5.578727781806845e-06,
+      "loss": 0.6048,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5172447106559752,
+      "grad_norm": 0.7265481352806091,
+      "learning_rate": 5.575378060820048e-06,
+      "loss": 0.6404,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5174379287025408,
+      "grad_norm": 0.6832833290100098,
+      "learning_rate": 5.572028078102917e-06,
+      "loss": 0.6212,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5176311467491064,
+      "grad_norm": 0.6340216994285583,
+      "learning_rate": 5.568677835179306e-06,
+      "loss": 0.6356,
+      "step": 2679
+    },
+    {
+      "epoch": 0.517824364795672,
+      "grad_norm": 0.6746296882629395,
+      "learning_rate": 5.565327333573187e-06,
+      "loss": 0.6612,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5180175828422374,
+      "grad_norm": 0.6806156039237976,
+      "learning_rate": 5.561976574808651e-06,
+      "loss": 0.5984,
+      "step": 2681
+    },
+    {
+      "epoch": 0.518210800888803,
+      "grad_norm": 0.6181798577308655,
+      "learning_rate": 5.558625560409904e-06,
+      "loss": 0.5767,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5184040189353686,
+      "grad_norm": 0.6050678491592407,
+      "learning_rate": 5.55527429190127e-06,
+      "loss": 0.5688,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5185972369819342,
+      "grad_norm": 0.6985223293304443,
+      "learning_rate": 5.551922770807186e-06,
+      "loss": 0.5797,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5187904550284996,
+      "grad_norm": 0.6985082626342773,
+      "learning_rate": 5.548570998652208e-06,
+      "loss": 0.6489,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5189836730750652,
+      "grad_norm": 0.6887945532798767,
+      "learning_rate": 5.545218976961005e-06,
+      "loss": 0.6799,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5191768911216308,
+      "grad_norm": 0.7127766609191895,
+      "learning_rate": 5.541866707258353e-06,
+      "loss": 0.6813,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5193701091681963,
+      "grad_norm": 0.6471123695373535,
+      "learning_rate": 5.53851419106915e-06,
+      "loss": 0.672,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5195633272147618,
+      "grad_norm": 0.5983400344848633,
+      "learning_rate": 5.535161429918402e-06,
+      "loss": 0.6049,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5197565452613274,
+      "grad_norm": 0.7132647037506104,
+      "learning_rate": 5.531808425331228e-06,
+      "loss": 0.6718,
+      "step": 2690
+    },
+    {
+      "epoch": 0.519949763307893,
+      "grad_norm": 0.6270169019699097,
+      "learning_rate": 5.528455178832853e-06,
+      "loss": 0.652,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5201429813544585,
+      "grad_norm": 0.5793806910514832,
+      "learning_rate": 5.525101691948618e-06,
+      "loss": 0.6197,
+      "step": 2692
+    },
+    {
+      "epoch": 0.520336199401024,
+      "grad_norm": 0.6260225176811218,
+      "learning_rate": 5.521747966203971e-06,
+      "loss": 0.6184,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5205294174475896,
+      "grad_norm": 0.6648169755935669,
+      "learning_rate": 5.518394003124468e-06,
+      "loss": 0.6137,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5207226354941552,
+      "grad_norm": 0.6524384021759033,
+      "learning_rate": 5.515039804235772e-06,
+      "loss": 0.5749,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5209158535407207,
+      "grad_norm": 0.670122504234314,
+      "learning_rate": 5.511685371063657e-06,
+      "loss": 0.6678,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5211090715872863,
+      "grad_norm": 0.7135493159294128,
+      "learning_rate": 5.508330705134001e-06,
+      "loss": 0.634,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5213022896338518,
+      "grad_norm": 0.709801971912384,
+      "learning_rate": 5.5049758079727875e-06,
+      "loss": 0.588,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5214955076804173,
+      "grad_norm": 0.690405547618866,
+      "learning_rate": 5.501620681106107e-06,
+      "loss": 0.6265,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5216887257269829,
+      "grad_norm": 0.610143780708313,
+      "learning_rate": 5.498265326060151e-06,
+      "loss": 0.6653,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5218819437735485,
+      "grad_norm": 0.6290842890739441,
+      "learning_rate": 5.494909744361218e-06,
+      "loss": 0.6639,
+      "step": 2701
+    },
+    {
+      "epoch": 0.522075161820114,
+      "grad_norm": 0.6794373989105225,
+      "learning_rate": 5.491553937535712e-06,
+      "loss": 0.6077,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5222683798666795,
+      "grad_norm": 0.6602514386177063,
+      "learning_rate": 5.488197907110131e-06,
+      "loss": 0.6327,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5224615979132451,
+      "grad_norm": 0.6371023058891296,
+      "learning_rate": 5.4848416546110836e-06,
+      "loss": 0.598,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5226548159598107,
+      "grad_norm": 0.653971254825592,
+      "learning_rate": 5.481485181565274e-06,
+      "loss": 0.6022,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5228480340063762,
+      "grad_norm": 0.6721585392951965,
+      "learning_rate": 5.47812848949951e-06,
+      "loss": 0.5938,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5230412520529417,
+      "grad_norm": 0.6991129517555237,
+      "learning_rate": 5.474771579940696e-06,
+      "loss": 0.6096,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5232344700995073,
+      "grad_norm": 0.7635659575462341,
+      "learning_rate": 5.471414454415839e-06,
+      "loss": 0.6556,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5234276881460729,
+      "grad_norm": 0.6575093269348145,
+      "learning_rate": 5.468057114452038e-06,
+      "loss": 0.5906,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5236209061926383,
+      "grad_norm": 0.639979898929596,
+      "learning_rate": 5.464699561576496e-06,
+      "loss": 0.5931,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5238141242392039,
+      "grad_norm": 0.6271804571151733,
+      "learning_rate": 5.46134179731651e-06,
+      "loss": 0.6339,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5240073422857695,
+      "grad_norm": 0.6619129180908203,
+      "learning_rate": 5.457983823199475e-06,
+      "loss": 0.6348,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5242005603323351,
+      "grad_norm": 0.6763337254524231,
+      "learning_rate": 5.4546256407528786e-06,
+      "loss": 0.6064,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5243937783789006,
+      "grad_norm": 0.5628504753112793,
+      "learning_rate": 5.451267251504306e-06,
+      "loss": 0.5652,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5245869964254661,
+      "grad_norm": 0.6760022640228271,
+      "learning_rate": 5.447908656981433e-06,
+      "loss": 0.6291,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5247802144720317,
+      "grad_norm": 0.6151069402694702,
+      "learning_rate": 5.444549858712034e-06,
+      "loss": 0.6108,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5249734325185972,
+      "grad_norm": 0.6776319146156311,
+      "learning_rate": 5.441190858223968e-06,
+      "loss": 0.6263,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5251666505651628,
+      "grad_norm": 0.6101654767990112,
+      "learning_rate": 5.437831657045194e-06,
+      "loss": 0.5692,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5253598686117283,
+      "grad_norm": 0.7197254300117493,
+      "learning_rate": 5.43447225670376e-06,
+      "loss": 0.611,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5255530866582939,
+      "grad_norm": 0.7295145988464355,
+      "learning_rate": 5.431112658727803e-06,
+      "loss": 0.5985,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5257463047048594,
+      "grad_norm": 0.6472116708755493,
+      "learning_rate": 5.42775286464555e-06,
+      "loss": 0.6156,
+      "step": 2721
+    },
+    {
+      "epoch": 0.525939522751425,
+      "grad_norm": 0.6525043249130249,
+      "learning_rate": 5.424392875985319e-06,
+      "loss": 0.5849,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5261327407979906,
+      "grad_norm": 0.6719191670417786,
+      "learning_rate": 5.421032694275516e-06,
+      "loss": 0.6853,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5263259588445561,
+      "grad_norm": 0.6869013905525208,
+      "learning_rate": 5.4176723210446315e-06,
+      "loss": 0.6454,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5265191768911216,
+      "grad_norm": 0.7117831707000732,
+      "learning_rate": 5.414311757821247e-06,
+      "loss": 0.6415,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5267123949376872,
+      "grad_norm": 0.6393850445747375,
+      "learning_rate": 5.41095100613403e-06,
+      "loss": 0.6151,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5269056129842528,
+      "grad_norm": 0.6451537013053894,
+      "learning_rate": 5.4075900675117334e-06,
+      "loss": 0.6382,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5270988310308182,
+      "grad_norm": 0.6481597423553467,
+      "learning_rate": 5.404228943483195e-06,
+      "loss": 0.6224,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5272920490773838,
+      "grad_norm": 0.647049605846405,
+      "learning_rate": 5.400867635577335e-06,
+      "loss": 0.5789,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5274852671239494,
+      "grad_norm": 0.6580525040626526,
+      "learning_rate": 5.397506145323159e-06,
+      "loss": 0.6395,
+      "step": 2730
+    },
+    {
+      "epoch": 0.527678485170515,
+      "grad_norm": 0.6456242799758911,
+      "learning_rate": 5.394144474249759e-06,
+      "loss": 0.5668,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5278717032170804,
+      "grad_norm": 0.6864328384399414,
+      "learning_rate": 5.390782623886299e-06,
+      "loss": 0.6263,
+      "step": 2732
+    },
+    {
+      "epoch": 0.528064921263646,
+      "grad_norm": 0.7010552287101746,
+      "learning_rate": 5.387420595762039e-06,
+      "loss": 0.6147,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5282581393102116,
+      "grad_norm": 0.6834889650344849,
+      "learning_rate": 5.384058391406305e-06,
+      "loss": 0.6157,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5284513573567772,
+      "grad_norm": 0.6397814154624939,
+      "learning_rate": 5.380696012348516e-06,
+      "loss": 0.5933,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5286445754033426,
+      "grad_norm": 0.6197842955589294,
+      "learning_rate": 5.377333460118163e-06,
+      "loss": 0.6352,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5288377934499082,
+      "grad_norm": 0.6409375667572021,
+      "learning_rate": 5.3739707362448165e-06,
+      "loss": 0.6173,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5290310114964738,
+      "grad_norm": 0.7432710528373718,
+      "learning_rate": 5.370607842258127e-06,
+      "loss": 0.6379,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5292242295430393,
+      "grad_norm": 0.6212089657783508,
+      "learning_rate": 5.36724477968782e-06,
+      "loss": 0.5886,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5294174475896049,
+      "grad_norm": 0.7244754433631897,
+      "learning_rate": 5.363881550063701e-06,
+      "loss": 0.6845,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5296106656361704,
+      "grad_norm": 0.6106608510017395,
+      "learning_rate": 5.36051815491565e-06,
+      "loss": 0.6456,
+      "step": 2741
+    },
+    {
+      "epoch": 0.529803883682736,
+      "grad_norm": 0.725473165512085,
+      "learning_rate": 5.3571545957736205e-06,
+      "loss": 0.6983,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5299971017293015,
+      "grad_norm": 0.6816869974136353,
+      "learning_rate": 5.353790874167642e-06,
+      "loss": 0.6069,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5301903197758671,
+      "grad_norm": 0.7198229432106018,
+      "learning_rate": 5.350426991627821e-06,
+      "loss": 0.6687,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5303835378224326,
+      "grad_norm": 0.7111002206802368,
+      "learning_rate": 5.347062949684331e-06,
+      "loss": 0.6791,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5305767558689982,
+      "grad_norm": 0.6627780795097351,
+      "learning_rate": 5.343698749867421e-06,
+      "loss": 0.6381,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5307699739155637,
+      "grad_norm": 0.6470596194267273,
+      "learning_rate": 5.340334393707414e-06,
+      "loss": 0.6352,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5309631919621293,
+      "grad_norm": 0.5991728901863098,
+      "learning_rate": 5.336969882734701e-06,
+      "loss": 0.6106,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5311564100086948,
+      "grad_norm": 0.621097207069397,
+      "learning_rate": 5.333605218479745e-06,
+      "loss": 0.5793,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5313496280552603,
+      "grad_norm": 0.6444972157478333,
+      "learning_rate": 5.330240402473078e-06,
+      "loss": 0.5619,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5315428461018259,
+      "grad_norm": 0.6247841715812683,
+      "learning_rate": 5.326875436245302e-06,
+      "loss": 0.5874,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5317360641483915,
+      "grad_norm": 0.6569274067878723,
+      "learning_rate": 5.323510321327086e-06,
+      "loss": 0.6102,
+      "step": 2752
+    },
+    {
+      "epoch": 0.531929282194957,
+      "grad_norm": 0.6525456309318542,
+      "learning_rate": 5.320145059249167e-06,
+      "loss": 0.5819,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5321225002415225,
+      "grad_norm": 0.5973703265190125,
+      "learning_rate": 5.316779651542349e-06,
+      "loss": 0.6013,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5323157182880881,
+      "grad_norm": 0.6485013365745544,
+      "learning_rate": 5.3134140997375025e-06,
+      "loss": 0.6058,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5325089363346537,
+      "grad_norm": 0.5931412577629089,
+      "learning_rate": 5.310048405365565e-06,
+      "loss": 0.605,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5327021543812192,
+      "grad_norm": 0.6626282334327698,
+      "learning_rate": 5.306682569957538e-06,
+      "loss": 0.6172,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5328953724277847,
+      "grad_norm": 0.6638219952583313,
+      "learning_rate": 5.3033165950444835e-06,
+      "loss": 0.6265,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5330885904743503,
+      "grad_norm": 0.6781473755836487,
+      "learning_rate": 5.299950482157531e-06,
+      "loss": 0.6288,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5332818085209159,
+      "grad_norm": 0.6562187075614929,
+      "learning_rate": 5.296584232827874e-06,
+      "loss": 0.6454,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5334750265674814,
+      "grad_norm": 0.6155015230178833,
+      "learning_rate": 5.2932178485867625e-06,
+      "loss": 0.5962,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5336682446140469,
+      "grad_norm": 0.6567010283470154,
+      "learning_rate": 5.289851330965512e-06,
+      "loss": 0.6302,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5338614626606125,
+      "grad_norm": 0.6853295564651489,
+      "learning_rate": 5.2864846814955e-06,
+      "loss": 0.6366,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5340546807071781,
+      "grad_norm": 0.6230853796005249,
+      "learning_rate": 5.283117901708159e-06,
+      "loss": 0.6156,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5342478987537436,
+      "grad_norm": 0.6507954001426697,
+      "learning_rate": 5.279750993134987e-06,
+      "loss": 0.6425,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5344411168003091,
+      "grad_norm": 0.6383042931556702,
+      "learning_rate": 5.276383957307535e-06,
+      "loss": 0.5884,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5346343348468747,
+      "grad_norm": 0.6133530735969543,
+      "learning_rate": 5.273016795757412e-06,
+      "loss": 0.5704,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5348275528934402,
+      "grad_norm": 0.6666024923324585,
+      "learning_rate": 5.269649510016291e-06,
+      "loss": 0.5935,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5350207709400058,
+      "grad_norm": 0.6686692237854004,
+      "learning_rate": 5.266282101615893e-06,
+      "loss": 0.6523,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5352139889865714,
+      "grad_norm": 0.6970983147621155,
+      "learning_rate": 5.262914572087999e-06,
+      "loss": 0.6232,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5354072070331369,
+      "grad_norm": 0.7154514193534851,
+      "learning_rate": 5.259546922964446e-06,
+      "loss": 0.5751,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5356004250797024,
+      "grad_norm": 0.6319283246994019,
+      "learning_rate": 5.2561791557771236e-06,
+      "loss": 0.592,
+      "step": 2772
+    },
+    {
+      "epoch": 0.535793643126268,
+      "grad_norm": 0.6348755955696106,
+      "learning_rate": 5.252811272057977e-06,
+      "loss": 0.5831,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5359868611728336,
+      "grad_norm": 0.610826849937439,
+      "learning_rate": 5.249443273339e-06,
+      "loss": 0.6003,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5361800792193991,
+      "grad_norm": 0.7023615837097168,
+      "learning_rate": 5.2460751611522445e-06,
+      "loss": 0.6054,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5363732972659646,
+      "grad_norm": 0.6908557415008545,
+      "learning_rate": 5.242706937029808e-06,
+      "loss": 0.6006,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5365665153125302,
+      "grad_norm": 0.6553770899772644,
+      "learning_rate": 5.239338602503844e-06,
+      "loss": 0.6673,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5367597333590958,
+      "grad_norm": 0.6191368699073792,
+      "learning_rate": 5.235970159106554e-06,
+      "loss": 0.5831,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5369529514056612,
+      "grad_norm": 0.6711176633834839,
+      "learning_rate": 5.232601608370191e-06,
+      "loss": 0.6292,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5371461694522268,
+      "grad_norm": 0.7053834199905396,
+      "learning_rate": 5.229232951827054e-06,
+      "loss": 0.6428,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5373393874987924,
+      "grad_norm": 0.6941760778427124,
+      "learning_rate": 5.22586419100949e-06,
+      "loss": 0.6578,
+      "step": 2781
+    },
+    {
+      "epoch": 0.537532605545358,
+      "grad_norm": 0.6883274912834167,
+      "learning_rate": 5.2224953274498965e-06,
+      "loss": 0.6503,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5377258235919234,
+      "grad_norm": 0.6377370953559875,
+      "learning_rate": 5.219126362680716e-06,
+      "loss": 0.6006,
+      "step": 2783
+    },
+    {
+      "epoch": 0.537919041638489,
+      "grad_norm": 0.67820143699646,
+      "learning_rate": 5.215757298234435e-06,
+      "loss": 0.5631,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5381122596850546,
+      "grad_norm": 0.6192267537117004,
+      "learning_rate": 5.212388135643588e-06,
+      "loss": 0.6487,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5383054777316202,
+      "grad_norm": 0.7302708029747009,
+      "learning_rate": 5.209018876440756e-06,
+      "loss": 0.5561,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5384986957781857,
+      "grad_norm": 0.7372840642929077,
+      "learning_rate": 5.20564952215856e-06,
+      "loss": 0.5972,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5386919138247512,
+      "grad_norm": 0.6722862124443054,
+      "learning_rate": 5.202280074329663e-06,
+      "loss": 0.6218,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5388851318713168,
+      "grad_norm": 0.6592112183570862,
+      "learning_rate": 5.198910534486779e-06,
+      "loss": 0.6406,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5390783499178823,
+      "grad_norm": 0.6313249468803406,
+      "learning_rate": 5.195540904162652e-06,
+      "loss": 0.6554,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5392715679644479,
+      "grad_norm": 0.6923301815986633,
+      "learning_rate": 5.192171184890078e-06,
+      "loss": 0.6236,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5394647860110134,
+      "grad_norm": 0.6240240335464478,
+      "learning_rate": 5.188801378201885e-06,
+      "loss": 0.6208,
+      "step": 2792
+    },
+    {
+      "epoch": 0.539658004057579,
+      "grad_norm": 0.5939702987670898,
+      "learning_rate": 5.1854314856309465e-06,
+      "loss": 0.5371,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5398512221041445,
+      "grad_norm": 0.6817887425422668,
+      "learning_rate": 5.182061508710173e-06,
+      "loss": 0.6337,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5400444401507101,
+      "grad_norm": 0.6512333154678345,
+      "learning_rate": 5.1786914489725145e-06,
+      "loss": 0.6065,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5402376581972756,
+      "grad_norm": 0.5865704417228699,
+      "learning_rate": 5.175321307950954e-06,
+      "loss": 0.5441,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5404308762438412,
+      "grad_norm": 0.7187792062759399,
+      "learning_rate": 5.17195108717852e-06,
+      "loss": 0.6072,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5406240942904067,
+      "grad_norm": 0.6709620952606201,
+      "learning_rate": 5.1685807881882685e-06,
+      "loss": 0.6712,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5408173123369723,
+      "grad_norm": 0.6507397890090942,
+      "learning_rate": 5.165210412513296e-06,
+      "loss": 0.5752,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5410105303835379,
+      "grad_norm": 0.6615001559257507,
+      "learning_rate": 5.161839961686734e-06,
+      "loss": 0.6274,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5412037484301033,
+      "grad_norm": 0.5951465964317322,
+      "learning_rate": 5.158469437241745e-06,
+      "loss": 0.5689,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5413969664766689,
+      "grad_norm": 0.6031231880187988,
+      "learning_rate": 5.155098840711529e-06,
+      "loss": 0.5474,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5415901845232345,
+      "grad_norm": 0.609968900680542,
+      "learning_rate": 5.151728173629316e-06,
+      "loss": 0.5362,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5417834025698001,
+      "grad_norm": 0.6773534417152405,
+      "learning_rate": 5.148357437528368e-06,
+      "loss": 0.6639,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5419766206163655,
+      "grad_norm": 0.5893373489379883,
+      "learning_rate": 5.144986633941979e-06,
+      "loss": 0.5127,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5421698386629311,
+      "grad_norm": 0.6556066274642944,
+      "learning_rate": 5.1416157644034766e-06,
+      "loss": 0.6174,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5423630567094967,
+      "grad_norm": 0.5999477505683899,
+      "learning_rate": 5.138244830446213e-06,
+      "loss": 0.5696,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5425562747560622,
+      "grad_norm": 0.9360123872756958,
+      "learning_rate": 5.1348738336035745e-06,
+      "loss": 0.6142,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5427494928026277,
+      "grad_norm": 0.7187207341194153,
+      "learning_rate": 5.1315027754089716e-06,
+      "loss": 0.5668,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5429427108491933,
+      "grad_norm": 0.6945245265960693,
+      "learning_rate": 5.128131657395848e-06,
+      "loss": 0.6547,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5431359288957589,
+      "grad_norm": 0.6497440934181213,
+      "learning_rate": 5.12476048109767e-06,
+      "loss": 0.6316,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5433291469423244,
+      "grad_norm": 0.7627323269844055,
+      "learning_rate": 5.121389248047933e-06,
+      "loss": 0.6357,
+      "step": 2812
+    },
+    {
+      "epoch": 0.54352236498889,
+      "grad_norm": 0.6285127401351929,
+      "learning_rate": 5.118017959780155e-06,
+      "loss": 0.5586,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5437155830354555,
+      "grad_norm": 0.6257187724113464,
+      "learning_rate": 5.114646617827884e-06,
+      "loss": 0.5946,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5439088010820211,
+      "grad_norm": 0.6658818125724792,
+      "learning_rate": 5.1112752237246885e-06,
+      "loss": 0.5868,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5441020191285866,
+      "grad_norm": 0.6484278440475464,
+      "learning_rate": 5.107903779004163e-06,
+      "loss": 0.6553,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5442952371751522,
+      "grad_norm": 0.6568118333816528,
+      "learning_rate": 5.104532285199924e-06,
+      "loss": 0.5869,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5444884552217177,
+      "grad_norm": 0.6952958703041077,
+      "learning_rate": 5.101160743845609e-06,
+      "loss": 0.668,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5446816732682832,
+      "grad_norm": 0.594314455986023,
+      "learning_rate": 5.097789156474882e-06,
+      "loss": 0.5134,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5448748913148488,
+      "grad_norm": 0.6452954411506653,
+      "learning_rate": 5.09441752462142e-06,
+      "loss": 0.6728,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5450681093614144,
+      "grad_norm": 0.6499065160751343,
+      "learning_rate": 5.091045849818927e-06,
+      "loss": 0.7006,
+      "step": 2821
+    },
+    {
+      "epoch": 0.54526132740798,
+      "grad_norm": 0.6829864382743835,
+      "learning_rate": 5.0876741336011246e-06,
+      "loss": 0.6209,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 0.6096253991127014,
+      "learning_rate": 5.084302377501753e-06,
+      "loss": 0.5886,
+      "step": 2823
+    },
+    {
+      "epoch": 0.545647763501111,
+      "grad_norm": 0.6730426549911499,
+      "learning_rate": 5.08093058305457e-06,
+      "loss": 0.6194,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5458409815476766,
+      "grad_norm": 0.6383257508277893,
+      "learning_rate": 5.077558751793351e-06,
+      "loss": 0.619,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5460341995942422,
+      "grad_norm": 0.6522809863090515,
+      "learning_rate": 5.074186885251888e-06,
+      "loss": 0.5653,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5462274176408076,
+      "grad_norm": 0.688464879989624,
+      "learning_rate": 5.070814984963991e-06,
+      "loss": 0.5841,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5464206356873732,
+      "grad_norm": 0.6673641800880432,
+      "learning_rate": 5.067443052463484e-06,
+      "loss": 0.6454,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5466138537339388,
+      "grad_norm": 0.6391381025314331,
+      "learning_rate": 5.0640710892842016e-06,
+      "loss": 0.6208,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5468070717805043,
+      "grad_norm": 0.6299818158149719,
+      "learning_rate": 5.06069909696e-06,
+      "loss": 0.557,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5470002898270698,
+      "grad_norm": 0.6129377484321594,
+      "learning_rate": 5.057327077024745e-06,
+      "loss": 0.5709,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5471935078736354,
+      "grad_norm": 0.706803560256958,
+      "learning_rate": 5.0539550310123144e-06,
+      "loss": 0.6093,
+      "step": 2832
+    },
+    {
+      "epoch": 0.547386725920201,
+      "grad_norm": 0.6250451803207397,
+      "learning_rate": 5.050582960456598e-06,
+      "loss": 0.571,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5475799439667665,
+      "grad_norm": 0.6042814254760742,
+      "learning_rate": 5.047210866891495e-06,
+      "loss": 0.612,
+      "step": 2834
+    },
+    {
+      "epoch": 0.547773162013332,
+      "grad_norm": 0.6108737587928772,
+      "learning_rate": 5.043838751850921e-06,
+      "loss": 0.6241,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5479663800598976,
+      "grad_norm": 0.6449500322341919,
+      "learning_rate": 5.040466616868795e-06,
+      "loss": 0.6088,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5481595981064632,
+      "grad_norm": 0.7373918294906616,
+      "learning_rate": 5.037094463479048e-06,
+      "loss": 0.665,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5483528161530287,
+      "grad_norm": 0.6938897371292114,
+      "learning_rate": 5.033722293215619e-06,
+      "loss": 0.6613,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5485460341995942,
+      "grad_norm": 0.6753829717636108,
+      "learning_rate": 5.030350107612456e-06,
+      "loss": 0.6688,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5487392522461598,
+      "grad_norm": 0.6768985390663147,
+      "learning_rate": 5.026977908203511e-06,
+      "loss": 0.5925,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5489324702927253,
+      "grad_norm": 0.6655199527740479,
+      "learning_rate": 5.023605696522744e-06,
+      "loss": 0.6436,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5491256883392909,
+      "grad_norm": 0.6161617636680603,
+      "learning_rate": 5.020233474104121e-06,
+      "loss": 0.5788,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5493189063858565,
+      "grad_norm": 0.6460961103439331,
+      "learning_rate": 5.016861242481614e-06,
+      "loss": 0.6116,
+      "step": 2843
+    },
+    {
+      "epoch": 0.549512124432422,
+      "grad_norm": 0.6198633313179016,
+      "learning_rate": 5.013489003189194e-06,
+      "loss": 0.5847,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5497053424789875,
+      "grad_norm": 0.6289231181144714,
+      "learning_rate": 5.01011675776084e-06,
+      "loss": 0.594,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5498985605255531,
+      "grad_norm": 0.7044505476951599,
+      "learning_rate": 5.006744507730534e-06,
+      "loss": 0.5382,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5500917785721187,
+      "grad_norm": 0.6441923975944519,
+      "learning_rate": 5.003372254632259e-06,
+      "loss": 0.6212,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5502849966186841,
+      "grad_norm": 0.6189219355583191,
+      "learning_rate": 5e-06,
+      "loss": 0.562,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5504782146652497,
+      "grad_norm": 0.6506729125976562,
+      "learning_rate": 4.99662774536774e-06,
+      "loss": 0.607,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5506714327118153,
+      "grad_norm": 0.6583701968193054,
+      "learning_rate": 4.9932554922694675e-06,
+      "loss": 0.6057,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5508646507583809,
+      "grad_norm": 0.6298441290855408,
+      "learning_rate": 4.989883242239162e-06,
+      "loss": 0.6183,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5510578688049463,
+      "grad_norm": 0.6135046482086182,
+      "learning_rate": 4.986510996810809e-06,
+      "loss": 0.5728,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5512510868515119,
+      "grad_norm": 0.6704480051994324,
+      "learning_rate": 4.983138757518389e-06,
+      "loss": 0.6431,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5514443048980775,
+      "grad_norm": 0.658631443977356,
+      "learning_rate": 4.979766525895879e-06,
+      "loss": 0.6001,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5516375229446431,
+      "grad_norm": 0.6455361843109131,
+      "learning_rate": 4.976394303477258e-06,
+      "loss": 0.5525,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5518307409912085,
+      "grad_norm": 0.6811034679412842,
+      "learning_rate": 4.973022091796491e-06,
+      "loss": 0.6297,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5520239590377741,
+      "grad_norm": 0.737478494644165,
+      "learning_rate": 4.969649892387545e-06,
+      "loss": 0.6757,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5522171770843397,
+      "grad_norm": 0.608394205570221,
+      "learning_rate": 4.966277706784383e-06,
+      "loss": 0.5519,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5524103951309052,
+      "grad_norm": 0.6599392294883728,
+      "learning_rate": 4.962905536520953e-06,
+      "loss": 0.6354,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5526036131774708,
+      "grad_norm": 0.7015693187713623,
+      "learning_rate": 4.959533383131207e-06,
+      "loss": 0.623,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5527968312240363,
+      "grad_norm": 0.6973515152931213,
+      "learning_rate": 4.9561612481490805e-06,
+      "loss": 0.6701,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5529900492706019,
+      "grad_norm": 0.6182733178138733,
+      "learning_rate": 4.952789133108505e-06,
+      "loss": 0.5824,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5531832673171674,
+      "grad_norm": 0.63578861951828,
+      "learning_rate": 4.949417039543405e-06,
+      "loss": 0.6016,
+      "step": 2863
+    },
+    {
+      "epoch": 0.553376485363733,
+      "grad_norm": 0.6409711837768555,
+      "learning_rate": 4.946044968987686e-06,
+      "loss": 0.5837,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5535697034102985,
+      "grad_norm": 0.6623215079307556,
+      "learning_rate": 4.942672922975255e-06,
+      "loss": 0.5036,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5537629214568641,
+      "grad_norm": 0.6077432036399841,
+      "learning_rate": 4.939300903040001e-06,
+      "loss": 0.6101,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5539561395034296,
+      "grad_norm": 0.6640979051589966,
+      "learning_rate": 4.9359289107158e-06,
+      "loss": 0.6067,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5541493575499952,
+      "grad_norm": 0.6606337428092957,
+      "learning_rate": 4.932556947536519e-06,
+      "loss": 0.6357,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5543425755965607,
+      "grad_norm": 0.703130304813385,
+      "learning_rate": 4.9291850150360105e-06,
+      "loss": 0.6063,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5545357936431262,
+      "grad_norm": 0.7002109289169312,
+      "learning_rate": 4.9258131147481125e-06,
+      "loss": 0.7258,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5547290116896918,
+      "grad_norm": 0.6449876427650452,
+      "learning_rate": 4.922441248206652e-06,
+      "loss": 0.6267,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5549222297362574,
+      "grad_norm": 0.7682011723518372,
+      "learning_rate": 4.919069416945432e-06,
+      "loss": 0.7303,
+      "step": 2872
+    },
+    {
+      "epoch": 0.555115447782823,
+      "grad_norm": 0.6637479662895203,
+      "learning_rate": 4.915697622498248e-06,
+      "loss": 0.5755,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5553086658293884,
+      "grad_norm": 0.634918212890625,
+      "learning_rate": 4.912325866398876e-06,
+      "loss": 0.6472,
+      "step": 2874
+    },
+    {
+      "epoch": 0.555501883875954,
+      "grad_norm": 0.6748210191726685,
+      "learning_rate": 4.908954150181074e-06,
+      "loss": 0.6403,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5556951019225196,
+      "grad_norm": 0.6177434325218201,
+      "learning_rate": 4.905582475378582e-06,
+      "loss": 0.6022,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5558883199690852,
+      "grad_norm": 0.7100134491920471,
+      "learning_rate": 4.9022108435251205e-06,
+      "loss": 0.6253,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5560815380156506,
+      "grad_norm": 0.5896530151367188,
+      "learning_rate": 4.898839256154391e-06,
+      "loss": 0.5456,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5562747560622162,
+      "grad_norm": 0.6749439239501953,
+      "learning_rate": 4.895467714800078e-06,
+      "loss": 0.6381,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5564679741087818,
+      "grad_norm": 0.6600642204284668,
+      "learning_rate": 4.892096220995838e-06,
+      "loss": 0.5839,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5566611921553473,
+      "grad_norm": 0.6440023183822632,
+      "learning_rate": 4.888724776275313e-06,
+      "loss": 0.5712,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5568544102019128,
+      "grad_norm": 0.6551118493080139,
+      "learning_rate": 4.8853533821721175e-06,
+      "loss": 0.6255,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5570476282484784,
+      "grad_norm": 0.6725713610649109,
+      "learning_rate": 4.881982040219845e-06,
+      "loss": 0.6212,
+      "step": 2883
+    },
+    {
+      "epoch": 0.557240846295044,
+      "grad_norm": 0.6112415194511414,
+      "learning_rate": 4.87861075195207e-06,
+      "loss": 0.5782,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5574340643416095,
+      "grad_norm": 0.6592001914978027,
+      "learning_rate": 4.875239518902331e-06,
+      "loss": 0.6531,
+      "step": 2885
+    },
+    {
+      "epoch": 0.557627282388175,
+      "grad_norm": 0.6550741195678711,
+      "learning_rate": 4.871868342604153e-06,
+      "loss": 0.6066,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5578205004347406,
+      "grad_norm": 0.6677191257476807,
+      "learning_rate": 4.86849722459103e-06,
+      "loss": 0.5607,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5580137184813062,
+      "grad_norm": 0.6839702725410461,
+      "learning_rate": 4.865126166396427e-06,
+      "loss": 0.5985,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5582069365278717,
+      "grad_norm": 0.617600679397583,
+      "learning_rate": 4.8617551695537885e-06,
+      "loss": 0.5393,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5584001545744373,
+      "grad_norm": 0.6372525095939636,
+      "learning_rate": 4.858384235596525e-06,
+      "loss": 0.5661,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5585933726210028,
+      "grad_norm": 0.6697811484336853,
+      "learning_rate": 4.8550133660580214e-06,
+      "loss": 0.6158,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5587865906675683,
+      "grad_norm": 0.679943323135376,
+      "learning_rate": 4.851642562471634e-06,
+      "loss": 0.6614,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5589798087141339,
+      "grad_norm": 0.5921047925949097,
+      "learning_rate": 4.848271826370686e-06,
+      "loss": 0.6384,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5591730267606995,
+      "grad_norm": 0.6646453142166138,
+      "learning_rate": 4.844901159288471e-06,
+      "loss": 0.5948,
+      "step": 2894
+    },
+    {
+      "epoch": 0.559366244807265,
+      "grad_norm": 0.7251352071762085,
+      "learning_rate": 4.841530562758256e-06,
+      "loss": 0.6084,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5595594628538305,
+      "grad_norm": 0.6890468001365662,
+      "learning_rate": 4.838160038313267e-06,
+      "loss": 0.6217,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5597526809003961,
+      "grad_norm": 0.6829890608787537,
+      "learning_rate": 4.8347895874867054e-06,
+      "loss": 0.5892,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5599458989469617,
+      "grad_norm": 0.6603589653968811,
+      "learning_rate": 4.831419211811732e-06,
+      "loss": 0.5909,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5601391169935271,
+      "grad_norm": 0.6246123313903809,
+      "learning_rate": 4.82804891282148e-06,
+      "loss": 0.5657,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5603323350400927,
+      "grad_norm": 0.6900237798690796,
+      "learning_rate": 4.8246786920490465e-06,
+      "loss": 0.7376,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5605255530866583,
+      "grad_norm": 0.6919370889663696,
+      "learning_rate": 4.821308551027487e-06,
+      "loss": 0.6089,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5607187711332239,
+      "grad_norm": 0.6730056405067444,
+      "learning_rate": 4.817938491289829e-06,
+      "loss": 0.7023,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5609119891797893,
+      "grad_norm": 0.6424972414970398,
+      "learning_rate": 4.814568514369055e-06,
+      "loss": 0.6134,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5611052072263549,
+      "grad_norm": 0.6296237111091614,
+      "learning_rate": 4.8111986217981165e-06,
+      "loss": 0.5504,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5612984252729205,
+      "grad_norm": 0.6533901691436768,
+      "learning_rate": 4.8078288151099245e-06,
+      "loss": 0.5818,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5614916433194861,
+      "grad_norm": 0.6038861274719238,
+      "learning_rate": 4.804459095837349e-06,
+      "loss": 0.5745,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5616848613660516,
+      "grad_norm": 0.6763200759887695,
+      "learning_rate": 4.801089465513222e-06,
+      "loss": 0.5984,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5618780794126171,
+      "grad_norm": 0.604251503944397,
+      "learning_rate": 4.797719925670338e-06,
+      "loss": 0.6329,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5620712974591827,
+      "grad_norm": 0.6831694841384888,
+      "learning_rate": 4.794350477841442e-06,
+      "loss": 0.6756,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5622645155057482,
+      "grad_norm": 0.6490294933319092,
+      "learning_rate": 4.7909811235592465e-06,
+      "loss": 0.5833,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5624577335523138,
+      "grad_norm": 0.6486014127731323,
+      "learning_rate": 4.787611864356413e-06,
+      "loss": 0.6104,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5626509515988793,
+      "grad_norm": 0.6566706299781799,
+      "learning_rate": 4.784242701765567e-06,
+      "loss": 0.5823,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5628441696454449,
+      "grad_norm": 0.5997588634490967,
+      "learning_rate": 4.780873637319287e-06,
+      "loss": 0.5958,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5630373876920104,
+      "grad_norm": 0.6183865070343018,
+      "learning_rate": 4.777504672550104e-06,
+      "loss": 0.5477,
+      "step": 2914
+    },
+    {
+      "epoch": 0.563230605738576,
+      "grad_norm": 0.6025052070617676,
+      "learning_rate": 4.77413580899051e-06,
+      "loss": 0.582,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5634238237851416,
+      "grad_norm": 0.6585809588432312,
+      "learning_rate": 4.770767048172948e-06,
+      "loss": 0.5544,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5636170418317071,
+      "grad_norm": 0.6207132339477539,
+      "learning_rate": 4.76739839162981e-06,
+      "loss": 0.6023,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5638102598782726,
+      "grad_norm": 0.6289366483688354,
+      "learning_rate": 4.764029840893446e-06,
+      "loss": 0.6121,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5640034779248382,
+      "grad_norm": 0.6573140025138855,
+      "learning_rate": 4.760661397496158e-06,
+      "loss": 0.6095,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5641966959714038,
+      "grad_norm": 0.7204758524894714,
+      "learning_rate": 4.757293062970194e-06,
+      "loss": 0.5964,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5643899140179692,
+      "grad_norm": 0.6620383262634277,
+      "learning_rate": 4.753924838847758e-06,
+      "loss": 0.5719,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5645831320645348,
+      "grad_norm": 0.6596202254295349,
+      "learning_rate": 4.750556726661001e-06,
+      "loss": 0.6676,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5647763501111004,
+      "grad_norm": 0.6132194995880127,
+      "learning_rate": 4.747188727942024e-06,
+      "loss": 0.5815,
+      "step": 2923
+    },
+    {
+      "epoch": 0.564969568157666,
+      "grad_norm": 0.6692056059837341,
+      "learning_rate": 4.743820844222877e-06,
+      "loss": 0.6516,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5651627862042314,
+      "grad_norm": 0.6077465415000916,
+      "learning_rate": 4.740453077035555e-06,
+      "loss": 0.5105,
+      "step": 2925
+    },
+    {
+      "epoch": 0.565356004250797,
+      "grad_norm": 0.5730334520339966,
+      "learning_rate": 4.737085427912003e-06,
+      "loss": 0.5085,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5655492222973626,
+      "grad_norm": 0.6925786137580872,
+      "learning_rate": 4.733717898384109e-06,
+      "loss": 0.642,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5657424403439282,
+      "grad_norm": 0.5852928757667542,
+      "learning_rate": 4.73035048998371e-06,
+      "loss": 0.5688,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5659356583904936,
+      "grad_norm": 0.6257668137550354,
+      "learning_rate": 4.72698320424259e-06,
+      "loss": 0.6616,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5661288764370592,
+      "grad_norm": 0.6533583998680115,
+      "learning_rate": 4.723616042692467e-06,
+      "loss": 0.636,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5663220944836248,
+      "grad_norm": 0.6532671451568604,
+      "learning_rate": 4.720249006865016e-06,
+      "loss": 0.6236,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5665153125301903,
+      "grad_norm": 0.6737897992134094,
+      "learning_rate": 4.7168820982918416e-06,
+      "loss": 0.6048,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5667085305767559,
+      "grad_norm": 0.6438847780227661,
+      "learning_rate": 4.713515318504501e-06,
+      "loss": 0.5749,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5669017486233214,
+      "grad_norm": 0.6344283223152161,
+      "learning_rate": 4.710148669034489e-06,
+      "loss": 0.6202,
+      "step": 2934
+    },
+    {
+      "epoch": 0.567094966669887,
+      "grad_norm": 0.5429412722587585,
+      "learning_rate": 4.706782151413239e-06,
+      "loss": 0.5609,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5672881847164525,
+      "grad_norm": 0.7498610615730286,
+      "learning_rate": 4.703415767172127e-06,
+      "loss": 0.6255,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5674814027630181,
+      "grad_norm": 0.679190456867218,
+      "learning_rate": 4.700049517842471e-06,
+      "loss": 0.6202,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5676746208095836,
+      "grad_norm": 0.672112226486206,
+      "learning_rate": 4.696683404955517e-06,
+      "loss": 0.6389,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5678678388561491,
+      "grad_norm": 0.6439762711524963,
+      "learning_rate": 4.693317430042465e-06,
+      "loss": 0.5587,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5680610569027147,
+      "grad_norm": 0.6281350255012512,
+      "learning_rate": 4.689951594634436e-06,
+      "loss": 0.5377,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5682542749492803,
+      "grad_norm": 0.675902783870697,
+      "learning_rate": 4.686585900262498e-06,
+      "loss": 0.6185,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5684474929958458,
+      "grad_norm": 0.677338182926178,
+      "learning_rate": 4.683220348457654e-06,
+      "loss": 0.5691,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5686407110424113,
+      "grad_norm": 0.7149407267570496,
+      "learning_rate": 4.679854940750835e-06,
+      "loss": 0.6566,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5688339290889769,
+      "grad_norm": 0.5532158017158508,
+      "learning_rate": 4.676489678672915e-06,
+      "loss": 0.5844,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5690271471355425,
+      "grad_norm": 0.6345710158348083,
+      "learning_rate": 4.6731245637547e-06,
+      "loss": 0.5789,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5692203651821081,
+      "grad_norm": 0.6669936180114746,
+      "learning_rate": 4.6697595975269236e-06,
+      "loss": 0.591,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5694135832286735,
+      "grad_norm": 0.6957157850265503,
+      "learning_rate": 4.666394781520257e-06,
+      "loss": 0.6169,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5696068012752391,
+      "grad_norm": 0.6178109049797058,
+      "learning_rate": 4.663030117265301e-06,
+      "loss": 0.6009,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5698000193218047,
+      "grad_norm": 0.6657687425613403,
+      "learning_rate": 4.659665606292587e-06,
+      "loss": 0.6214,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5699932373683702,
+      "grad_norm": 0.69082111120224,
+      "learning_rate": 4.656301250132581e-06,
+      "loss": 0.6158,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5701864554149357,
+      "grad_norm": 0.6488152742385864,
+      "learning_rate": 4.652937050315671e-06,
+      "loss": 0.6321,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5703796734615013,
+      "grad_norm": 0.6634805202484131,
+      "learning_rate": 4.649573008372182e-06,
+      "loss": 0.6367,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5705728915080669,
+      "grad_norm": 0.5881782174110413,
+      "learning_rate": 4.6462091258323586e-06,
+      "loss": 0.5885,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5707661095546324,
+      "grad_norm": 0.6777159571647644,
+      "learning_rate": 4.64284540422638e-06,
+      "loss": 0.678,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5709593276011979,
+      "grad_norm": 0.6688756346702576,
+      "learning_rate": 4.639481845084353e-06,
+      "loss": 0.5666,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5711525456477635,
+      "grad_norm": 0.689289927482605,
+      "learning_rate": 4.636118449936301e-06,
+      "loss": 0.6445,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5713457636943291,
+      "grad_norm": 0.7141535878181458,
+      "learning_rate": 4.6327552203121805e-06,
+      "loss": 0.6201,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5715389817408946,
+      "grad_norm": 0.6926259994506836,
+      "learning_rate": 4.629392157741876e-06,
+      "loss": 0.6818,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5717321997874601,
+      "grad_norm": 0.6699407696723938,
+      "learning_rate": 4.626029263755184e-06,
+      "loss": 0.6057,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5719254178340257,
+      "grad_norm": 0.688433825969696,
+      "learning_rate": 4.62266653988184e-06,
+      "loss": 0.6273,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5721186358805912,
+      "grad_norm": 0.6757651567459106,
+      "learning_rate": 4.619303987651485e-06,
+      "loss": 0.5936,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5723118539271568,
+      "grad_norm": 0.7014297842979431,
+      "learning_rate": 4.615941608593695e-06,
+      "loss": 0.649,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5725050719737224,
+      "grad_norm": 0.6267008185386658,
+      "learning_rate": 4.6125794042379645e-06,
+      "loss": 0.5332,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5726982900202879,
+      "grad_norm": 0.6772094964981079,
+      "learning_rate": 4.6092173761137025e-06,
+      "loss": 0.6803,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5728915080668534,
+      "grad_norm": 0.6433650255203247,
+      "learning_rate": 4.605855525750244e-06,
+      "loss": 0.5902,
+      "step": 2965
+    },
+    {
+      "epoch": 0.573084726113419,
+      "grad_norm": 0.7123873829841614,
+      "learning_rate": 4.602493854676842e-06,
+      "loss": 0.6462,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5732779441599846,
+      "grad_norm": 0.6595927476882935,
+      "learning_rate": 4.599132364422666e-06,
+      "loss": 0.67,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5734711622065501,
+      "grad_norm": 0.6980518698692322,
+      "learning_rate": 4.595771056516808e-06,
+      "loss": 0.6811,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5736643802531156,
+      "grad_norm": 0.6137881278991699,
+      "learning_rate": 4.592409932488267e-06,
+      "loss": 0.5305,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5738575982996812,
+      "grad_norm": 0.5613073110580444,
+      "learning_rate": 4.589048993865971e-06,
+      "loss": 0.563,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5740508163462468,
+      "grad_norm": 0.7470640540122986,
+      "learning_rate": 4.585688242178755e-06,
+      "loss": 0.6222,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5742440343928122,
+      "grad_norm": 0.6520352363586426,
+      "learning_rate": 4.58232767895537e-06,
+      "loss": 0.6506,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5744372524393778,
+      "grad_norm": 0.6754294633865356,
+      "learning_rate": 4.578967305724487e-06,
+      "loss": 0.6463,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5746304704859434,
+      "grad_norm": 0.7000090479850769,
+      "learning_rate": 4.575607124014682e-06,
+      "loss": 0.6771,
+      "step": 2974
+    },
+    {
+      "epoch": 0.574823688532509,
+      "grad_norm": 0.6640840172767639,
+      "learning_rate": 4.572247135354451e-06,
+      "loss": 0.6189,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5750169065790744,
+      "grad_norm": 0.6900555491447449,
+      "learning_rate": 4.568887341272199e-06,
+      "loss": 0.6091,
+      "step": 2976
+    },
+    {
+      "epoch": 0.57521012462564,
+      "grad_norm": 0.6496033668518066,
+      "learning_rate": 4.565527743296241e-06,
+      "loss": 0.551,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5754033426722056,
+      "grad_norm": 0.6979163885116577,
+      "learning_rate": 4.562168342954807e-06,
+      "loss": 0.5195,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5755965607187712,
+      "grad_norm": 0.7177830934524536,
+      "learning_rate": 4.558809141776034e-06,
+      "loss": 0.6643,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5757897787653367,
+      "grad_norm": 0.7084118127822876,
+      "learning_rate": 4.5554501412879685e-06,
+      "loss": 0.608,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5759829968119022,
+      "grad_norm": 0.6018376350402832,
+      "learning_rate": 4.552091343018568e-06,
+      "loss": 0.539,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5761762148584678,
+      "grad_norm": 0.5696696043014526,
+      "learning_rate": 4.5487327484956955e-06,
+      "loss": 0.5449,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5763694329050333,
+      "grad_norm": 0.6319146752357483,
+      "learning_rate": 4.545374359247121e-06,
+      "loss": 0.6223,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5765626509515989,
+      "grad_norm": 0.683325469493866,
+      "learning_rate": 4.542016176800527e-06,
+      "loss": 0.5642,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5767558689981644,
+      "grad_norm": 0.7015407681465149,
+      "learning_rate": 4.53865820268349e-06,
+      "loss": 0.6624,
+      "step": 2985
+    },
+    {
+      "epoch": 0.57694908704473,
+      "grad_norm": 0.733903706073761,
+      "learning_rate": 4.535300438423505e-06,
+      "loss": 0.6248,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5771423050912955,
+      "grad_norm": 0.7721959352493286,
+      "learning_rate": 4.5319428855479645e-06,
+      "loss": 0.6379,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5773355231378611,
+      "grad_norm": 0.6301102638244629,
+      "learning_rate": 4.528585545584164e-06,
+      "loss": 0.5911,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5775287411844267,
+      "grad_norm": 0.6760264039039612,
+      "learning_rate": 4.525228420059306e-06,
+      "loss": 0.6453,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5777219592309921,
+      "grad_norm": 0.6454445719718933,
+      "learning_rate": 4.5218715105004915e-06,
+      "loss": 0.5684,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5779151772775577,
+      "grad_norm": 0.6514859795570374,
+      "learning_rate": 4.518514818434726e-06,
+      "loss": 0.5438,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5781083953241233,
+      "grad_norm": 0.6480543613433838,
+      "learning_rate": 4.515158345388919e-06,
+      "loss": 0.5762,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5783016133706889,
+      "grad_norm": 0.6662483811378479,
+      "learning_rate": 4.51180209288987e-06,
+      "loss": 0.5878,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5784948314172543,
+      "grad_norm": 0.6456397771835327,
+      "learning_rate": 4.5084460624642904e-06,
+      "loss": 0.6312,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5786880494638199,
+      "grad_norm": 0.6031267642974854,
+      "learning_rate": 4.5050902556387825e-06,
+      "loss": 0.6059,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5788812675103855,
+      "grad_norm": 0.6470152139663696,
+      "learning_rate": 4.5017346739398495e-06,
+      "loss": 0.5947,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5790744855569511,
+      "grad_norm": 0.7066733837127686,
+      "learning_rate": 4.4983793188938965e-06,
+      "loss": 0.7198,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5792677036035165,
+      "grad_norm": 0.6355428099632263,
+      "learning_rate": 4.495024192027213e-06,
+      "loss": 0.5955,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5794609216500821,
+      "grad_norm": 0.6686714887619019,
+      "learning_rate": 4.491669294865999e-06,
+      "loss": 0.5864,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5796541396966477,
+      "grad_norm": 0.6329473853111267,
+      "learning_rate": 4.488314628936344e-06,
+      "loss": 0.6077,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5798473577432132,
+      "grad_norm": 0.6995109915733337,
+      "learning_rate": 4.4849601957642295e-06,
+      "loss": 0.6182,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5800405757897787,
+      "grad_norm": 0.6510365009307861,
+      "learning_rate": 4.4816059968755345e-06,
+      "loss": 0.5853,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5802337938363443,
+      "grad_norm": 0.6233105063438416,
+      "learning_rate": 4.47825203379603e-06,
+      "loss": 0.6063,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5804270118829099,
+      "grad_norm": 0.6427648663520813,
+      "learning_rate": 4.474898308051382e-06,
+      "loss": 0.539,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5806202299294754,
+      "grad_norm": 0.6461838483810425,
+      "learning_rate": 4.471544821167149e-06,
+      "loss": 0.6896,
+      "step": 3005
+    },
+    {
+      "epoch": 0.580813447976041,
+      "grad_norm": 0.6679365038871765,
+      "learning_rate": 4.468191574668774e-06,
+      "loss": 0.6525,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5810066660226065,
+      "grad_norm": 0.7439361214637756,
+      "learning_rate": 4.464838570081598e-06,
+      "loss": 0.6546,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5811998840691721,
+      "grad_norm": 0.6675832271575928,
+      "learning_rate": 4.4614858089308525e-06,
+      "loss": 0.5851,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5813931021157376,
+      "grad_norm": 0.6346164345741272,
+      "learning_rate": 4.458133292741649e-06,
+      "loss": 0.5647,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5815863201623032,
+      "grad_norm": 0.6693183779716492,
+      "learning_rate": 4.454781023038998e-06,
+      "loss": 0.6751,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5817795382088687,
+      "grad_norm": 0.6972511410713196,
+      "learning_rate": 4.4514290013477925e-06,
+      "loss": 0.6068,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5819727562554342,
+      "grad_norm": 0.6598070859909058,
+      "learning_rate": 4.448077229192814e-06,
+      "loss": 0.592,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5821659743019998,
+      "grad_norm": 0.6673294305801392,
+      "learning_rate": 4.444725708098732e-06,
+      "loss": 0.5689,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5823591923485654,
+      "grad_norm": 0.6916127800941467,
+      "learning_rate": 4.441374439590098e-06,
+      "loss": 0.5958,
+      "step": 3014
+    },
+    {
+      "epoch": 0.582552410395131,
+      "grad_norm": 0.6576099991798401,
+      "learning_rate": 4.43802342519135e-06,
+      "loss": 0.5939,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5827456284416964,
+      "grad_norm": 0.6135929822921753,
+      "learning_rate": 4.434672666426814e-06,
+      "loss": 0.6262,
+      "step": 3016
+    },
+    {
+      "epoch": 0.582938846488262,
+      "grad_norm": 0.7327417731285095,
+      "learning_rate": 4.431322164820696e-06,
+      "loss": 0.6285,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5831320645348276,
+      "grad_norm": 0.6808791756629944,
+      "learning_rate": 4.427971921897086e-06,
+      "loss": 0.5759,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5833252825813932,
+      "grad_norm": 0.6076217293739319,
+      "learning_rate": 4.424621939179953e-06,
+      "loss": 0.5904,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5835185006279586,
+      "grad_norm": 0.6603553295135498,
+      "learning_rate": 4.421272218193155e-06,
+      "loss": 0.6176,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5837117186745242,
+      "grad_norm": 0.6446527242660522,
+      "learning_rate": 4.417922760460428e-06,
+      "loss": 0.5855,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5839049367210898,
+      "grad_norm": 0.6930314898490906,
+      "learning_rate": 4.414573567505381e-06,
+      "loss": 0.6543,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5840981547676553,
+      "grad_norm": 0.5961169600486755,
+      "learning_rate": 4.411224640851513e-06,
+      "loss": 0.6081,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5842913728142208,
+      "grad_norm": 0.611328661441803,
+      "learning_rate": 4.407875982022197e-06,
+      "loss": 0.5932,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5844845908607864,
+      "grad_norm": 0.6698623299598694,
+      "learning_rate": 4.404527592540684e-06,
+      "loss": 0.583,
+      "step": 3025
+    },
+    {
+      "epoch": 0.584677808907352,
+      "grad_norm": 0.6034017205238342,
+      "learning_rate": 4.401179473930105e-06,
+      "loss": 0.622,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5848710269539175,
+      "grad_norm": 0.6452416181564331,
+      "learning_rate": 4.397831627713463e-06,
+      "loss": 0.6167,
+      "step": 3027
+    },
+    {
+      "epoch": 0.585064245000483,
+      "grad_norm": 0.6414170265197754,
+      "learning_rate": 4.39448405541364e-06,
+      "loss": 0.6447,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5852574630470486,
+      "grad_norm": 0.5835801362991333,
+      "learning_rate": 4.391136758553399e-06,
+      "loss": 0.5273,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5854506810936141,
+      "grad_norm": 0.6473480463027954,
+      "learning_rate": 4.387789738655365e-06,
+      "loss": 0.5787,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5856438991401797,
+      "grad_norm": 0.6619698405265808,
+      "learning_rate": 4.384442997242048e-06,
+      "loss": 0.5895,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5858371171867452,
+      "grad_norm": 0.6222323775291443,
+      "learning_rate": 4.381096535835825e-06,
+      "loss": 0.5934,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5860303352333108,
+      "grad_norm": 0.6699633598327637,
+      "learning_rate": 4.377750355958951e-06,
+      "loss": 0.5848,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5862235532798763,
+      "grad_norm": 0.7049209475517273,
+      "learning_rate": 4.37440445913355e-06,
+      "loss": 0.6468,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5864167713264419,
+      "grad_norm": 0.6795954704284668,
+      "learning_rate": 4.371058846881614e-06,
+      "loss": 0.6272,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5866099893730075,
+      "grad_norm": 0.6409685611724854,
+      "learning_rate": 4.367713520725011e-06,
+      "loss": 0.5475,
+      "step": 3036
+    },
+    {
+      "epoch": 0.586803207419573,
+      "grad_norm": 0.7316529750823975,
+      "learning_rate": 4.364368482185479e-06,
+      "loss": 0.6666,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5869964254661385,
+      "grad_norm": 0.6956066489219666,
+      "learning_rate": 4.361023732784617e-06,
+      "loss": 0.6167,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5871896435127041,
+      "grad_norm": 0.6802467107772827,
+      "learning_rate": 4.357679274043904e-06,
+      "loss": 0.6488,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5873828615592697,
+      "grad_norm": 0.6987923979759216,
+      "learning_rate": 4.354335107484678e-06,
+      "loss": 0.6129,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5875760796058351,
+      "grad_norm": 0.6633896827697754,
+      "learning_rate": 4.3509912346281475e-06,
+      "loss": 0.6297,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5877692976524007,
+      "grad_norm": 0.6447646021842957,
+      "learning_rate": 4.34764765699539e-06,
+      "loss": 0.6392,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5879625156989663,
+      "grad_norm": 0.6072761416435242,
+      "learning_rate": 4.344304376107341e-06,
+      "loss": 0.5689,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5881557337455319,
+      "grad_norm": 0.6146666407585144,
+      "learning_rate": 4.340961393484808e-06,
+      "loss": 0.5453,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5883489517920973,
+      "grad_norm": 0.6065789461135864,
+      "learning_rate": 4.337618710648465e-06,
+      "loss": 0.6176,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5885421698386629,
+      "grad_norm": 0.6589290499687195,
+      "learning_rate": 4.334276329118837e-06,
+      "loss": 0.6185,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5887353878852285,
+      "grad_norm": 0.6515271663665771,
+      "learning_rate": 4.330934250416327e-06,
+      "loss": 0.604,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5889286059317941,
+      "grad_norm": 0.6307036876678467,
+      "learning_rate": 4.327592476061189e-06,
+      "loss": 0.5665,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5891218239783595,
+      "grad_norm": 0.6181390285491943,
+      "learning_rate": 4.324251007573545e-06,
+      "loss": 0.5975,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5893150420249251,
+      "grad_norm": 0.7591662406921387,
+      "learning_rate": 4.320909846473379e-06,
+      "loss": 0.5705,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5895082600714907,
+      "grad_norm": 0.65484619140625,
+      "learning_rate": 4.317568994280525e-06,
+      "loss": 0.5636,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5897014781180562,
+      "grad_norm": 0.6706452965736389,
+      "learning_rate": 4.3142284525146915e-06,
+      "loss": 0.5855,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5898946961646218,
+      "grad_norm": 0.6329277753829956,
+      "learning_rate": 4.3108882226954305e-06,
+      "loss": 0.6209,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5900879142111873,
+      "grad_norm": 0.7050097584724426,
+      "learning_rate": 4.3075483063421636e-06,
+      "loss": 0.6407,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5902811322577529,
+      "grad_norm": 0.6224166750907898,
+      "learning_rate": 4.3042087049741654e-06,
+      "loss": 0.5488,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5904743503043184,
+      "grad_norm": 0.6544567346572876,
+      "learning_rate": 4.300869420110567e-06,
+      "loss": 0.658,
+      "step": 3056
+    },
+    {
+      "epoch": 0.590667568350884,
+      "grad_norm": 0.6712908148765564,
+      "learning_rate": 4.2975304532703545e-06,
+      "loss": 0.5929,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5908607863974495,
+      "grad_norm": 0.6762823462486267,
+      "learning_rate": 4.294191805972375e-06,
+      "loss": 0.6468,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5910540044440151,
+      "grad_norm": 0.7492475509643555,
+      "learning_rate": 4.290853479735321e-06,
+      "loss": 0.6623,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5912472224905806,
+      "grad_norm": 0.6332654356956482,
+      "learning_rate": 4.287515476077749e-06,
+      "loss": 0.602,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5914404405371462,
+      "grad_norm": 0.6222278475761414,
+      "learning_rate": 4.284177796518059e-06,
+      "loss": 0.6741,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5916336585837118,
+      "grad_norm": 0.6935966610908508,
+      "learning_rate": 4.280840442574511e-06,
+      "loss": 0.6934,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5918268766302772,
+      "grad_norm": 0.6677472591400146,
+      "learning_rate": 4.277503415765214e-06,
+      "loss": 0.6055,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5920200946768428,
+      "grad_norm": 0.6669917106628418,
+      "learning_rate": 4.274166717608128e-06,
+      "loss": 0.56,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5922133127234084,
+      "grad_norm": 0.6815713047981262,
+      "learning_rate": 4.270830349621064e-06,
+      "loss": 0.5902,
+      "step": 3065
+    },
+    {
+      "epoch": 0.592406530769974,
+      "grad_norm": 0.6139689683914185,
+      "learning_rate": 4.267494313321685e-06,
+      "loss": 0.5386,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5925997488165394,
+      "grad_norm": 0.6703992486000061,
+      "learning_rate": 4.264158610227497e-06,
+      "loss": 0.6304,
+      "step": 3067
+    },
+    {
+      "epoch": 0.592792966863105,
+      "grad_norm": 0.6625550985336304,
+      "learning_rate": 4.260823241855861e-06,
+      "loss": 0.5919,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5929861849096706,
+      "grad_norm": 0.693695068359375,
+      "learning_rate": 4.257488209723981e-06,
+      "loss": 0.6259,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5931794029562361,
+      "grad_norm": 0.7083082795143127,
+      "learning_rate": 4.254153515348913e-06,
+      "loss": 0.6501,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5933726210028016,
+      "grad_norm": 0.6486281156539917,
+      "learning_rate": 4.250819160247555e-06,
+      "loss": 0.5973,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5935658390493672,
+      "grad_norm": 0.624309778213501,
+      "learning_rate": 4.247485145936651e-06,
+      "loss": 0.5881,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5937590570959328,
+      "grad_norm": 0.615766704082489,
+      "learning_rate": 4.244151473932791e-06,
+      "loss": 0.6076,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5939522751424983,
+      "grad_norm": 0.6588672995567322,
+      "learning_rate": 4.240818145752414e-06,
+      "loss": 0.6329,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5941454931890638,
+      "grad_norm": 0.6576484441757202,
+      "learning_rate": 4.237485162911791e-06,
+      "loss": 0.6201,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5943387112356294,
+      "grad_norm": 0.6750476360321045,
+      "learning_rate": 4.234152526927048e-06,
+      "loss": 0.625,
+      "step": 3076
+    },
+    {
+      "epoch": 0.594531929282195,
+      "grad_norm": 0.6797803640365601,
+      "learning_rate": 4.2308202393141455e-06,
+      "loss": 0.6168,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5947251473287605,
+      "grad_norm": 0.6943289041519165,
+      "learning_rate": 4.227488301588889e-06,
+      "loss": 0.6226,
+      "step": 3078
+    },
+    {
+      "epoch": 0.594918365375326,
+      "grad_norm": 0.6153055429458618,
+      "learning_rate": 4.224156715266927e-06,
+      "loss": 0.5888,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5951115834218916,
+      "grad_norm": 0.6653738617897034,
+      "learning_rate": 4.22082548186374e-06,
+      "loss": 0.5847,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5953048014684571,
+      "grad_norm": 0.6674055457115173,
+      "learning_rate": 4.217494602894657e-06,
+      "loss": 0.6082,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5954980195150227,
+      "grad_norm": 0.646702766418457,
+      "learning_rate": 4.2141640798748386e-06,
+      "loss": 0.5685,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5956912375615883,
+      "grad_norm": 0.7081882953643799,
+      "learning_rate": 4.210833914319289e-06,
+      "loss": 0.6288,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5958844556081538,
+      "grad_norm": 0.7005528211593628,
+      "learning_rate": 4.207504107742848e-06,
+      "loss": 0.6172,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5960776736547193,
+      "grad_norm": 0.624311089515686,
+      "learning_rate": 4.204174661660189e-06,
+      "loss": 0.5202,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5962708917012849,
+      "grad_norm": 0.6389459371566772,
+      "learning_rate": 4.200845577585827e-06,
+      "loss": 0.583,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5964641097478505,
+      "grad_norm": 0.6603552103042603,
+      "learning_rate": 4.19751685703411e-06,
+      "loss": 0.5283,
+      "step": 3087
+    },
+    {
+      "epoch": 0.596657327794416,
+      "grad_norm": 0.6210460066795349,
+      "learning_rate": 4.194188501519215e-06,
+      "loss": 0.6898,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5968505458409815,
+      "grad_norm": 0.6569671034812927,
+      "learning_rate": 4.190860512555166e-06,
+      "loss": 0.6006,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5970437638875471,
+      "grad_norm": 0.6777694225311279,
+      "learning_rate": 4.1875328916558035e-06,
+      "loss": 0.5868,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5972369819341127,
+      "grad_norm": 0.6970713138580322,
+      "learning_rate": 4.184205640334814e-06,
+      "loss": 0.5922,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5974301999806781,
+      "grad_norm": 0.6388897895812988,
+      "learning_rate": 4.180878760105712e-06,
+      "loss": 0.5776,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5976234180272437,
+      "grad_norm": 0.6125625371932983,
+      "learning_rate": 4.177552252481841e-06,
+      "loss": 0.6124,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5978166360738093,
+      "grad_norm": 0.6168357729911804,
+      "learning_rate": 4.174226118976376e-06,
+      "loss": 0.5961,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5980098541203749,
+      "grad_norm": 0.649751603603363,
+      "learning_rate": 4.170900361102325e-06,
+      "loss": 0.6265,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5982030721669404,
+      "grad_norm": 0.7361144423484802,
+      "learning_rate": 4.167574980372518e-06,
+      "loss": 0.6751,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5983962902135059,
+      "grad_norm": 0.6557120084762573,
+      "learning_rate": 4.164249978299622e-06,
+      "loss": 0.6079,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5985895082600715,
+      "grad_norm": 0.6769595742225647,
+      "learning_rate": 4.160925356396125e-06,
+      "loss": 0.6333,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5987827263066371,
+      "grad_norm": 0.6935896873474121,
+      "learning_rate": 4.157601116174344e-06,
+      "loss": 0.6695,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5989759443532026,
+      "grad_norm": 0.6438478827476501,
+      "learning_rate": 4.154277259146426e-06,
+      "loss": 0.6174,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5991691623997681,
+      "grad_norm": 0.6886301636695862,
+      "learning_rate": 4.150953786824338e-06,
+      "loss": 0.5829,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5993623804463337,
+      "grad_norm": 0.6287814378738403,
+      "learning_rate": 4.147630700719878e-06,
+      "loss": 0.5737,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5995555984928992,
+      "grad_norm": 0.6288377046585083,
+      "learning_rate": 4.1443080023446605e-06,
+      "loss": 0.6323,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5997488165394648,
+      "grad_norm": 0.6151443123817444,
+      "learning_rate": 4.140985693210131e-06,
+      "loss": 0.6182,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5999420345860303,
+      "grad_norm": 0.667919397354126,
+      "learning_rate": 4.1376637748275566e-06,
+      "loss": 0.6162,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6001352526325959,
+      "grad_norm": 0.6643292307853699,
+      "learning_rate": 4.134342248708021e-06,
+      "loss": 0.6846,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6003284706791614,
+      "grad_norm": 0.7210611701011658,
+      "learning_rate": 4.131021116362437e-06,
+      "loss": 0.631,
+      "step": 3107
+    },
+    {
+      "epoch": 0.600521688725727,
+      "grad_norm": 0.7405746579170227,
+      "learning_rate": 4.127700379301534e-06,
+      "loss": 0.6081,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6007149067722926,
+      "grad_norm": 0.5900118947029114,
+      "learning_rate": 4.124380039035861e-06,
+      "loss": 0.5814,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6009081248188581,
+      "grad_norm": 0.6563978791236877,
+      "learning_rate": 4.1210600970757915e-06,
+      "loss": 0.5147,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6011013428654236,
+      "grad_norm": 0.6381539106369019,
+      "learning_rate": 4.11774055493151e-06,
+      "loss": 0.6691,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6012945609119892,
+      "grad_norm": 0.6472948789596558,
+      "learning_rate": 4.1144214141130254e-06,
+      "loss": 0.554,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6014877789585548,
+      "grad_norm": 0.6395518183708191,
+      "learning_rate": 4.111102676130164e-06,
+      "loss": 0.5778,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6016809970051202,
+      "grad_norm": 0.6823351979255676,
+      "learning_rate": 4.107784342492565e-06,
+      "loss": 0.6007,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6018742150516858,
+      "grad_norm": 0.6574017405509949,
+      "learning_rate": 4.104466414709685e-06,
+      "loss": 0.6306,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6020674330982514,
+      "grad_norm": 0.694908618927002,
+      "learning_rate": 4.101148894290798e-06,
+      "loss": 0.6145,
+      "step": 3116
+    },
+    {
+      "epoch": 0.602260651144817,
+      "grad_norm": 0.5726170539855957,
+      "learning_rate": 4.09783178274499e-06,
+      "loss": 0.6373,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6024538691913824,
+      "grad_norm": 0.5958892107009888,
+      "learning_rate": 4.094515081581166e-06,
+      "loss": 0.6192,
+      "step": 3118
+    },
+    {
+      "epoch": 0.602647087237948,
+      "grad_norm": 0.6423183083534241,
+      "learning_rate": 4.091198792308034e-06,
+      "loss": 0.6271,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6028403052845136,
+      "grad_norm": 0.6943200826644897,
+      "learning_rate": 4.087882916434126e-06,
+      "loss": 0.6304,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6030335233310791,
+      "grad_norm": 0.7092317342758179,
+      "learning_rate": 4.084567455467782e-06,
+      "loss": 0.6263,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6032267413776446,
+      "grad_norm": 0.7151508331298828,
+      "learning_rate": 4.081252410917148e-06,
+      "loss": 0.5964,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6034199594242102,
+      "grad_norm": 0.6772610545158386,
+      "learning_rate": 4.0779377842901895e-06,
+      "loss": 0.5915,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6036131774707758,
+      "grad_norm": 0.6566310524940491,
+      "learning_rate": 4.0746235770946765e-06,
+      "loss": 0.6453,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6038063955173413,
+      "grad_norm": 0.6607003808021545,
+      "learning_rate": 4.071309790838186e-06,
+      "loss": 0.6569,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6039996135639069,
+      "grad_norm": 0.6405696272850037,
+      "learning_rate": 4.067996427028113e-06,
+      "loss": 0.6101,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6041928316104724,
+      "grad_norm": 0.5764988660812378,
+      "learning_rate": 4.064683487171647e-06,
+      "loss": 0.5958,
+      "step": 3127
+    },
+    {
+      "epoch": 0.604386049657038,
+      "grad_norm": 0.6420401334762573,
+      "learning_rate": 4.061370972775795e-06,
+      "loss": 0.6421,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6045792677036035,
+      "grad_norm": 0.6833434104919434,
+      "learning_rate": 4.058058885347368e-06,
+      "loss": 0.6452,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6047724857501691,
+      "grad_norm": 0.6693868041038513,
+      "learning_rate": 4.0547472263929796e-06,
+      "loss": 0.6434,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6049657037967346,
+      "grad_norm": 0.591437578201294,
+      "learning_rate": 4.051435997419054e-06,
+      "loss": 0.6091,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6051589218433001,
+      "grad_norm": 0.6181185841560364,
+      "learning_rate": 4.048125199931813e-06,
+      "loss": 0.5442,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6053521398898657,
+      "grad_norm": 0.6737474203109741,
+      "learning_rate": 4.044814835437287e-06,
+      "loss": 0.5597,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6055453579364313,
+      "grad_norm": 0.7228105664253235,
+      "learning_rate": 4.041504905441312e-06,
+      "loss": 0.5918,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6057385759829969,
+      "grad_norm": 0.7036318182945251,
+      "learning_rate": 4.038195411449516e-06,
+      "loss": 0.6169,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6059317940295623,
+      "grad_norm": 0.6438372135162354,
+      "learning_rate": 4.034886354967339e-06,
+      "loss": 0.6613,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6061250120761279,
+      "grad_norm": 0.6414713859558105,
+      "learning_rate": 4.0315777375000185e-06,
+      "loss": 0.5718,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6063182301226935,
+      "grad_norm": 0.6760304570198059,
+      "learning_rate": 4.02826956055259e-06,
+      "loss": 0.5665,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6065114481692591,
+      "grad_norm": 0.6948249340057373,
+      "learning_rate": 4.024961825629893e-06,
+      "loss": 0.6618,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6067046662158245,
+      "grad_norm": 0.6644945740699768,
+      "learning_rate": 4.021654534236561e-06,
+      "loss": 0.6531,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6068978842623901,
+      "grad_norm": 0.7183898687362671,
+      "learning_rate": 4.01834768787703e-06,
+      "loss": 0.6418,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6070911023089557,
+      "grad_norm": 0.6848105788230896,
+      "learning_rate": 4.015041288055533e-06,
+      "loss": 0.5061,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6072843203555212,
+      "grad_norm": 0.6214210391044617,
+      "learning_rate": 4.011735336276094e-06,
+      "loss": 0.5793,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6074775384020867,
+      "grad_norm": 0.7016898393630981,
+      "learning_rate": 4.008429834042542e-06,
+      "loss": 0.6406,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6076707564486523,
+      "grad_norm": 0.615348219871521,
+      "learning_rate": 4.005124782858497e-06,
+      "loss": 0.5558,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6078639744952179,
+      "grad_norm": 0.6601006388664246,
+      "learning_rate": 4.001820184227373e-06,
+      "loss": 0.5933,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6080571925417834,
+      "grad_norm": 0.7249693274497986,
+      "learning_rate": 3.998516039652382e-06,
+      "loss": 0.6947,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6082504105883489,
+      "grad_norm": 0.6220579743385315,
+      "learning_rate": 3.995212350636523e-06,
+      "loss": 0.5406,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6084436286349145,
+      "grad_norm": 0.7134974002838135,
+      "learning_rate": 3.9919091186825935e-06,
+      "loss": 0.589,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6086368466814801,
+      "grad_norm": 0.6719569563865662,
+      "learning_rate": 3.988606345293183e-06,
+      "loss": 0.6127,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6088300647280456,
+      "grad_norm": 0.6442338228225708,
+      "learning_rate": 3.985304031970667e-06,
+      "loss": 0.6007,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6090232827746112,
+      "grad_norm": 0.6382316946983337,
+      "learning_rate": 3.982002180217218e-06,
+      "loss": 0.6055,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6092165008211767,
+      "grad_norm": 0.6540247797966003,
+      "learning_rate": 3.978700791534794e-06,
+      "loss": 0.6002,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6094097188677422,
+      "grad_norm": 0.6548553705215454,
+      "learning_rate": 3.975399867425146e-06,
+      "loss": 0.5409,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6096029369143078,
+      "grad_norm": 0.706505298614502,
+      "learning_rate": 3.9720994093898125e-06,
+      "loss": 0.5849,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6097961549608734,
+      "grad_norm": 0.6191006302833557,
+      "learning_rate": 3.968799418930116e-06,
+      "loss": 0.6606,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6099893730074389,
+      "grad_norm": 0.6425442099571228,
+      "learning_rate": 3.965499897547172e-06,
+      "loss": 0.5722,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6101825910540044,
+      "grad_norm": 0.6770890951156616,
+      "learning_rate": 3.962200846741882e-06,
+      "loss": 0.6476,
+      "step": 3158
+    },
+    {
+      "epoch": 0.61037580910057,
+      "grad_norm": 0.6407185792922974,
+      "learning_rate": 3.958902268014928e-06,
+      "loss": 0.5508,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6105690271471356,
+      "grad_norm": 0.6762952208518982,
+      "learning_rate": 3.955604162866784e-06,
+      "loss": 0.6261,
+      "step": 3160
+    },
+    {
+      "epoch": 0.610762245193701,
+      "grad_norm": 0.6587925553321838,
+      "learning_rate": 3.952306532797703e-06,
+      "loss": 0.6043,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6109554632402666,
+      "grad_norm": 0.6428242325782776,
+      "learning_rate": 3.949009379307726e-06,
+      "loss": 0.6023,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6111486812868322,
+      "grad_norm": 0.6767794489860535,
+      "learning_rate": 3.9457127038966795e-06,
+      "loss": 0.5936,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6113418993333978,
+      "grad_norm": 0.6042398810386658,
+      "learning_rate": 3.942416508064161e-06,
+      "loss": 0.6314,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6115351173799632,
+      "grad_norm": 0.6889746785163879,
+      "learning_rate": 3.939120793309562e-06,
+      "loss": 0.6057,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6117283354265288,
+      "grad_norm": 0.6757248640060425,
+      "learning_rate": 3.935825561132052e-06,
+      "loss": 0.5991,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6119215534730944,
+      "grad_norm": 0.7389405965805054,
+      "learning_rate": 3.932530813030576e-06,
+      "loss": 0.6518,
+      "step": 3167
+    },
+    {
+      "epoch": 0.61211477151966,
+      "grad_norm": 0.6088101863861084,
+      "learning_rate": 3.929236550503867e-06,
+      "loss": 0.5635,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6123079895662255,
+      "grad_norm": 0.5785213112831116,
+      "learning_rate": 3.925942775050429e-06,
+      "loss": 0.5472,
+      "step": 3169
+    },
+    {
+      "epoch": 0.612501207612791,
+      "grad_norm": 0.6213880777359009,
+      "learning_rate": 3.922649488168548e-06,
+      "loss": 0.5217,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6126944256593566,
+      "grad_norm": 0.6397833824157715,
+      "learning_rate": 3.9193566913562915e-06,
+      "loss": 0.652,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6128876437059221,
+      "grad_norm": 0.6624322533607483,
+      "learning_rate": 3.9160643861114965e-06,
+      "loss": 0.6115,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6130808617524877,
+      "grad_norm": 0.6826323866844177,
+      "learning_rate": 3.912772573931782e-06,
+      "loss": 0.6469,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6132740797990532,
+      "grad_norm": 0.7239211797714233,
+      "learning_rate": 3.9094812563145425e-06,
+      "loss": 0.6942,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6134672978456188,
+      "grad_norm": 0.6406311988830566,
+      "learning_rate": 3.906190434756942e-06,
+      "loss": 0.6105,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6136605158921843,
+      "grad_norm": 0.6539967656135559,
+      "learning_rate": 3.902900110755929e-06,
+      "loss": 0.5918,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6138537339387499,
+      "grad_norm": 0.608866274356842,
+      "learning_rate": 3.899610285808213e-06,
+      "loss": 0.5654,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6140469519853154,
+      "grad_norm": 0.7390885949134827,
+      "learning_rate": 3.896320961410286e-06,
+      "loss": 0.6596,
+      "step": 3178
+    },
+    {
+      "epoch": 0.614240170031881,
+      "grad_norm": 0.6658431887626648,
+      "learning_rate": 3.893032139058413e-06,
+      "loss": 0.6292,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6144333880784465,
+      "grad_norm": 0.6442549824714661,
+      "learning_rate": 3.88974382024862e-06,
+      "loss": 0.6348,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6146266061250121,
+      "grad_norm": 0.6331585645675659,
+      "learning_rate": 3.886456006476716e-06,
+      "loss": 0.5812,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6148198241715777,
+      "grad_norm": 0.6637753248214722,
+      "learning_rate": 3.8831686992382725e-06,
+      "loss": 0.6232,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6150130422181431,
+      "grad_norm": 0.6564189195632935,
+      "learning_rate": 3.879881900028634e-06,
+      "loss": 0.5707,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6152062602647087,
+      "grad_norm": 0.6711429357528687,
+      "learning_rate": 3.876595610342917e-06,
+      "loss": 0.6315,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6153994783112743,
+      "grad_norm": 0.7061915397644043,
+      "learning_rate": 3.873309831675996e-06,
+      "loss": 0.629,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6155926963578399,
+      "grad_norm": 0.5790874361991882,
+      "learning_rate": 3.870024565522523e-06,
+      "loss": 0.5787,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6157859144044053,
+      "grad_norm": 0.6867915987968445,
+      "learning_rate": 3.866739813376915e-06,
+      "loss": 0.5791,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6159791324509709,
+      "grad_norm": 0.6547077894210815,
+      "learning_rate": 3.863455576733349e-06,
+      "loss": 0.6645,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6161723504975365,
+      "grad_norm": 0.6758971810340881,
+      "learning_rate": 3.860171857085776e-06,
+      "loss": 0.6169,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6163655685441021,
+      "grad_norm": 0.7291857600212097,
+      "learning_rate": 3.8568886559279054e-06,
+      "loss": 0.6001,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6165587865906675,
+      "grad_norm": 0.6720841526985168,
+      "learning_rate": 3.853605974753214e-06,
+      "loss": 0.6613,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6167520046372331,
+      "grad_norm": 0.6910128593444824,
+      "learning_rate": 3.850323815054944e-06,
+      "loss": 0.6924,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6169452226837987,
+      "grad_norm": 0.6642037630081177,
+      "learning_rate": 3.847042178326093e-06,
+      "loss": 0.6224,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6171384407303642,
+      "grad_norm": 0.6651246547698975,
+      "learning_rate": 3.843761066059427e-06,
+      "loss": 0.6251,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6173316587769297,
+      "grad_norm": 0.7186761498451233,
+      "learning_rate": 3.840480479747476e-06,
+      "loss": 0.6417,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6175248768234953,
+      "grad_norm": 0.6934588551521301,
+      "learning_rate": 3.83720042088252e-06,
+      "loss": 0.6355,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6177180948700609,
+      "grad_norm": 0.6388946771621704,
+      "learning_rate": 3.8339208909566085e-06,
+      "loss": 0.6117,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6179113129166264,
+      "grad_norm": 0.6608049273490906,
+      "learning_rate": 3.830641891461548e-06,
+      "loss": 0.5629,
+      "step": 3198
+    },
+    {
+      "epoch": 0.618104530963192,
+      "grad_norm": 0.6073363423347473,
+      "learning_rate": 3.827363423888903e-06,
+      "loss": 0.5774,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6182977490097575,
+      "grad_norm": 0.6730050444602966,
+      "learning_rate": 3.824085489729997e-06,
+      "loss": 0.5929,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6184909670563231,
+      "grad_norm": 0.6740630269050598,
+      "learning_rate": 3.820808090475907e-06,
+      "loss": 0.5794,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6186841851028886,
+      "grad_norm": 0.640056848526001,
+      "learning_rate": 3.817531227617476e-06,
+      "loss": 0.5868,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6188774031494542,
+      "grad_norm": 0.6819208264350891,
+      "learning_rate": 3.8142549026452898e-06,
+      "loss": 0.5774,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6190706211960197,
+      "grad_norm": 0.6388457417488098,
+      "learning_rate": 3.8109791170496996e-06,
+      "loss": 0.5722,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6192638392425852,
+      "grad_norm": 0.6221622228622437,
+      "learning_rate": 3.807703872320809e-06,
+      "loss": 0.6525,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6194570572891508,
+      "grad_norm": 0.6043922305107117,
+      "learning_rate": 3.8044291699484742e-06,
+      "loss": 0.5724,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6196502753357164,
+      "grad_norm": 0.5924354195594788,
+      "learning_rate": 3.801155011422304e-06,
+      "loss": 0.5907,
+      "step": 3207
+    },
+    {
+      "epoch": 0.619843493382282,
+      "grad_norm": 0.689477264881134,
+      "learning_rate": 3.797881398231665e-06,
+      "loss": 0.5825,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6200367114288474,
+      "grad_norm": 0.74201500415802,
+      "learning_rate": 3.7946083318656667e-06,
+      "loss": 0.6568,
+      "step": 3209
+    },
+    {
+      "epoch": 0.620229929475413,
+      "grad_norm": 0.7243762612342834,
+      "learning_rate": 3.7913358138131777e-06,
+      "loss": 0.6389,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6204231475219786,
+      "grad_norm": 0.6657256484031677,
+      "learning_rate": 3.788063845562815e-06,
+      "loss": 0.5433,
+      "step": 3211
+    },
+    {
+      "epoch": 0.620616365568544,
+      "grad_norm": 0.8499423861503601,
+      "learning_rate": 3.784792428602942e-06,
+      "loss": 0.6677,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6208095836151096,
+      "grad_norm": 0.6967642903327942,
+      "learning_rate": 3.7815215644216776e-06,
+      "loss": 0.5654,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6210028016616752,
+      "grad_norm": 0.6031731367111206,
+      "learning_rate": 3.7782512545068835e-06,
+      "loss": 0.611,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6211960197082408,
+      "grad_norm": 0.6164298057556152,
+      "learning_rate": 3.774981500346172e-06,
+      "loss": 0.5819,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6213892377548063,
+      "grad_norm": 0.7172102332115173,
+      "learning_rate": 3.771712303426906e-06,
+      "loss": 0.6206,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6215824558013718,
+      "grad_norm": 0.699630081653595,
+      "learning_rate": 3.7684436652361855e-06,
+      "loss": 0.5906,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6217756738479374,
+      "grad_norm": 0.5986834764480591,
+      "learning_rate": 3.7651755872608663e-06,
+      "loss": 0.5374,
+      "step": 3218
+    },
+    {
+      "epoch": 0.621968891894503,
+      "grad_norm": 0.6978106498718262,
+      "learning_rate": 3.7619080709875428e-06,
+      "loss": 0.654,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6221621099410685,
+      "grad_norm": 0.6366021037101746,
+      "learning_rate": 3.7586411179025562e-06,
+      "loss": 0.5798,
+      "step": 3220
+    },
+    {
+      "epoch": 0.622355327987634,
+      "grad_norm": 0.6151782870292664,
+      "learning_rate": 3.755374729491995e-06,
+      "loss": 0.5771,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6225485460341996,
+      "grad_norm": 0.7137126326560974,
+      "learning_rate": 3.752108907241682e-06,
+      "loss": 0.5622,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6227417640807651,
+      "grad_norm": 0.7322572469711304,
+      "learning_rate": 3.7488436526371915e-06,
+      "loss": 0.6978,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6229349821273307,
+      "grad_norm": 0.6011915802955627,
+      "learning_rate": 3.7455789671638363e-06,
+      "loss": 0.5993,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6231282001738963,
+      "grad_norm": 0.6842527389526367,
+      "learning_rate": 3.7423148523066664e-06,
+      "loss": 0.6356,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6233214182204618,
+      "grad_norm": 0.639638364315033,
+      "learning_rate": 3.7390513095504793e-06,
+      "loss": 0.5992,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6235146362670273,
+      "grad_norm": 0.6641109585762024,
+      "learning_rate": 3.735788340379806e-06,
+      "loss": 0.6287,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6237078543135929,
+      "grad_norm": 0.6206823587417603,
+      "learning_rate": 3.732525946278921e-06,
+      "loss": 0.5655,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6239010723601585,
+      "grad_norm": 0.6205646991729736,
+      "learning_rate": 3.7292641287318372e-06,
+      "loss": 0.5625,
+      "step": 3229
+    },
+    {
+      "epoch": 0.624094290406724,
+      "grad_norm": 0.6578320264816284,
+      "learning_rate": 3.7260028892222993e-06,
+      "loss": 0.609,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6242875084532895,
+      "grad_norm": 0.7106614708900452,
+      "learning_rate": 3.7227422292337978e-06,
+      "loss": 0.6553,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6244807264998551,
+      "grad_norm": 0.673772394657135,
+      "learning_rate": 3.7194821502495503e-06,
+      "loss": 0.6377,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6246739445464207,
+      "grad_norm": 0.6769312024116516,
+      "learning_rate": 3.716222653752518e-06,
+      "loss": 0.6373,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6248671625929861,
+      "grad_norm": 0.6099493503570557,
+      "learning_rate": 3.7129637412253937e-06,
+      "loss": 0.5603,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6250603806395517,
+      "grad_norm": 0.683137059211731,
+      "learning_rate": 3.709705414150605e-06,
+      "loss": 0.5888,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6252535986861173,
+      "grad_norm": 0.6549149751663208,
+      "learning_rate": 3.7064476740103115e-06,
+      "loss": 0.5821,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6254468167326829,
+      "grad_norm": 0.6492919325828552,
+      "learning_rate": 3.7031905222864116e-06,
+      "loss": 0.6255,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6256400347792483,
+      "grad_norm": 0.7013925909996033,
+      "learning_rate": 3.6999339604605266e-06,
+      "loss": 0.6261,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6258332528258139,
+      "grad_norm": 0.6394886374473572,
+      "learning_rate": 3.6966779900140193e-06,
+      "loss": 0.5939,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6260264708723795,
+      "grad_norm": 0.657923698425293,
+      "learning_rate": 3.693422612427974e-06,
+      "loss": 0.6478,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6262196889189451,
+      "grad_norm": 0.6997016072273254,
+      "learning_rate": 3.690167829183213e-06,
+      "loss": 0.6012,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6264129069655106,
+      "grad_norm": 0.6997055411338806,
+      "learning_rate": 3.6869136417602862e-06,
+      "loss": 0.6564,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6266061250120761,
+      "grad_norm": 0.6959564089775085,
+      "learning_rate": 3.6836600516394705e-06,
+      "loss": 0.5504,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6267993430586417,
+      "grad_norm": 0.7039137482643127,
+      "learning_rate": 3.680407060300772e-06,
+      "loss": 0.5987,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6269925611052072,
+      "grad_norm": 0.6388022899627686,
+      "learning_rate": 3.677154669223928e-06,
+      "loss": 0.6175,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6271857791517728,
+      "grad_norm": 0.6844452023506165,
+      "learning_rate": 3.6739028798883945e-06,
+      "loss": 0.5983,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6273789971983383,
+      "grad_norm": 0.692890465259552,
+      "learning_rate": 3.6706516937733634e-06,
+      "loss": 0.6443,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6275722152449039,
+      "grad_norm": 0.6511271595954895,
+      "learning_rate": 3.667401112357743e-06,
+      "loss": 0.6067,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6277654332914694,
+      "grad_norm": 0.618399977684021,
+      "learning_rate": 3.6641511371201737e-06,
+      "loss": 0.6139,
+      "step": 3249
+    },
+    {
+      "epoch": 0.627958651338035,
+      "grad_norm": 0.634907603263855,
+      "learning_rate": 3.6609017695390187e-06,
+      "loss": 0.6346,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6281518693846005,
+      "grad_norm": 0.6481157541275024,
+      "learning_rate": 3.657653011092361e-06,
+      "loss": 0.5487,
+      "step": 3251
+    },
+    {
+      "epoch": 0.628345087431166,
+      "grad_norm": 0.6779963970184326,
+      "learning_rate": 3.6544048632580127e-06,
+      "loss": 0.5302,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6285383054777316,
+      "grad_norm": 0.6489970088005066,
+      "learning_rate": 3.6511573275135004e-06,
+      "loss": 0.637,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6287315235242972,
+      "grad_norm": 0.795566201210022,
+      "learning_rate": 3.647910405336078e-06,
+      "loss": 0.6085,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6289247415708628,
+      "grad_norm": 0.6367505788803101,
+      "learning_rate": 3.6446640982027217e-06,
+      "loss": 0.6398,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6291179596174282,
+      "grad_norm": 0.6749064326286316,
+      "learning_rate": 3.6414184075901206e-06,
+      "loss": 0.5846,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6293111776639938,
+      "grad_norm": 0.6027222871780396,
+      "learning_rate": 3.6381733349746895e-06,
+      "loss": 0.5859,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6295043957105594,
+      "grad_norm": 0.5859025716781616,
+      "learning_rate": 3.6349288818325624e-06,
+      "loss": 0.5705,
+      "step": 3258
+    },
+    {
+      "epoch": 0.629697613757125,
+      "grad_norm": 0.7081364989280701,
+      "learning_rate": 3.6316850496395863e-06,
+      "loss": 0.6552,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6298908318036904,
+      "grad_norm": 0.6194645166397095,
+      "learning_rate": 3.628441839871332e-06,
+      "loss": 0.6389,
+      "step": 3260
+    },
+    {
+      "epoch": 0.630084049850256,
+      "grad_norm": 0.669784426689148,
+      "learning_rate": 3.62519925400308e-06,
+      "loss": 0.6153,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6302772678968216,
+      "grad_norm": 0.6537437438964844,
+      "learning_rate": 3.621957293509834e-06,
+      "loss": 0.5991,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6304704859433871,
+      "grad_norm": 0.7451925277709961,
+      "learning_rate": 3.6187159598663103e-06,
+      "loss": 0.6191,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6306637039899526,
+      "grad_norm": 0.6728523969650269,
+      "learning_rate": 3.615475254546939e-06,
+      "loss": 0.6236,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6308569220365182,
+      "grad_norm": 0.6399971842765808,
+      "learning_rate": 3.612235179025867e-06,
+      "loss": 0.6147,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6310501400830838,
+      "grad_norm": 0.6588414907455444,
+      "learning_rate": 3.608995734776951e-06,
+      "loss": 0.5754,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6312433581296493,
+      "grad_norm": 0.6864632964134216,
+      "learning_rate": 3.605756923273764e-06,
+      "loss": 0.6194,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6314365761762148,
+      "grad_norm": 0.6336121559143066,
+      "learning_rate": 3.6025187459895907e-06,
+      "loss": 0.5371,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6316297942227804,
+      "grad_norm": 0.6088796854019165,
+      "learning_rate": 3.5992812043974236e-06,
+      "loss": 0.6337,
+      "step": 3269
+    },
+    {
+      "epoch": 0.631823012269346,
+      "grad_norm": 0.7010225057601929,
+      "learning_rate": 3.5960442999699707e-06,
+      "loss": 0.6051,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6320162303159115,
+      "grad_norm": 0.6600781083106995,
+      "learning_rate": 3.59280803417965e-06,
+      "loss": 0.5972,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6322094483624771,
+      "grad_norm": 0.6569859981536865,
+      "learning_rate": 3.589572408498584e-06,
+      "loss": 0.626,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6324026664090426,
+      "grad_norm": 0.6797444820404053,
+      "learning_rate": 3.586337424398609e-06,
+      "loss": 0.5846,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6325958844556081,
+      "grad_norm": 0.6384918689727783,
+      "learning_rate": 3.5831030833512704e-06,
+      "loss": 0.5974,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6327891025021737,
+      "grad_norm": 0.5791540741920471,
+      "learning_rate": 3.579869386827814e-06,
+      "loss": 0.6056,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6329823205487393,
+      "grad_norm": 0.6680526733398438,
+      "learning_rate": 3.5766363362992007e-06,
+      "loss": 0.5872,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6331755385953048,
+      "grad_norm": 0.6964307427406311,
+      "learning_rate": 3.5734039332360902e-06,
+      "loss": 0.5956,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6333687566418703,
+      "grad_norm": 0.602342963218689,
+      "learning_rate": 3.5701721791088536e-06,
+      "loss": 0.503,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6335619746884359,
+      "grad_norm": 0.7309747934341431,
+      "learning_rate": 3.5669410753875643e-06,
+      "loss": 0.6285,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6337551927350015,
+      "grad_norm": 0.6663960218429565,
+      "learning_rate": 3.5637106235419994e-06,
+      "loss": 0.6029,
+      "step": 3280
+    },
+    {
+      "epoch": 0.633948410781567,
+      "grad_norm": 0.6081252098083496,
+      "learning_rate": 3.560480825041643e-06,
+      "loss": 0.5294,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6341416288281325,
+      "grad_norm": 0.6353676319122314,
+      "learning_rate": 3.5572516813556736e-06,
+      "loss": 0.5957,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6343348468746981,
+      "grad_norm": 0.6464443802833557,
+      "learning_rate": 3.5540231939529806e-06,
+      "loss": 0.6019,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6345280649212637,
+      "grad_norm": 0.6857627034187317,
+      "learning_rate": 3.5507953643021532e-06,
+      "loss": 0.6425,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6347212829678291,
+      "grad_norm": 0.64161217212677,
+      "learning_rate": 3.547568193871476e-06,
+      "loss": 0.576,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6349145010143947,
+      "grad_norm": 0.5956976413726807,
+      "learning_rate": 3.5443416841289397e-06,
+      "loss": 0.5318,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6351077190609603,
+      "grad_norm": 0.6842436194419861,
+      "learning_rate": 3.5411158365422325e-06,
+      "loss": 0.6532,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6353009371075259,
+      "grad_norm": 0.6533595323562622,
+      "learning_rate": 3.5378906525787404e-06,
+      "loss": 0.5929,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6354941551540914,
+      "grad_norm": 0.7157329320907593,
+      "learning_rate": 3.5346661337055506e-06,
+      "loss": 0.6844,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6356873732006569,
+      "grad_norm": 0.6525877714157104,
+      "learning_rate": 3.5314422813894413e-06,
+      "loss": 0.5706,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6358805912472225,
+      "grad_norm": 0.6242631077766418,
+      "learning_rate": 3.5282190970968935e-06,
+      "loss": 0.5562,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6360738092937881,
+      "grad_norm": 0.6606700420379639,
+      "learning_rate": 3.524996582294085e-06,
+      "loss": 0.5717,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6362670273403536,
+      "grad_norm": 0.6717787981033325,
+      "learning_rate": 3.5217747384468827e-06,
+      "loss": 0.6289,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6364602453869191,
+      "grad_norm": 0.718144416809082,
+      "learning_rate": 3.518553567020854e-06,
+      "loss": 0.6369,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6366534634334847,
+      "grad_norm": 0.6808081269264221,
+      "learning_rate": 3.515333069481259e-06,
+      "loss": 0.6371,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6368466814800502,
+      "grad_norm": 0.6526963710784912,
+      "learning_rate": 3.5121132472930498e-06,
+      "loss": 0.6405,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6370398995266158,
+      "grad_norm": 0.6198930144309998,
+      "learning_rate": 3.5088941019208744e-06,
+      "loss": 0.5938,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6372331175731814,
+      "grad_norm": 0.6532193422317505,
+      "learning_rate": 3.505675634829068e-06,
+      "loss": 0.5732,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6374263356197469,
+      "grad_norm": 0.6823054552078247,
+      "learning_rate": 3.5024578474816606e-06,
+      "loss": 0.611,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6376195536663124,
+      "grad_norm": 0.6453916430473328,
+      "learning_rate": 3.4992407413423756e-06,
+      "loss": 0.5938,
+      "step": 3300
+    },
+    {
+      "epoch": 0.637812771712878,
+      "grad_norm": 0.6656875014305115,
+      "learning_rate": 3.49602431787462e-06,
+      "loss": 0.55,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6380059897594436,
+      "grad_norm": 0.6388484239578247,
+      "learning_rate": 3.4928085785414944e-06,
+      "loss": 0.6029,
+      "step": 3302
+    },
+    {
+      "epoch": 0.638199207806009,
+      "grad_norm": 0.6700331568717957,
+      "learning_rate": 3.489593524805788e-06,
+      "loss": 0.5922,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6383924258525746,
+      "grad_norm": 0.6447831988334656,
+      "learning_rate": 3.486379158129978e-06,
+      "loss": 0.5599,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6385856438991402,
+      "grad_norm": 0.6811746954917908,
+      "learning_rate": 3.4831654799762294e-06,
+      "loss": 0.5693,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6387788619457058,
+      "grad_norm": 0.622677206993103,
+      "learning_rate": 3.4799524918063905e-06,
+      "loss": 0.548,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6389720799922712,
+      "grad_norm": 0.6524781584739685,
+      "learning_rate": 3.4767401950820003e-06,
+      "loss": 0.5814,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6391652980388368,
+      "grad_norm": 0.6539818048477173,
+      "learning_rate": 3.4735285912642822e-06,
+      "loss": 0.5682,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6393585160854024,
+      "grad_norm": 0.6521338820457458,
+      "learning_rate": 3.4703176818141436e-06,
+      "loss": 0.6227,
+      "step": 3309
+    },
+    {
+      "epoch": 0.639551734131968,
+      "grad_norm": 0.6621224880218506,
+      "learning_rate": 3.467107468192176e-06,
+      "loss": 0.6383,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6397449521785334,
+      "grad_norm": 0.6227132678031921,
+      "learning_rate": 3.463897951858652e-06,
+      "loss": 0.4813,
+      "step": 3311
+    },
+    {
+      "epoch": 0.639938170225099,
+      "grad_norm": 0.6808221936225891,
+      "learning_rate": 3.4606891342735323e-06,
+      "loss": 0.6189,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6401313882716646,
+      "grad_norm": 0.7317914962768555,
+      "learning_rate": 3.457481016896458e-06,
+      "loss": 0.6188,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6403246063182301,
+      "grad_norm": 0.7073050141334534,
+      "learning_rate": 3.4542736011867474e-06,
+      "loss": 0.6205,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6405178243647957,
+      "grad_norm": 0.6699276566505432,
+      "learning_rate": 3.4510668886034047e-06,
+      "loss": 0.5834,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6407110424113612,
+      "grad_norm": 0.6689836978912354,
+      "learning_rate": 3.447860880605113e-06,
+      "loss": 0.6398,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6409042604579268,
+      "grad_norm": 0.6237653493881226,
+      "learning_rate": 3.4446555786502336e-06,
+      "loss": 0.5898,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6410974785044923,
+      "grad_norm": 0.6792764067649841,
+      "learning_rate": 3.441450984196809e-06,
+      "loss": 0.5745,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6412906965510579,
+      "grad_norm": 0.6743780374526978,
+      "learning_rate": 3.4382470987025553e-06,
+      "loss": 0.5985,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6414839145976234,
+      "grad_norm": 0.7176421880722046,
+      "learning_rate": 3.43504392362487e-06,
+      "loss": 0.6403,
+      "step": 3320
+    },
+    {
+      "epoch": 0.641677132644189,
+      "grad_norm": 0.6396244168281555,
+      "learning_rate": 3.4318414604208296e-06,
+      "loss": 0.6304,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6418703506907545,
+      "grad_norm": 0.6680925488471985,
+      "learning_rate": 3.428639710547179e-06,
+      "loss": 0.563,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6420635687373201,
+      "grad_norm": 0.6626508831977844,
+      "learning_rate": 3.4254386754603464e-06,
+      "loss": 0.5873,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6422567867838856,
+      "grad_norm": 0.6570268273353577,
+      "learning_rate": 3.4222383566164314e-06,
+      "loss": 0.5435,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6424500048304511,
+      "grad_norm": 0.6346259713172913,
+      "learning_rate": 3.419038755471207e-06,
+      "loss": 0.592,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6426432228770167,
+      "grad_norm": 0.6961902379989624,
+      "learning_rate": 3.4158398734801234e-06,
+      "loss": 0.6231,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6428364409235823,
+      "grad_norm": 0.5682681798934937,
+      "learning_rate": 3.4126417120982968e-06,
+      "loss": 0.5094,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6430296589701479,
+      "grad_norm": 0.6552253365516663,
+      "learning_rate": 3.409444272780523e-06,
+      "loss": 0.6258,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6432228770167133,
+      "grad_norm": 0.6391724944114685,
+      "learning_rate": 3.4062475569812693e-06,
+      "loss": 0.6178,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6434160950632789,
+      "grad_norm": 0.6486668586730957,
+      "learning_rate": 3.4030515661546635e-06,
+      "loss": 0.5886,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6436093131098445,
+      "grad_norm": 0.6479765176773071,
+      "learning_rate": 3.399856301754517e-06,
+      "loss": 0.6002,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6438025311564101,
+      "grad_norm": 0.7845220565795898,
+      "learning_rate": 3.396661765234302e-06,
+      "loss": 0.7074,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6439957492029755,
+      "grad_norm": 0.6333609223365784,
+      "learning_rate": 3.393467958047163e-06,
+      "loss": 0.5613,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6441889672495411,
+      "grad_norm": 0.6251965165138245,
+      "learning_rate": 3.390274881645914e-06,
+      "loss": 0.5772,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6443821852961067,
+      "grad_norm": 0.6808993220329285,
+      "learning_rate": 3.3870825374830318e-06,
+      "loss": 0.6028,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6445754033426722,
+      "grad_norm": 0.7167524099349976,
+      "learning_rate": 3.3838909270106634e-06,
+      "loss": 0.6356,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6447686213892377,
+      "grad_norm": 0.6394233107566833,
+      "learning_rate": 3.380700051680625e-06,
+      "loss": 0.5467,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6449618394358033,
+      "grad_norm": 0.6992518305778503,
+      "learning_rate": 3.37750991294439e-06,
+      "loss": 0.5962,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6451550574823689,
+      "grad_norm": 0.697227954864502,
+      "learning_rate": 3.3743205122531065e-06,
+      "loss": 0.5782,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6453482755289344,
+      "grad_norm": 0.672706663608551,
+      "learning_rate": 3.3711318510575773e-06,
+      "loss": 0.5791,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6455414935755,
+      "grad_norm": 0.6067805290222168,
+      "learning_rate": 3.3679439308082777e-06,
+      "loss": 0.624,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6457347116220655,
+      "grad_norm": 0.6672163605690002,
+      "learning_rate": 3.364756752955343e-06,
+      "loss": 0.6197,
+      "step": 3342
+    },
+    {
+      "epoch": 0.645927929668631,
+      "grad_norm": 0.5996965765953064,
+      "learning_rate": 3.361570318948565e-06,
+      "loss": 0.5335,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6461211477151966,
+      "grad_norm": 0.6017184853553772,
+      "learning_rate": 3.3583846302374033e-06,
+      "loss": 0.582,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6463143657617622,
+      "grad_norm": 0.6459488868713379,
+      "learning_rate": 3.3551996882709813e-06,
+      "loss": 0.536,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6465075838083277,
+      "grad_norm": 0.584374189376831,
+      "learning_rate": 3.352015494498073e-06,
+      "loss": 0.5958,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6467008018548932,
+      "grad_norm": 0.6581851840019226,
+      "learning_rate": 3.3488320503671206e-06,
+      "loss": 0.5405,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6468940199014588,
+      "grad_norm": 0.6161103844642639,
+      "learning_rate": 3.3456493573262203e-06,
+      "loss": 0.5278,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6470872379480244,
+      "grad_norm": 0.617865800857544,
+      "learning_rate": 3.34246741682313e-06,
+      "loss": 0.5654,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6472804559945899,
+      "grad_norm": 0.6709005832672119,
+      "learning_rate": 3.3392862303052653e-06,
+      "loss": 0.599,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6474736740411554,
+      "grad_norm": 0.6249212026596069,
+      "learning_rate": 3.3361057992196933e-06,
+      "loss": 0.5262,
+      "step": 3351
+    },
+    {
+      "epoch": 0.647666892087721,
+      "grad_norm": 0.6177813410758972,
+      "learning_rate": 3.332926125013145e-06,
+      "loss": 0.6006,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6478601101342866,
+      "grad_norm": 0.6987733244895935,
+      "learning_rate": 3.3297472091320026e-06,
+      "loss": 0.6315,
+      "step": 3353
+    },
+    {
+      "epoch": 0.648053328180852,
+      "grad_norm": 0.6020432710647583,
+      "learning_rate": 3.326569053022304e-06,
+      "loss": 0.5435,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6482465462274176,
+      "grad_norm": 0.6674201488494873,
+      "learning_rate": 3.323391658129742e-06,
+      "loss": 0.6204,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6484397642739832,
+      "grad_norm": 0.6322950124740601,
+      "learning_rate": 3.320215025899663e-06,
+      "loss": 0.5814,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6486329823205488,
+      "grad_norm": 0.5877676010131836,
+      "learning_rate": 3.3170391577770666e-06,
+      "loss": 0.5856,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6488262003671142,
+      "grad_norm": 0.6406484246253967,
+      "learning_rate": 3.313864055206607e-06,
+      "loss": 0.5775,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6490194184136798,
+      "grad_norm": 0.6323333978652954,
+      "learning_rate": 3.3106897196325833e-06,
+      "loss": 0.6154,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6492126364602454,
+      "grad_norm": 0.7128861546516418,
+      "learning_rate": 3.307516152498954e-06,
+      "loss": 0.6377,
+      "step": 3360
+    },
+    {
+      "epoch": 0.649405854506811,
+      "grad_norm": 0.6893148422241211,
+      "learning_rate": 3.304343355249322e-06,
+      "loss": 0.5883,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6495990725533765,
+      "grad_norm": 0.6999078989028931,
+      "learning_rate": 3.301171329326943e-06,
+      "loss": 0.622,
+      "step": 3362
+    },
+    {
+      "epoch": 0.649792290599942,
+      "grad_norm": 0.6050905585289001,
+      "learning_rate": 3.2980000761747246e-06,
+      "loss": 0.6342,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6499855086465076,
+      "grad_norm": 0.6924591064453125,
+      "learning_rate": 3.294829597235213e-06,
+      "loss": 0.6254,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6501787266930731,
+      "grad_norm": 0.673215389251709,
+      "learning_rate": 3.2916598939506126e-06,
+      "loss": 0.5671,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6503719447396387,
+      "grad_norm": 0.6511792540550232,
+      "learning_rate": 3.288490967762773e-06,
+      "loss": 0.5686,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6505651627862042,
+      "grad_norm": 0.6443485021591187,
+      "learning_rate": 3.285322820113184e-06,
+      "loss": 0.6128,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6507583808327698,
+      "grad_norm": 0.6270350217819214,
+      "learning_rate": 3.282155452442988e-06,
+      "loss": 0.6034,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6509515988793353,
+      "grad_norm": 0.6430886387825012,
+      "learning_rate": 3.2789888661929693e-06,
+      "loss": 0.5902,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6511448169259009,
+      "grad_norm": 0.6649411916732788,
+      "learning_rate": 3.275823062803557e-06,
+      "loss": 0.5929,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6513380349724665,
+      "grad_norm": 0.6746648550033569,
+      "learning_rate": 3.2726580437148284e-06,
+      "loss": 0.5901,
+      "step": 3371
+    },
+    {
+      "epoch": 0.651531253019032,
+      "grad_norm": 0.6224015951156616,
+      "learning_rate": 3.269493810366496e-06,
+      "loss": 0.5614,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6517244710655975,
+      "grad_norm": 0.5421116948127747,
+      "learning_rate": 3.2663303641979207e-06,
+      "loss": 0.5158,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6519176891121631,
+      "grad_norm": 0.656063437461853,
+      "learning_rate": 3.263167706648106e-06,
+      "loss": 0.6248,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6521109071587287,
+      "grad_norm": 0.5643991827964783,
+      "learning_rate": 3.260005839155691e-06,
+      "loss": 0.5216,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6523041252052941,
+      "grad_norm": 0.6067699193954468,
+      "learning_rate": 3.2568447631589616e-06,
+      "loss": 0.5703,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6524973432518597,
+      "grad_norm": 0.7164930701255798,
+      "learning_rate": 3.253684480095839e-06,
+      "loss": 0.6281,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6526905612984253,
+      "grad_norm": 0.6151798367500305,
+      "learning_rate": 3.250524991403887e-06,
+      "loss": 0.5451,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6528837793449909,
+      "grad_norm": 0.7195934057235718,
+      "learning_rate": 3.247366298520309e-06,
+      "loss": 0.5961,
+      "step": 3379
+    },
+    {
+      "epoch": 0.6530769973915563,
+      "grad_norm": 0.6130334734916687,
+      "learning_rate": 3.244208402881939e-06,
+      "loss": 0.5935,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6532702154381219,
+      "grad_norm": 0.6749882698059082,
+      "learning_rate": 3.241051305925259e-06,
+      "loss": 0.5864,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6534634334846875,
+      "grad_norm": 0.5927901268005371,
+      "learning_rate": 3.237895009086377e-06,
+      "loss": 0.5859,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6536566515312531,
+      "grad_norm": 0.6817566156387329,
+      "learning_rate": 3.234739513801045e-06,
+      "loss": 0.6087,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6538498695778185,
+      "grad_norm": 0.665349543094635,
+      "learning_rate": 3.2315848215046487e-06,
+      "loss": 0.6072,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6540430876243841,
+      "grad_norm": 0.6708278059959412,
+      "learning_rate": 3.2284309336322052e-06,
+      "loss": 0.6059,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6542363056709497,
+      "grad_norm": 0.7216925024986267,
+      "learning_rate": 3.2252778516183693e-06,
+      "loss": 0.6564,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6544295237175152,
+      "grad_norm": 0.677300751209259,
+      "learning_rate": 3.2221255768974285e-06,
+      "loss": 0.5992,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6546227417640808,
+      "grad_norm": 0.7097252607345581,
+      "learning_rate": 3.2189741109033e-06,
+      "loss": 0.6382,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6548159598106463,
+      "grad_norm": 0.7245694994926453,
+      "learning_rate": 3.21582345506954e-06,
+      "loss": 0.7068,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6550091778572119,
+      "grad_norm": 0.6451432704925537,
+      "learning_rate": 3.2126736108293257e-06,
+      "loss": 0.642,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6552023959037774,
+      "grad_norm": 0.6093419194221497,
+      "learning_rate": 3.209524579615475e-06,
+      "loss": 0.5976,
+      "step": 3391
+    },
+    {
+      "epoch": 0.655395613950343,
+      "grad_norm": 0.657567024230957,
+      "learning_rate": 3.206376362860432e-06,
+      "loss": 0.575,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6555888319969085,
+      "grad_norm": 0.6776841282844543,
+      "learning_rate": 3.2032289619962696e-06,
+      "loss": 0.6274,
+      "step": 3393
+    },
+    {
+      "epoch": 0.655782050043474,
+      "grad_norm": 0.6294445991516113,
+      "learning_rate": 3.20008237845469e-06,
+      "loss": 0.6628,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6559752680900396,
+      "grad_norm": 0.690061628818512,
+      "learning_rate": 3.196936613667028e-06,
+      "loss": 0.6198,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6561684861366052,
+      "grad_norm": 0.6565684080123901,
+      "learning_rate": 3.1937916690642356e-06,
+      "loss": 0.5649,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6563617041831707,
+      "grad_norm": 0.6539074182510376,
+      "learning_rate": 3.1906475460769037e-06,
+      "loss": 0.5889,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6565549222297362,
+      "grad_norm": 0.6750060319900513,
+      "learning_rate": 3.1875042461352397e-06,
+      "loss": 0.6354,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6567481402763018,
+      "grad_norm": 0.6550947427749634,
+      "learning_rate": 3.1843617706690823e-06,
+      "loss": 0.5717,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6569413583228674,
+      "grad_norm": 0.6761435866355896,
+      "learning_rate": 3.1812201211078952e-06,
+      "loss": 0.6246,
+      "step": 3400
+    },
+    {
+      "epoch": 0.657134576369433,
+      "grad_norm": 0.6549978256225586,
+      "learning_rate": 3.178079298880762e-06,
+      "loss": 0.6013,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6573277944159984,
+      "grad_norm": 0.6521122455596924,
+      "learning_rate": 3.1749393054163968e-06,
+      "loss": 0.5917,
+      "step": 3402
+    },
+    {
+      "epoch": 0.657521012462564,
+      "grad_norm": 0.6262401342391968,
+      "learning_rate": 3.1718001421431284e-06,
+      "loss": 0.5929,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6577142305091296,
+      "grad_norm": 0.6625175476074219,
+      "learning_rate": 3.1686618104889133e-06,
+      "loss": 0.661,
+      "step": 3404
+    },
+    {
+      "epoch": 0.657907448555695,
+      "grad_norm": 0.6762275099754333,
+      "learning_rate": 3.1655243118813316e-06,
+      "loss": 0.6285,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6581006666022606,
+      "grad_norm": 0.656437337398529,
+      "learning_rate": 3.162387647747579e-06,
+      "loss": 0.6297,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6582938846488262,
+      "grad_norm": 0.7066957354545593,
+      "learning_rate": 3.159251819514473e-06,
+      "loss": 0.5797,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6584871026953918,
+      "grad_norm": 0.6464317440986633,
+      "learning_rate": 3.1561168286084553e-06,
+      "loss": 0.6321,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6586803207419573,
+      "grad_norm": 0.6305776834487915,
+      "learning_rate": 3.152982676455581e-06,
+      "loss": 0.5582,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6588735387885228,
+      "grad_norm": 0.6569595336914062,
+      "learning_rate": 3.1498493644815287e-06,
+      "loss": 0.6229,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6590667568350884,
+      "grad_norm": 0.6136631965637207,
+      "learning_rate": 3.1467168941115877e-06,
+      "loss": 0.6252,
+      "step": 3411
+    },
+    {
+      "epoch": 0.659259974881654,
+      "grad_norm": 0.6575597524642944,
+      "learning_rate": 3.1435852667706725e-06,
+      "loss": 0.6172,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6594531929282195,
+      "grad_norm": 0.6091388463973999,
+      "learning_rate": 3.14045448388331e-06,
+      "loss": 0.5756,
+      "step": 3413
+    },
+    {
+      "epoch": 0.659646410974785,
+      "grad_norm": 0.5932101607322693,
+      "learning_rate": 3.1373245468736424e-06,
+      "loss": 0.5673,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6598396290213506,
+      "grad_norm": 0.6358563303947449,
+      "learning_rate": 3.1341954571654297e-06,
+      "loss": 0.6236,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6600328470679161,
+      "grad_norm": 0.6411874890327454,
+      "learning_rate": 3.1310672161820465e-06,
+      "loss": 0.5924,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6602260651144817,
+      "grad_norm": 0.6185294389724731,
+      "learning_rate": 3.1279398253464755e-06,
+      "loss": 0.5862,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6604192831610473,
+      "grad_norm": 0.7433662414550781,
+      "learning_rate": 3.1248132860813218e-06,
+      "loss": 0.5979,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6606125012076128,
+      "grad_norm": 0.6251358985900879,
+      "learning_rate": 3.1216875998087938e-06,
+      "loss": 0.6142,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6608057192541783,
+      "grad_norm": 0.6764047145843506,
+      "learning_rate": 3.118562767950719e-06,
+      "loss": 0.5911,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6609989373007439,
+      "grad_norm": 0.6275711059570312,
+      "learning_rate": 3.115438791928535e-06,
+      "loss": 0.5832,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6611921553473095,
+      "grad_norm": 0.6400980949401855,
+      "learning_rate": 3.1123156731632866e-06,
+      "loss": 0.6306,
+      "step": 3422
+    },
+    {
+      "epoch": 0.661385373393875,
+      "grad_norm": 0.5926651954650879,
+      "learning_rate": 3.1091934130756307e-06,
+      "loss": 0.566,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6615785914404405,
+      "grad_norm": 0.5518404245376587,
+      "learning_rate": 3.106072013085838e-06,
+      "loss": 0.5312,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6617718094870061,
+      "grad_norm": 0.6512795090675354,
+      "learning_rate": 3.102951474613777e-06,
+      "loss": 0.5685,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6619650275335717,
+      "grad_norm": 0.679987370967865,
+      "learning_rate": 3.0998317990789378e-06,
+      "loss": 0.624,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6621582455801371,
+      "grad_norm": 0.7382001280784607,
+      "learning_rate": 3.0967129879004055e-06,
+      "loss": 0.6039,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6623514636267027,
+      "grad_norm": 0.6749094128608704,
+      "learning_rate": 3.093595042496881e-06,
+      "loss": 0.6251,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6625446816732683,
+      "grad_norm": 0.6222795844078064,
+      "learning_rate": 3.090477964286667e-06,
+      "loss": 0.6173,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6627378997198339,
+      "grad_norm": 0.6688793301582336,
+      "learning_rate": 3.087361754687673e-06,
+      "loss": 0.5754,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6629311177663993,
+      "grad_norm": 0.6260875463485718,
+      "learning_rate": 3.0842464151174144e-06,
+      "loss": 0.6019,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6631243358129649,
+      "grad_norm": 0.6877782344818115,
+      "learning_rate": 3.0811319469930077e-06,
+      "loss": 0.6376,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6633175538595305,
+      "grad_norm": 0.6777331233024597,
+      "learning_rate": 3.0780183517311755e-06,
+      "loss": 0.6031,
+      "step": 3433
+    },
+    {
+      "epoch": 0.663510771906096,
+      "grad_norm": 0.6745155453681946,
+      "learning_rate": 3.074905630748245e-06,
+      "loss": 0.6717,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6637039899526616,
+      "grad_norm": 0.6700016856193542,
+      "learning_rate": 3.07179378546014e-06,
+      "loss": 0.6051,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6638972079992271,
+      "grad_norm": 0.7014890313148499,
+      "learning_rate": 3.06868281728239e-06,
+      "loss": 0.6976,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6640904260457927,
+      "grad_norm": 0.6400468349456787,
+      "learning_rate": 3.0655727276301283e-06,
+      "loss": 0.6493,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6642836440923582,
+      "grad_norm": 0.6822721362113953,
+      "learning_rate": 3.0624635179180817e-06,
+      "loss": 0.631,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6644768621389238,
+      "grad_norm": 0.6331565380096436,
+      "learning_rate": 3.059355189560583e-06,
+      "loss": 0.6097,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6646700801854893,
+      "grad_norm": 0.6206365823745728,
+      "learning_rate": 3.0562477439715588e-06,
+      "loss": 0.5174,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6648632982320549,
+      "grad_norm": 0.7027522921562195,
+      "learning_rate": 3.053141182564537e-06,
+      "loss": 0.6487,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6650565162786204,
+      "grad_norm": 0.674957811832428,
+      "learning_rate": 3.0500355067526462e-06,
+      "loss": 0.5892,
+      "step": 3442
+    },
+    {
+      "epoch": 0.665249734325186,
+      "grad_norm": 0.6515098810195923,
+      "learning_rate": 3.046930717948604e-06,
+      "loss": 0.5355,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6654429523717516,
+      "grad_norm": 0.7066481113433838,
+      "learning_rate": 3.0438268175647327e-06,
+      "loss": 0.6147,
+      "step": 3444
+    },
+    {
+      "epoch": 0.665636170418317,
+      "grad_norm": 0.6602864861488342,
+      "learning_rate": 3.0407238070129473e-06,
+      "loss": 0.5969,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6658293884648826,
+      "grad_norm": 0.6818235516548157,
+      "learning_rate": 3.037621687704757e-06,
+      "loss": 0.6463,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6660226065114482,
+      "grad_norm": 0.6196058988571167,
+      "learning_rate": 3.034520461051268e-06,
+      "loss": 0.6255,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6662158245580138,
+      "grad_norm": 0.6728675961494446,
+      "learning_rate": 3.0314201284631763e-06,
+      "loss": 0.5795,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6664090426045792,
+      "grad_norm": 0.5710001587867737,
+      "learning_rate": 3.0283206913507755e-06,
+      "loss": 0.6019,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6666022606511448,
+      "grad_norm": 0.594716489315033,
+      "learning_rate": 3.025222151123952e-06,
+      "loss": 0.5945,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6667954786977104,
+      "grad_norm": 0.6558206081390381,
+      "learning_rate": 3.022124509192179e-06,
+      "loss": 0.5986,
+      "step": 3451
+    },
+    {
+      "epoch": 0.666988696744276,
+      "grad_norm": 0.644540011882782,
+      "learning_rate": 3.0190277669645274e-06,
+      "loss": 0.5935,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6671819147908414,
+      "grad_norm": 0.6220780611038208,
+      "learning_rate": 3.0159319258496535e-06,
+      "loss": 0.6041,
+      "step": 3453
+    },
+    {
+      "epoch": 0.667375132837407,
+      "grad_norm": 0.6915618777275085,
+      "learning_rate": 3.012836987255807e-06,
+      "loss": 0.5789,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6675683508839726,
+      "grad_norm": 0.596890926361084,
+      "learning_rate": 3.0097429525908306e-06,
+      "loss": 0.5095,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6677615689305381,
+      "grad_norm": 0.6206494569778442,
+      "learning_rate": 3.0066498232621446e-06,
+      "loss": 0.5109,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6679547869771036,
+      "grad_norm": 0.627731204032898,
+      "learning_rate": 3.0035576006767673e-06,
+      "loss": 0.56,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6681480050236692,
+      "grad_norm": 0.6419090032577515,
+      "learning_rate": 3.000466286241304e-06,
+      "loss": 0.5957,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6683412230702348,
+      "grad_norm": 0.5856388807296753,
+      "learning_rate": 2.99737588136194e-06,
+      "loss": 0.6183,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6685344411168003,
+      "grad_norm": 0.6836070418357849,
+      "learning_rate": 2.9942863874444565e-06,
+      "loss": 0.6426,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6687276591633659,
+      "grad_norm": 0.668258786201477,
+      "learning_rate": 2.9911978058942104e-06,
+      "loss": 0.5687,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6689208772099314,
+      "grad_norm": 0.6604463458061218,
+      "learning_rate": 2.98811013811615e-06,
+      "loss": 0.5567,
+      "step": 3462
+    },
+    {
+      "epoch": 0.669114095256497,
+      "grad_norm": 0.6705439686775208,
+      "learning_rate": 2.985023385514809e-06,
+      "loss": 0.6742,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6693073133030625,
+      "grad_norm": 0.6783331036567688,
+      "learning_rate": 2.9819375494942974e-06,
+      "loss": 0.6123,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6695005313496281,
+      "grad_norm": 0.7344334125518799,
+      "learning_rate": 2.9788526314583153e-06,
+      "loss": 0.6926,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6696937493961936,
+      "grad_norm": 0.7066696882247925,
+      "learning_rate": 2.9757686328101443e-06,
+      "loss": 0.6648,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6698869674427591,
+      "grad_norm": 0.7319726347923279,
+      "learning_rate": 2.9726855549526434e-06,
+      "loss": 0.6382,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6700801854893247,
+      "grad_norm": 0.6848384737968445,
+      "learning_rate": 2.969603399288259e-06,
+      "loss": 0.6684,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6702734035358903,
+      "grad_norm": 0.6153566837310791,
+      "learning_rate": 2.9665221672190113e-06,
+      "loss": 0.6599,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6704666215824558,
+      "grad_norm": 0.6265047788619995,
+      "learning_rate": 2.9634418601465055e-06,
+      "loss": 0.6157,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6706598396290213,
+      "grad_norm": 0.6532571315765381,
+      "learning_rate": 2.960362479471926e-06,
+      "loss": 0.6488,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6708530576755869,
+      "grad_norm": 0.6168240904808044,
+      "learning_rate": 2.95728402659603e-06,
+      "loss": 0.5796,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6710462757221525,
+      "grad_norm": 0.652706503868103,
+      "learning_rate": 2.95420650291916e-06,
+      "loss": 0.6069,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6712394937687179,
+      "grad_norm": 0.6177634000778198,
+      "learning_rate": 2.951129909841232e-06,
+      "loss": 0.6483,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6714327118152835,
+      "grad_norm": 0.6367934942245483,
+      "learning_rate": 2.9480542487617383e-06,
+      "loss": 0.5724,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6716259298618491,
+      "grad_norm": 0.6717763543128967,
+      "learning_rate": 2.944979521079751e-06,
+      "loss": 0.6108,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6718191479084147,
+      "grad_norm": 0.6688340306282043,
+      "learning_rate": 2.9419057281939106e-06,
+      "loss": 0.6535,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6720123659549802,
+      "grad_norm": 0.6718976497650146,
+      "learning_rate": 2.938832871502439e-06,
+      "loss": 0.589,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6722055840015457,
+      "grad_norm": 0.6231658458709717,
+      "learning_rate": 2.9357609524031315e-06,
+      "loss": 0.5948,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6723988020481113,
+      "grad_norm": 0.7021034359931946,
+      "learning_rate": 2.9326899722933516e-06,
+      "loss": 0.6552,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6725920200946769,
+      "grad_norm": 0.654029130935669,
+      "learning_rate": 2.9296199325700423e-06,
+      "loss": 0.5902,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6727852381412424,
+      "grad_norm": 0.6762625575065613,
+      "learning_rate": 2.9265508346297133e-06,
+      "loss": 0.5504,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6729784561878079,
+      "grad_norm": 0.6844524145126343,
+      "learning_rate": 2.92348267986845e-06,
+      "loss": 0.609,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6731716742343735,
+      "grad_norm": 0.6178730130195618,
+      "learning_rate": 2.92041546968191e-06,
+      "loss": 0.6475,
+      "step": 3484
+    },
+    {
+      "epoch": 0.673364892280939,
+      "grad_norm": 0.6185630559921265,
+      "learning_rate": 2.917349205465313e-06,
+      "loss": 0.5629,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6735581103275046,
+      "grad_norm": 0.6928721070289612,
+      "learning_rate": 2.914283888613455e-06,
+      "loss": 0.5963,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6737513283740701,
+      "grad_norm": 0.672309398651123,
+      "learning_rate": 2.911219520520706e-06,
+      "loss": 0.6646,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6739445464206357,
+      "grad_norm": 0.626837432384491,
+      "learning_rate": 2.908156102580989e-06,
+      "loss": 0.5786,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6741377644672012,
+      "grad_norm": 0.6432395577430725,
+      "learning_rate": 2.9050936361878136e-06,
+      "loss": 0.5297,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6743309825137668,
+      "grad_norm": 0.6385375261306763,
+      "learning_rate": 2.902032122734238e-06,
+      "loss": 0.6224,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6745242005603324,
+      "grad_norm": 0.6779659986495972,
+      "learning_rate": 2.8989715636129024e-06,
+      "loss": 0.5564,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6747174186068979,
+      "grad_norm": 0.7246143817901611,
+      "learning_rate": 2.8959119602160043e-06,
+      "loss": 0.6235,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6749106366534634,
+      "grad_norm": 0.6050474047660828,
+      "learning_rate": 2.8928533139353067e-06,
+      "loss": 0.5222,
+      "step": 3493
+    },
+    {
+      "epoch": 0.675103854700029,
+      "grad_norm": 0.6382969617843628,
+      "learning_rate": 2.889795626162143e-06,
+      "loss": 0.6027,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6752970727465946,
+      "grad_norm": 0.7218477129936218,
+      "learning_rate": 2.8867388982874056e-06,
+      "loss": 0.6376,
+      "step": 3495
+    },
+    {
+      "epoch": 0.67549029079316,
+      "grad_norm": 0.6787936091423035,
+      "learning_rate": 2.8836831317015503e-06,
+      "loss": 0.6541,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6756835088397256,
+      "grad_norm": 0.6661674380302429,
+      "learning_rate": 2.8806283277945973e-06,
+      "loss": 0.6021,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6758767268862912,
+      "grad_norm": 0.6974189281463623,
+      "learning_rate": 2.8775744879561283e-06,
+      "loss": 0.6286,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6760699449328568,
+      "grad_norm": 0.6904579401016235,
+      "learning_rate": 2.874521613575285e-06,
+      "loss": 0.6274,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6762631629794222,
+      "grad_norm": 0.6546990871429443,
+      "learning_rate": 2.8714697060407744e-06,
+      "loss": 0.602,
+      "step": 3500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5176,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2688046994817024.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}