diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8029 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 7.0,
+  "eval_steps": 500,
+  "global_step": 1141,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.006134969325153374,
+      "grad_norm": 0.14804252982139587,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.5673,
+      "step": 1
+    },
+    {
+      "epoch": 0.012269938650306749,
+      "grad_norm": 0.1421855241060257,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.5703,
+      "step": 2
+    },
+    {
+      "epoch": 0.018404907975460124,
+      "grad_norm": 0.14651049673557281,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5759,
+      "step": 3
+    },
+    {
+      "epoch": 0.024539877300613498,
+      "grad_norm": 0.14899852871894836,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6024,
+      "step": 4
+    },
+    {
+      "epoch": 0.03067484662576687,
+      "grad_norm": 0.1509060114622116,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.5794,
+      "step": 5
+    },
+    {
+      "epoch": 0.03680981595092025,
+      "grad_norm": 0.14855755865573883,
+      "learning_rate": 5e-06,
+      "loss": 0.5726,
+      "step": 6
+    },
+    {
+      "epoch": 0.04294478527607362,
+      "grad_norm": 0.161569744348526,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.5992,
+      "step": 7
+    },
+    {
+      "epoch": 0.049079754601226995,
+      "grad_norm": 0.1378004103899002,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5801,
+      "step": 8
+    },
+    {
+      "epoch": 0.05521472392638037,
+      "grad_norm": 0.1505780816078186,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5923,
+      "step": 9
+    },
+    {
+      "epoch": 0.06134969325153374,
+      "grad_norm": 0.15210111439228058,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.5794,
+      "step": 10
+    },
+    {
+      "epoch": 0.06748466257668712,
+      "grad_norm": 0.16127604246139526,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.5925,
+      "step": 11
+    },
+    {
+      "epoch": 0.0736196319018405,
+      "grad_norm": 0.15838903188705444,
+      "learning_rate": 1e-05,
+      "loss": 0.5879,
+      "step": 12
+    },
+    {
+      "epoch": 0.07975460122699386,
+      "grad_norm": 0.15082281827926636,
+      "learning_rate": 9.999980642396502e-06,
+      "loss": 0.6217,
+      "step": 13
+    },
+    {
+      "epoch": 0.08588957055214724,
+      "grad_norm": 0.16154126822948456,
+      "learning_rate": 9.999922569735891e-06,
+      "loss": 0.5725,
+      "step": 14
+    },
+    {
+      "epoch": 0.09202453987730061,
+      "grad_norm": 0.16344521939754486,
+      "learning_rate": 9.999825782467827e-06,
+      "loss": 0.6094,
+      "step": 15
+    },
+    {
+      "epoch": 0.09815950920245399,
+      "grad_norm": 0.16911642253398895,
+      "learning_rate": 9.99969028134174e-06,
+      "loss": 0.5627,
+      "step": 16
+    },
+    {
+      "epoch": 0.10429447852760736,
+      "grad_norm": 0.16826483607292175,
+      "learning_rate": 9.999516067406818e-06,
+      "loss": 0.5996,
+      "step": 17
+    },
+    {
+      "epoch": 0.11042944785276074,
+      "grad_norm": 0.16453488171100616,
+      "learning_rate": 9.999303142012008e-06,
+      "loss": 0.5984,
+      "step": 18
+    },
+    {
+      "epoch": 0.1165644171779141,
+      "grad_norm": 0.16882814466953278,
+      "learning_rate": 9.999051506806e-06,
+      "loss": 0.5684,
+      "step": 19
+    },
+    {
+      "epoch": 0.12269938650306748,
+      "grad_norm": 0.15942800045013428,
+      "learning_rate": 9.998761163737217e-06,
+      "loss": 0.5879,
+      "step": 20
+    },
+    {
+      "epoch": 0.12883435582822086,
+      "grad_norm": 0.17792896926403046,
+      "learning_rate": 9.998432115053796e-06,
+      "loss": 0.5572,
+      "step": 21
+    },
+    {
+      "epoch": 0.13496932515337423,
+      "grad_norm": 0.1712748408317566,
+      "learning_rate": 9.998064363303573e-06,
+      "loss": 0.5693,
+      "step": 22
+    },
+    {
+      "epoch": 0.1411042944785276,
+      "grad_norm": 0.1689537614583969,
+      "learning_rate": 9.997657911334068e-06,
+      "loss": 0.5584,
+      "step": 23
+    },
+    {
+      "epoch": 0.147239263803681,
+      "grad_norm": 0.17206093668937683,
+      "learning_rate": 9.997212762292453e-06,
+      "loss": 0.5406,
+      "step": 24
+    },
+    {
+      "epoch": 0.15337423312883436,
+      "grad_norm": 0.1795872151851654,
+      "learning_rate": 9.996728919625538e-06,
+      "loss": 0.5767,
+      "step": 25
+    },
+    {
+      "epoch": 0.15950920245398773,
+      "grad_norm": 0.16666671633720398,
+      "learning_rate": 9.996206387079736e-06,
+      "loss": 0.5792,
+      "step": 26
+    },
+    {
+      "epoch": 0.1656441717791411,
+      "grad_norm": 0.15964794158935547,
+      "learning_rate": 9.995645168701038e-06,
+      "loss": 0.5771,
+      "step": 27
+    },
+    {
+      "epoch": 0.17177914110429449,
+      "grad_norm": 0.1661137342453003,
+      "learning_rate": 9.995045268834979e-06,
+      "loss": 0.5695,
+      "step": 28
+    },
+    {
+      "epoch": 0.17791411042944785,
+      "grad_norm": 0.16640233993530273,
+      "learning_rate": 9.99440669212661e-06,
+      "loss": 0.6123,
+      "step": 29
+    },
+    {
+      "epoch": 0.18404907975460122,
+      "grad_norm": 0.14831523597240448,
+      "learning_rate": 9.99372944352046e-06,
+      "loss": 0.5309,
+      "step": 30
+    },
+    {
+      "epoch": 0.1901840490797546,
+      "grad_norm": 0.13072659075260162,
+      "learning_rate": 9.993013528260486e-06,
+      "loss": 0.5782,
+      "step": 31
+    },
+    {
+      "epoch": 0.19631901840490798,
+      "grad_norm": 0.12631100416183472,
+      "learning_rate": 9.992258951890057e-06,
+      "loss": 0.5361,
+      "step": 32
+    },
+    {
+      "epoch": 0.20245398773006135,
+      "grad_norm": 0.1304839849472046,
+      "learning_rate": 9.991465720251885e-06,
+      "loss": 0.5673,
+      "step": 33
+    },
+    {
+      "epoch": 0.2085889570552147,
+      "grad_norm": 0.12783069908618927,
+      "learning_rate": 9.990633839487997e-06,
+      "loss": 0.5662,
+      "step": 34
+    },
+    {
+      "epoch": 0.2147239263803681,
+      "grad_norm": 0.12577638030052185,
+      "learning_rate": 9.989763316039678e-06,
+      "loss": 0.5328,
+      "step": 35
+    },
+    {
+      "epoch": 0.22085889570552147,
+      "grad_norm": 0.12679477035999298,
+      "learning_rate": 9.988854156647428e-06,
+      "loss": 0.5758,
+      "step": 36
+    },
+    {
+      "epoch": 0.22699386503067484,
+      "grad_norm": 0.1201774999499321,
+      "learning_rate": 9.987906368350908e-06,
+      "loss": 0.5598,
+      "step": 37
+    },
+    {
+      "epoch": 0.2331288343558282,
+      "grad_norm": 0.1210893765091896,
+      "learning_rate": 9.98691995848888e-06,
+      "loss": 0.5902,
+      "step": 38
+    },
+    {
+      "epoch": 0.2392638036809816,
+      "grad_norm": 0.10759458690881729,
+      "learning_rate": 9.985894934699154e-06,
+      "loss": 0.524,
+      "step": 39
+    },
+    {
+      "epoch": 0.24539877300613497,
+      "grad_norm": 0.10544677078723907,
+      "learning_rate": 9.984831304918537e-06,
+      "loss": 0.5535,
+      "step": 40
+    },
+    {
+      "epoch": 0.25153374233128833,
+      "grad_norm": 0.1004803329706192,
+      "learning_rate": 9.983729077382755e-06,
+      "loss": 0.5204,
+      "step": 41
+    },
+    {
+      "epoch": 0.25766871165644173,
+      "grad_norm": 0.10377446562051773,
+      "learning_rate": 9.982588260626402e-06,
+      "loss": 0.5319,
+      "step": 42
+    },
+    {
+      "epoch": 0.26380368098159507,
+      "grad_norm": 0.09945333003997803,
+      "learning_rate": 9.981408863482872e-06,
+      "loss": 0.5473,
+      "step": 43
+    },
+    {
+      "epoch": 0.26993865030674846,
+      "grad_norm": 0.09462588280439377,
+      "learning_rate": 9.98019089508428e-06,
+      "loss": 0.5455,
+      "step": 44
+    },
+    {
+      "epoch": 0.27607361963190186,
+      "grad_norm": 0.09673507511615753,
+      "learning_rate": 9.97893436486141e-06,
+      "loss": 0.5175,
+      "step": 45
+    },
+    {
+      "epoch": 0.2822085889570552,
+      "grad_norm": 0.09385982155799866,
+      "learning_rate": 9.977639282543627e-06,
+      "loss": 0.5519,
+      "step": 46
+    },
+    {
+      "epoch": 0.2883435582822086,
+      "grad_norm": 0.10252804309129715,
+      "learning_rate": 9.976305658158806e-06,
+      "loss": 0.5197,
+      "step": 47
+    },
+    {
+      "epoch": 0.294478527607362,
+      "grad_norm": 0.09260332584381104,
+      "learning_rate": 9.97493350203326e-06,
+      "loss": 0.5454,
+      "step": 48
+    },
+    {
+      "epoch": 0.3006134969325153,
+      "grad_norm": 0.09458579868078232,
+      "learning_rate": 9.973522824791643e-06,
+      "loss": 0.5376,
+      "step": 49
+    },
+    {
+      "epoch": 0.3067484662576687,
+      "grad_norm": 0.10349428653717041,
+      "learning_rate": 9.972073637356894e-06,
+      "loss": 0.5747,
+      "step": 50
+    },
+    {
+      "epoch": 0.3128834355828221,
+      "grad_norm": 0.09267648309469223,
+      "learning_rate": 9.970585950950129e-06,
+      "loss": 0.525,
+      "step": 51
+    },
+    {
+      "epoch": 0.31901840490797545,
+      "grad_norm": 0.08669942617416382,
+      "learning_rate": 9.969059777090564e-06,
+      "loss": 0.5121,
+      "step": 52
+    },
+    {
+      "epoch": 0.32515337423312884,
+      "grad_norm": 0.09225864708423615,
+      "learning_rate": 9.967495127595427e-06,
+      "loss": 0.523,
+      "step": 53
+    },
+    {
+      "epoch": 0.3312883435582822,
+      "grad_norm": 0.09258433431386948,
+      "learning_rate": 9.965892014579867e-06,
+      "loss": 0.5275,
+      "step": 54
+    },
+    {
+      "epoch": 0.3374233128834356,
+      "grad_norm": 0.08944110572338104,
+      "learning_rate": 9.96425045045685e-06,
+      "loss": 0.5338,
+      "step": 55
+    },
+    {
+      "epoch": 0.34355828220858897,
+      "grad_norm": 0.09073130786418915,
+      "learning_rate": 9.962570447937077e-06,
+      "loss": 0.5379,
+      "step": 56
+    },
+    {
+      "epoch": 0.3496932515337423,
+      "grad_norm": 0.10125018656253815,
+      "learning_rate": 9.960852020028877e-06,
+      "loss": 0.5216,
+      "step": 57
+    },
+    {
+      "epoch": 0.3558282208588957,
+      "grad_norm": 0.09612809121608734,
+      "learning_rate": 9.95909518003811e-06,
+      "loss": 0.5076,
+      "step": 58
+    },
+    {
+      "epoch": 0.3619631901840491,
+      "grad_norm": 0.09233148396015167,
+      "learning_rate": 9.957299941568058e-06,
+      "loss": 0.5443,
+      "step": 59
+    },
+    {
+      "epoch": 0.36809815950920244,
+      "grad_norm": 0.14108285307884216,
+      "learning_rate": 9.955466318519327e-06,
+      "loss": 0.5588,
+      "step": 60
+    },
+    {
+      "epoch": 0.37423312883435583,
+      "grad_norm": 0.09588061273097992,
+      "learning_rate": 9.953594325089738e-06,
+      "loss": 0.506,
+      "step": 61
+    },
+    {
+      "epoch": 0.3803680981595092,
+      "grad_norm": 0.08795749396085739,
+      "learning_rate": 9.951683975774213e-06,
+      "loss": 0.5158,
+      "step": 62
+    },
+    {
+      "epoch": 0.38650306748466257,
+      "grad_norm": 0.0903969258069992,
+      "learning_rate": 9.949735285364666e-06,
+      "loss": 0.5337,
+      "step": 63
+    },
+    {
+      "epoch": 0.39263803680981596,
+      "grad_norm": 0.09337311238050461,
+      "learning_rate": 9.947748268949885e-06,
+      "loss": 0.5313,
+      "step": 64
+    },
+    {
+      "epoch": 0.3987730061349693,
+      "grad_norm": 0.08544723689556122,
+      "learning_rate": 9.945722941915424e-06,
+      "loss": 0.5398,
+      "step": 65
+    },
+    {
+      "epoch": 0.4049079754601227,
+      "grad_norm": 0.08584097027778625,
+      "learning_rate": 9.943659319943472e-06,
+      "loss": 0.5116,
+      "step": 66
+    },
+    {
+      "epoch": 0.4110429447852761,
+      "grad_norm": 0.08828586339950562,
+      "learning_rate": 9.941557419012742e-06,
+      "loss": 0.5004,
+      "step": 67
+    },
+    {
+      "epoch": 0.4171779141104294,
+      "grad_norm": 0.0884728655219078,
+      "learning_rate": 9.939417255398336e-06,
+      "loss": 0.5463,
+      "step": 68
+    },
+    {
+      "epoch": 0.4233128834355828,
+      "grad_norm": 0.08237382769584656,
+      "learning_rate": 9.93723884567163e-06,
+      "loss": 0.4999,
+      "step": 69
+    },
+    {
+      "epoch": 0.4294478527607362,
+      "grad_norm": 0.07927916198968887,
+      "learning_rate": 9.935022206700145e-06,
+      "loss": 0.5056,
+      "step": 70
+    },
+    {
+      "epoch": 0.43558282208588955,
+      "grad_norm": 0.08695581555366516,
+      "learning_rate": 9.932767355647404e-06,
+      "loss": 0.5327,
+      "step": 71
+    },
+    {
+      "epoch": 0.44171779141104295,
+      "grad_norm": 0.0842549055814743,
+      "learning_rate": 9.930474309972813e-06,
+      "loss": 0.5371,
+      "step": 72
+    },
+    {
+      "epoch": 0.44785276073619634,
+      "grad_norm": 0.09977614879608154,
+      "learning_rate": 9.92814308743152e-06,
+      "loss": 0.5404,
+      "step": 73
+    },
+    {
+      "epoch": 0.4539877300613497,
+      "grad_norm": 0.08078821003437042,
+      "learning_rate": 9.925773706074278e-06,
+      "loss": 0.5223,
+      "step": 74
+    },
+    {
+      "epoch": 0.4601226993865031,
+      "grad_norm": 0.08402088284492493,
+      "learning_rate": 9.923366184247306e-06,
+      "loss": 0.4808,
+      "step": 75
+    },
+    {
+      "epoch": 0.4662576687116564,
+      "grad_norm": 0.08832000941038132,
+      "learning_rate": 9.920920540592141e-06,
+      "loss": 0.5187,
+      "step": 76
+    },
+    {
+      "epoch": 0.4723926380368098,
+      "grad_norm": 0.0865439772605896,
+      "learning_rate": 9.918436794045507e-06,
+      "loss": 0.5013,
+      "step": 77
+    },
+    {
+      "epoch": 0.4785276073619632,
+      "grad_norm": 0.0843573585152626,
+      "learning_rate": 9.915914963839154e-06,
+      "loss": 0.5203,
+      "step": 78
+    },
+    {
+      "epoch": 0.48466257668711654,
+      "grad_norm": 0.08602918684482574,
+      "learning_rate": 9.91335506949972e-06,
+      "loss": 0.509,
+      "step": 79
+    },
+    {
+      "epoch": 0.49079754601226994,
+      "grad_norm": 0.11076867580413818,
+      "learning_rate": 9.910757130848571e-06,
+      "loss": 0.5022,
+      "step": 80
+    },
+    {
+      "epoch": 0.49693251533742333,
+      "grad_norm": 0.08907414972782135,
+      "learning_rate": 9.908121168001657e-06,
+      "loss": 0.524,
+      "step": 81
+    },
+    {
+      "epoch": 0.5030674846625767,
+      "grad_norm": 0.09341124445199966,
+      "learning_rate": 9.90544720136934e-06,
+      "loss": 0.4973,
+      "step": 82
+    },
+    {
+      "epoch": 0.50920245398773,
+      "grad_norm": 0.10275765508413315,
+      "learning_rate": 9.902735251656263e-06,
+      "loss": 0.5031,
+      "step": 83
+    },
+    {
+      "epoch": 0.5153374233128835,
+      "grad_norm": 0.08170262724161148,
+      "learning_rate": 9.89998533986116e-06,
+      "loss": 0.4829,
+      "step": 84
+    },
+    {
+      "epoch": 0.5214723926380368,
+      "grad_norm": 0.08169631659984589,
+      "learning_rate": 9.897197487276712e-06,
+      "loss": 0.4893,
+      "step": 85
+    },
+    {
+      "epoch": 0.5276073619631901,
+      "grad_norm": 0.07817935198545456,
+      "learning_rate": 9.894371715489376e-06,
+      "loss": 0.5155,
+      "step": 86
+    },
+    {
+      "epoch": 0.5337423312883436,
+      "grad_norm": 0.0947134792804718,
+      "learning_rate": 9.891508046379225e-06,
+      "loss": 0.5513,
+      "step": 87
+    },
+    {
+      "epoch": 0.5398773006134969,
+      "grad_norm": 0.08149490505456924,
+      "learning_rate": 9.888606502119763e-06,
+      "loss": 0.4797,
+      "step": 88
+    },
+    {
+      "epoch": 0.5460122699386503,
+      "grad_norm": 0.09330648183822632,
+      "learning_rate": 9.885667105177769e-06,
+      "loss": 0.4906,
+      "step": 89
+    },
+    {
+      "epoch": 0.5521472392638037,
+      "grad_norm": 0.07798902690410614,
+      "learning_rate": 9.882689878313114e-06,
+      "loss": 0.4751,
+      "step": 90
+    },
+    {
+      "epoch": 0.558282208588957,
+      "grad_norm": 0.08369090408086777,
+      "learning_rate": 9.879674844578588e-06,
+      "loss": 0.5378,
+      "step": 91
+    },
+    {
+      "epoch": 0.5644171779141104,
+      "grad_norm": 0.09182888269424438,
+      "learning_rate": 9.876622027319726e-06,
+      "loss": 0.5062,
+      "step": 92
+    },
+    {
+      "epoch": 0.5705521472392638,
+      "grad_norm": 0.07969052344560623,
+      "learning_rate": 9.873531450174616e-06,
+      "loss": 0.5108,
+      "step": 93
+    },
+    {
+      "epoch": 0.5766871165644172,
+      "grad_norm": 0.1103825718164444,
+      "learning_rate": 9.870403137073723e-06,
+      "loss": 0.5127,
+      "step": 94
+    },
+    {
+      "epoch": 0.5828220858895705,
+      "grad_norm": 0.0967809185385704,
+      "learning_rate": 9.867237112239708e-06,
+      "loss": 0.4884,
+      "step": 95
+    },
+    {
+      "epoch": 0.588957055214724,
+      "grad_norm": 0.07554975152015686,
+      "learning_rate": 9.86403340018723e-06,
+      "loss": 0.4743,
+      "step": 96
+    },
+    {
+      "epoch": 0.5950920245398773,
+      "grad_norm": 0.09686005860567093,
+      "learning_rate": 9.860792025722768e-06,
+      "loss": 0.5033,
+      "step": 97
+    },
+    {
+      "epoch": 0.6012269938650306,
+      "grad_norm": 0.07932725548744202,
+      "learning_rate": 9.857513013944413e-06,
+      "loss": 0.5121,
+      "step": 98
+    },
+    {
+      "epoch": 0.6073619631901841,
+      "grad_norm": 0.08431375026702881,
+      "learning_rate": 9.854196390241691e-06,
+      "loss": 0.5058,
+      "step": 99
+    },
+    {
+      "epoch": 0.6134969325153374,
+      "grad_norm": 0.081431545317173,
+      "learning_rate": 9.85084218029536e-06,
+      "loss": 0.5138,
+      "step": 100
+    },
+    {
+      "epoch": 0.6196319018404908,
+      "grad_norm": 0.08625893294811249,
+      "learning_rate": 9.847450410077202e-06,
+      "loss": 0.5052,
+      "step": 101
+    },
+    {
+      "epoch": 0.6257668711656442,
+      "grad_norm": 0.08018495887517929,
+      "learning_rate": 9.844021105849837e-06,
+      "loss": 0.5018,
+      "step": 102
+    },
+    {
+      "epoch": 0.6319018404907976,
+      "grad_norm": 0.07922230660915375,
+      "learning_rate": 9.840554294166507e-06,
+      "loss": 0.4949,
+      "step": 103
+    },
+    {
+      "epoch": 0.6380368098159509,
+      "grad_norm": 0.07988546043634415,
+      "learning_rate": 9.83705000187088e-06,
+      "loss": 0.4774,
+      "step": 104
+    },
+    {
+      "epoch": 0.6441717791411042,
+      "grad_norm": 0.08108440041542053,
+      "learning_rate": 9.833508256096837e-06,
+      "loss": 0.508,
+      "step": 105
+    },
+    {
+      "epoch": 0.6503067484662577,
+      "grad_norm": 0.07891630381345749,
+      "learning_rate": 9.829929084268262e-06,
+      "loss": 0.4729,
+      "step": 106
+    },
+    {
+      "epoch": 0.656441717791411,
+      "grad_norm": 0.07818285375833511,
+      "learning_rate": 9.82631251409883e-06,
+      "loss": 0.4535,
+      "step": 107
+    },
+    {
+      "epoch": 0.6625766871165644,
+      "grad_norm": 0.07954879850149155,
+      "learning_rate": 9.822658573591794e-06,
+      "loss": 0.5095,
+      "step": 108
+    },
+    {
+      "epoch": 0.6687116564417178,
+      "grad_norm": 0.08702922612428665,
+      "learning_rate": 9.818967291039767e-06,
+      "loss": 0.5049,
+      "step": 109
+    },
+    {
+      "epoch": 0.6748466257668712,
+      "grad_norm": 0.07859531790018082,
+      "learning_rate": 9.8152386950245e-06,
+      "loss": 0.4565,
+      "step": 110
+    },
+    {
+      "epoch": 0.6809815950920245,
+      "grad_norm": 0.09203047305345535,
+      "learning_rate": 9.811472814416669e-06,
+      "loss": 0.487,
+      "step": 111
+    },
+    {
+      "epoch": 0.6871165644171779,
+      "grad_norm": 0.08309823274612427,
+      "learning_rate": 9.807669678375643e-06,
+      "loss": 0.4731,
+      "step": 112
+    },
+    {
+      "epoch": 0.6932515337423313,
+      "grad_norm": 0.0878622755408287,
+      "learning_rate": 9.803829316349262e-06,
+      "loss": 0.5039,
+      "step": 113
+    },
+    {
+      "epoch": 0.6993865030674846,
+      "grad_norm": 0.09177032113075256,
+      "learning_rate": 9.799951758073607e-06,
+      "loss": 0.4605,
+      "step": 114
+    },
+    {
+      "epoch": 0.7055214723926381,
+      "grad_norm": 0.08374316245317459,
+      "learning_rate": 9.796037033572771e-06,
+      "loss": 0.4926,
+      "step": 115
+    },
+    {
+      "epoch": 0.7116564417177914,
+      "grad_norm": 0.08500220626592636,
+      "learning_rate": 9.792085173158633e-06,
+      "loss": 0.5156,
+      "step": 116
+    },
+    {
+      "epoch": 0.7177914110429447,
+      "grad_norm": 0.10756376385688782,
+      "learning_rate": 9.788096207430608e-06,
+      "loss": 0.4892,
+      "step": 117
+    },
+    {
+      "epoch": 0.7239263803680982,
+      "grad_norm": 0.08716326951980591,
+      "learning_rate": 9.784070167275422e-06,
+      "loss": 0.488,
+      "step": 118
+    },
+    {
+      "epoch": 0.7300613496932515,
+      "grad_norm": 0.08646374195814133,
+      "learning_rate": 9.780007083866872e-06,
+      "loss": 0.5176,
+      "step": 119
+    },
+    {
+      "epoch": 0.7361963190184049,
+      "grad_norm": 0.08787130564451218,
+      "learning_rate": 9.775906988665583e-06,
+      "loss": 0.5049,
+      "step": 120
+    },
+    {
+      "epoch": 0.7423312883435583,
+      "grad_norm": 0.11348209530115128,
+      "learning_rate": 9.771769913418758e-06,
+      "loss": 0.4736,
+      "step": 121
+    },
+    {
+      "epoch": 0.7484662576687117,
+      "grad_norm": 0.09234916418790817,
+      "learning_rate": 9.767595890159944e-06,
+      "loss": 0.4943,
+      "step": 122
+    },
+    {
+      "epoch": 0.754601226993865,
+      "grad_norm": 0.08790959417819977,
+      "learning_rate": 9.763384951208776e-06,
+      "loss": 0.5097,
+      "step": 123
+    },
+    {
+      "epoch": 0.7607361963190185,
+      "grad_norm": 0.08863342553377151,
+      "learning_rate": 9.759137129170728e-06,
+      "loss": 0.5241,
+      "step": 124
+    },
+    {
+      "epoch": 0.7668711656441718,
+      "grad_norm": 0.08124402910470963,
+      "learning_rate": 9.754852456936862e-06,
+      "loss": 0.4759,
+      "step": 125
+    },
+    {
+      "epoch": 0.7730061349693251,
+      "grad_norm": 0.08492821455001831,
+      "learning_rate": 9.750530967683573e-06,
+      "loss": 0.478,
+      "step": 126
+    },
+    {
+      "epoch": 0.7791411042944786,
+      "grad_norm": 0.09666764736175537,
+      "learning_rate": 9.746172694872332e-06,
+      "loss": 0.4842,
+      "step": 127
+    },
+    {
+      "epoch": 0.7852760736196319,
+      "grad_norm": 0.08827357739210129,
+      "learning_rate": 9.741777672249424e-06,
+      "loss": 0.5038,
+      "step": 128
+    },
+    {
+      "epoch": 0.7914110429447853,
+      "grad_norm": 0.09018061310052872,
+      "learning_rate": 9.737345933845692e-06,
+      "loss": 0.5357,
+      "step": 129
+    },
+    {
+      "epoch": 0.7975460122699386,
+      "grad_norm": 0.09164313971996307,
+      "learning_rate": 9.732877513976269e-06,
+      "loss": 0.5167,
+      "step": 130
+    },
+    {
+      "epoch": 0.803680981595092,
+      "grad_norm": 0.08618041127920151,
+      "learning_rate": 9.728372447240315e-06,
+      "loss": 0.4807,
+      "step": 131
+    },
+    {
+      "epoch": 0.8098159509202454,
+      "grad_norm": 0.10933379828929901,
+      "learning_rate": 9.72383076852075e-06,
+      "loss": 0.5073,
+      "step": 132
+    },
+    {
+      "epoch": 0.8159509202453987,
+      "grad_norm": 0.08988262712955475,
+      "learning_rate": 9.71925251298398e-06,
+      "loss": 0.5095,
+      "step": 133
+    },
+    {
+      "epoch": 0.8220858895705522,
+      "grad_norm": 0.08735327422618866,
+      "learning_rate": 9.714637716079627e-06,
+      "loss": 0.4838,
+      "step": 134
+    },
+    {
+      "epoch": 0.8282208588957055,
+      "grad_norm": 0.08958423137664795,
+      "learning_rate": 9.709986413540254e-06,
+      "loss": 0.4931,
+      "step": 135
+    },
+    {
+      "epoch": 0.8343558282208589,
+      "grad_norm": 0.08985975384712219,
+      "learning_rate": 9.705298641381089e-06,
+      "loss": 0.4776,
+      "step": 136
+    },
+    {
+      "epoch": 0.8404907975460123,
+      "grad_norm": 0.10067565739154816,
+      "learning_rate": 9.700574435899745e-06,
+      "loss": 0.4672,
+      "step": 137
+    },
+    {
+      "epoch": 0.8466257668711656,
+      "grad_norm": 0.09510832279920578,
+      "learning_rate": 9.695813833675943e-06,
+      "loss": 0.4497,
+      "step": 138
+    },
+    {
+      "epoch": 0.852760736196319,
+      "grad_norm": 0.09438912570476532,
+      "learning_rate": 9.691016871571219e-06,
+      "loss": 0.4935,
+      "step": 139
+    },
+    {
+      "epoch": 0.8588957055214724,
+      "grad_norm": 0.0871388390660286,
+      "learning_rate": 9.686183586728654e-06,
+      "loss": 0.5168,
+      "step": 140
+    },
+    {
+      "epoch": 0.8650306748466258,
+      "grad_norm": 0.08199909329414368,
+      "learning_rate": 9.681314016572572e-06,
+      "loss": 0.4852,
+      "step": 141
+    },
+    {
+      "epoch": 0.8711656441717791,
+      "grad_norm": 0.08726981282234192,
+      "learning_rate": 9.676408198808253e-06,
+      "loss": 0.468,
+      "step": 142
+    },
+    {
+      "epoch": 0.8773006134969326,
+      "grad_norm": 0.09949006140232086,
+      "learning_rate": 9.671466171421651e-06,
+      "loss": 0.4706,
+      "step": 143
+    },
+    {
+      "epoch": 0.8834355828220859,
+      "grad_norm": 0.0843021422624588,
+      "learning_rate": 9.666487972679085e-06,
+      "loss": 0.4729,
+      "step": 144
+    },
+    {
+      "epoch": 0.8895705521472392,
+      "grad_norm": 0.0966368094086647,
+      "learning_rate": 9.661473641126954e-06,
+      "loss": 0.4628,
+      "step": 145
+    },
+    {
+      "epoch": 0.8957055214723927,
+      "grad_norm": 0.09247634559869766,
+      "learning_rate": 9.65642321559144e-06,
+      "loss": 0.4907,
+      "step": 146
+    },
+    {
+      "epoch": 0.901840490797546,
+      "grad_norm": 0.09630908071994781,
+      "learning_rate": 9.651336735178191e-06,
+      "loss": 0.4878,
+      "step": 147
+    },
+    {
+      "epoch": 0.9079754601226994,
+      "grad_norm": 0.09485550224781036,
+      "learning_rate": 9.646214239272038e-06,
+      "loss": 0.4735,
+      "step": 148
+    },
+    {
+      "epoch": 0.9141104294478528,
+      "grad_norm": 0.09002482891082764,
+      "learning_rate": 9.64105576753668e-06,
+      "loss": 0.4755,
+      "step": 149
+    },
+    {
+      "epoch": 0.9202453987730062,
+      "grad_norm": 0.09443383663892746,
+      "learning_rate": 9.635861359914374e-06,
+      "loss": 0.486,
+      "step": 150
+    },
+    {
+      "epoch": 0.9263803680981595,
+      "grad_norm": 0.08977822959423065,
+      "learning_rate": 9.630631056625635e-06,
+      "loss": 0.4966,
+      "step": 151
+    },
+    {
+      "epoch": 0.9325153374233128,
+      "grad_norm": 0.09370667487382889,
+      "learning_rate": 9.62536489816892e-06,
+      "loss": 0.4792,
+      "step": 152
+    },
+    {
+      "epoch": 0.9386503067484663,
+      "grad_norm": 0.09303581714630127,
+      "learning_rate": 9.620062925320309e-06,
+      "loss": 0.4648,
+      "step": 153
+    },
+    {
+      "epoch": 0.9447852760736196,
+      "grad_norm": 0.08958807587623596,
+      "learning_rate": 9.614725179133197e-06,
+      "loss": 0.4786,
+      "step": 154
+    },
+    {
+      "epoch": 0.950920245398773,
+      "grad_norm": 0.10791140049695969,
+      "learning_rate": 9.609351700937976e-06,
+      "loss": 0.472,
+      "step": 155
+    },
+    {
+      "epoch": 0.9570552147239264,
+      "grad_norm": 0.09399349987506866,
+      "learning_rate": 9.60394253234171e-06,
+      "loss": 0.4529,
+      "step": 156
+    },
+    {
+      "epoch": 0.9631901840490797,
+      "grad_norm": 0.09015163034200668,
+      "learning_rate": 9.598497715227815e-06,
+      "loss": 0.4839,
+      "step": 157
+    },
+    {
+      "epoch": 0.9693251533742331,
+      "grad_norm": 0.09602297842502594,
+      "learning_rate": 9.593017291755733e-06,
+      "loss": 0.4806,
+      "step": 158
+    },
+    {
+      "epoch": 0.9754601226993865,
+      "grad_norm": 0.097842738032341,
+      "learning_rate": 9.587501304360612e-06,
+      "loss": 0.4518,
+      "step": 159
+    },
+    {
+      "epoch": 0.9815950920245399,
+      "grad_norm": 0.11291999369859695,
+      "learning_rate": 9.581949795752972e-06,
+      "loss": 0.4375,
+      "step": 160
+    },
+    {
+      "epoch": 0.9877300613496932,
+      "grad_norm": 0.0913916602730751,
+      "learning_rate": 9.576362808918368e-06,
+      "loss": 0.4771,
+      "step": 161
+    },
+    {
+      "epoch": 0.9938650306748467,
+      "grad_norm": 0.10012573003768921,
+      "learning_rate": 9.570740387117078e-06,
+      "loss": 0.4775,
+      "step": 162
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09484586119651794,
+      "learning_rate": 9.565082573883745e-06,
+      "loss": 0.4859,
+      "step": 163
+    },
+    {
+      "epoch": 1.0061349693251533,
+      "grad_norm": 0.09999288618564606,
+      "learning_rate": 9.559389413027048e-06,
+      "loss": 0.4723,
+      "step": 164
+    },
+    {
+      "epoch": 1.0122699386503067,
+      "grad_norm": 0.09577009081840515,
+      "learning_rate": 9.553660948629369e-06,
+      "loss": 0.492,
+      "step": 165
+    },
+    {
+      "epoch": 1.01840490797546,
+      "grad_norm": 0.1010693684220314,
+      "learning_rate": 9.547897225046445e-06,
+      "loss": 0.489,
+      "step": 166
+    },
+    {
+      "epoch": 1.0245398773006136,
+      "grad_norm": 0.10461648553609848,
+      "learning_rate": 9.542098286907024e-06,
+      "loss": 0.4795,
+      "step": 167
+    },
+    {
+      "epoch": 1.030674846625767,
+      "grad_norm": 0.10396389663219452,
+      "learning_rate": 9.536264179112529e-06,
+      "loss": 0.464,
+      "step": 168
+    },
+    {
+      "epoch": 1.0368098159509203,
+      "grad_norm": 0.09160111844539642,
+      "learning_rate": 9.530394946836694e-06,
+      "loss": 0.4997,
+      "step": 169
+    },
+    {
+      "epoch": 1.0429447852760736,
+      "grad_norm": 0.09714648127555847,
+      "learning_rate": 9.524490635525228e-06,
+      "loss": 0.4724,
+      "step": 170
+    },
+    {
+      "epoch": 1.049079754601227,
+      "grad_norm": 0.10137296468019485,
+      "learning_rate": 9.51855129089546e-06,
+      "loss": 0.4802,
+      "step": 171
+    },
+    {
+      "epoch": 1.0552147239263803,
+      "grad_norm": 0.09933356940746307,
+      "learning_rate": 9.51257695893598e-06,
+      "loss": 0.4478,
+      "step": 172
+    },
+    {
+      "epoch": 1.0613496932515338,
+      "grad_norm": 0.0967121571302414,
+      "learning_rate": 9.506567685906289e-06,
+      "loss": 0.4496,
+      "step": 173
+    },
+    {
+      "epoch": 1.0674846625766872,
+      "grad_norm": 0.10460793972015381,
+      "learning_rate": 9.500523518336435e-06,
+      "loss": 0.4569,
+      "step": 174
+    },
+    {
+      "epoch": 1.0736196319018405,
+      "grad_norm": 0.11570947617292404,
+      "learning_rate": 9.494444503026656e-06,
+      "loss": 0.4808,
+      "step": 175
+    },
+    {
+      "epoch": 1.0797546012269938,
+      "grad_norm": 0.1063118651509285,
+      "learning_rate": 9.488330687047025e-06,
+      "loss": 0.4889,
+      "step": 176
+    },
+    {
+      "epoch": 1.0858895705521472,
+      "grad_norm": 0.09846626967191696,
+      "learning_rate": 9.482182117737066e-06,
+      "loss": 0.4876,
+      "step": 177
+    },
+    {
+      "epoch": 1.0920245398773005,
+      "grad_norm": 0.09458671510219574,
+      "learning_rate": 9.475998842705412e-06,
+      "loss": 0.4633,
+      "step": 178
+    },
+    {
+      "epoch": 1.098159509202454,
+      "grad_norm": 0.0982397273182869,
+      "learning_rate": 9.469780909829411e-06,
+      "loss": 0.462,
+      "step": 179
+    },
+    {
+      "epoch": 1.1042944785276074,
+      "grad_norm": 0.09949040412902832,
+      "learning_rate": 9.46352836725478e-06,
+      "loss": 0.4677,
+      "step": 180
+    },
+    {
+      "epoch": 1.1104294478527608,
+      "grad_norm": 0.10287673771381378,
+      "learning_rate": 9.457241263395212e-06,
+      "loss": 0.4743,
+      "step": 181
+    },
+    {
+      "epoch": 1.116564417177914,
+      "grad_norm": 0.10356439650058746,
+      "learning_rate": 9.450919646932013e-06,
+      "loss": 0.483,
+      "step": 182
+    },
+    {
+      "epoch": 1.1226993865030674,
+      "grad_norm": 0.09517747908830643,
+      "learning_rate": 9.44456356681372e-06,
+      "loss": 0.4462,
+      "step": 183
+    },
+    {
+      "epoch": 1.1288343558282208,
+      "grad_norm": 0.12741521000862122,
+      "learning_rate": 9.438173072255727e-06,
+      "loss": 0.4577,
+      "step": 184
+    },
+    {
+      "epoch": 1.1349693251533743,
+      "grad_norm": 0.10354242473840714,
+      "learning_rate": 9.431748212739897e-06,
+      "loss": 0.4308,
+      "step": 185
+    },
+    {
+      "epoch": 1.1411042944785277,
+      "grad_norm": 0.11123675853013992,
+      "learning_rate": 9.425289038014184e-06,
+      "loss": 0.4937,
+      "step": 186
+    },
+    {
+      "epoch": 1.147239263803681,
+      "grad_norm": 0.09957747161388397,
+      "learning_rate": 9.418795598092243e-06,
+      "loss": 0.4646,
+      "step": 187
+    },
+    {
+      "epoch": 1.1533742331288344,
+      "grad_norm": 0.10711025446653366,
+      "learning_rate": 9.41226794325305e-06,
+      "loss": 0.5061,
+      "step": 188
+    },
+    {
+      "epoch": 1.1595092024539877,
+      "grad_norm": 0.09933678060770035,
+      "learning_rate": 9.405706124040506e-06,
+      "loss": 0.4725,
+      "step": 189
+    },
+    {
+      "epoch": 1.165644171779141,
+      "grad_norm": 0.10218477994203568,
+      "learning_rate": 9.399110191263048e-06,
+      "loss": 0.4745,
+      "step": 190
+    },
+    {
+      "epoch": 1.1717791411042944,
+      "grad_norm": 0.10867977887392044,
+      "learning_rate": 9.392480195993258e-06,
+      "loss": 0.4866,
+      "step": 191
+    },
+    {
+      "epoch": 1.177914110429448,
+      "grad_norm": 0.10019568353891373,
+      "learning_rate": 9.385816189567462e-06,
+      "loss": 0.4638,
+      "step": 192
+    },
+    {
+      "epoch": 1.1840490797546013,
+      "grad_norm": 0.0998578816652298,
+      "learning_rate": 9.379118223585342e-06,
+      "loss": 0.4609,
+      "step": 193
+    },
+    {
+      "epoch": 1.1901840490797546,
+      "grad_norm": 0.10544786602258682,
+      "learning_rate": 9.372386349909521e-06,
+      "loss": 0.4702,
+      "step": 194
+    },
+    {
+      "epoch": 1.196319018404908,
+      "grad_norm": 0.11477886140346527,
+      "learning_rate": 9.365620620665176e-06,
+      "loss": 0.4664,
+      "step": 195
+    },
+    {
+      "epoch": 1.2024539877300613,
+      "grad_norm": 0.09951955080032349,
+      "learning_rate": 9.358821088239632e-06,
+      "loss": 0.4672,
+      "step": 196
+    },
+    {
+      "epoch": 1.2085889570552146,
+      "grad_norm": 0.10202579945325851,
+      "learning_rate": 9.351987805281949e-06,
+      "loss": 0.4728,
+      "step": 197
+    },
+    {
+      "epoch": 1.2147239263803682,
+      "grad_norm": 0.11044152826070786,
+      "learning_rate": 9.345120824702515e-06,
+      "loss": 0.4306,
+      "step": 198
+    },
+    {
+      "epoch": 1.2208588957055215,
+      "grad_norm": 0.10835869610309601,
+      "learning_rate": 9.338220199672652e-06,
+      "loss": 0.4879,
+      "step": 199
+    },
+    {
+      "epoch": 1.2269938650306749,
+      "grad_norm": 0.10329335927963257,
+      "learning_rate": 9.331285983624182e-06,
+      "loss": 0.4728,
+      "step": 200
+    },
+    {
+      "epoch": 1.2331288343558282,
+      "grad_norm": 0.10831797868013382,
+      "learning_rate": 9.324318230249026e-06,
+      "loss": 0.4789,
+      "step": 201
+    },
+    {
+      "epoch": 1.2392638036809815,
+      "grad_norm": 0.10910773277282715,
+      "learning_rate": 9.317316993498788e-06,
+      "loss": 0.4642,
+      "step": 202
+    },
+    {
+      "epoch": 1.2453987730061349,
+      "grad_norm": 0.10746899992227554,
+      "learning_rate": 9.310282327584335e-06,
+      "loss": 0.4659,
+      "step": 203
+    },
+    {
+      "epoch": 1.2515337423312882,
+      "grad_norm": 0.11725542694330215,
+      "learning_rate": 9.303214286975373e-06,
+      "loss": 0.4629,
+      "step": 204
+    },
+    {
+      "epoch": 1.2576687116564418,
+      "grad_norm": 0.10740689188241959,
+      "learning_rate": 9.296112926400038e-06,
+      "loss": 0.4835,
+      "step": 205
+    },
+    {
+      "epoch": 1.2638036809815951,
+      "grad_norm": 0.10859764367341995,
+      "learning_rate": 9.288978300844456e-06,
+      "loss": 0.4555,
+      "step": 206
+    },
+    {
+      "epoch": 1.2699386503067485,
+      "grad_norm": 0.11524678766727448,
+      "learning_rate": 9.281810465552327e-06,
+      "loss": 0.4597,
+      "step": 207
+    },
+    {
+      "epoch": 1.2760736196319018,
+      "grad_norm": 0.11158965528011322,
+      "learning_rate": 9.274609476024499e-06,
+      "loss": 0.4903,
+      "step": 208
+    },
+    {
+      "epoch": 1.2822085889570551,
+      "grad_norm": 0.10941608250141144,
+      "learning_rate": 9.26737538801853e-06,
+      "loss": 0.4778,
+      "step": 209
+    },
+    {
+      "epoch": 1.2883435582822087,
+      "grad_norm": 0.10840465873479843,
+      "learning_rate": 9.260108257548264e-06,
+      "loss": 0.4635,
+      "step": 210
+    },
+    {
+      "epoch": 1.294478527607362,
+      "grad_norm": 0.11878825724124908,
+      "learning_rate": 9.252808140883393e-06,
+      "loss": 0.4668,
+      "step": 211
+    },
+    {
+      "epoch": 1.3006134969325154,
+      "grad_norm": 0.11123567819595337,
+      "learning_rate": 9.24547509454902e-06,
+      "loss": 0.4587,
+      "step": 212
+    },
+    {
+      "epoch": 1.3067484662576687,
+      "grad_norm": 0.1218094527721405,
+      "learning_rate": 9.238109175325232e-06,
+      "loss": 0.4741,
+      "step": 213
+    },
+    {
+      "epoch": 1.312883435582822,
+      "grad_norm": 0.11062923818826675,
+      "learning_rate": 9.230710440246642e-06,
+      "loss": 0.4712,
+      "step": 214
+    },
+    {
+      "epoch": 1.3190184049079754,
+      "grad_norm": 0.11379806697368622,
+      "learning_rate": 9.223278946601963e-06,
+      "loss": 0.4822,
+      "step": 215
+    },
+    {
+      "epoch": 1.3251533742331287,
+      "grad_norm": 0.11029627174139023,
+      "learning_rate": 9.215814751933559e-06,
+      "loss": 0.4479,
+      "step": 216
+    },
+    {
+      "epoch": 1.331288343558282,
+      "grad_norm": 0.10827028751373291,
+      "learning_rate": 9.208317914036997e-06,
+      "loss": 0.4707,
+      "step": 217
+    },
+    {
+      "epoch": 1.3374233128834356,
+      "grad_norm": 0.11715540289878845,
+      "learning_rate": 9.200788490960605e-06,
+      "loss": 0.4576,
+      "step": 218
+    },
+    {
+      "epoch": 1.343558282208589,
+      "grad_norm": 0.11236132681369781,
+      "learning_rate": 9.193226541005015e-06,
+      "loss": 0.4646,
+      "step": 219
+    },
+    {
+      "epoch": 1.3496932515337423,
+      "grad_norm": 0.1052820086479187,
+      "learning_rate": 9.185632122722719e-06,
+      "loss": 0.4761,
+      "step": 220
+    },
+    {
+      "epoch": 1.3558282208588956,
+      "grad_norm": 0.12960059940814972,
+      "learning_rate": 9.178005294917615e-06,
+      "loss": 0.4741,
+      "step": 221
+    },
+    {
+      "epoch": 1.3619631901840492,
+      "grad_norm": 0.11204280704259872,
+      "learning_rate": 9.170346116644545e-06,
+      "loss": 0.4752,
+      "step": 222
+    },
+    {
+      "epoch": 1.3680981595092025,
+      "grad_norm": 0.11414473503828049,
+      "learning_rate": 9.16265464720884e-06,
+      "loss": 0.479,
+      "step": 223
+    },
+    {
+      "epoch": 1.3742331288343559,
+      "grad_norm": 0.12588758766651154,
+      "learning_rate": 9.154930946165872e-06,
+      "loss": 0.4299,
+      "step": 224
+    },
+    {
+      "epoch": 1.3803680981595092,
+      "grad_norm": 0.12927886843681335,
+      "learning_rate": 9.147175073320574e-06,
+      "loss": 0.4669,
+      "step": 225
+    },
+    {
+      "epoch": 1.3865030674846626,
+      "grad_norm": 0.1076217070221901,
+      "learning_rate": 9.13938708872699e-06,
+      "loss": 0.4788,
+      "step": 226
+    },
+    {
+      "epoch": 1.392638036809816,
+      "grad_norm": 0.11360085755586624,
+      "learning_rate": 9.131567052687811e-06,
+      "loss": 0.4374,
+      "step": 227
+    },
+    {
+      "epoch": 1.3987730061349692,
+      "grad_norm": 0.11873716115951538,
+      "learning_rate": 9.123715025753896e-06,
+      "loss": 0.4401,
+      "step": 228
+    },
+    {
+      "epoch": 1.4049079754601226,
+      "grad_norm": 0.11451287567615509,
+      "learning_rate": 9.115831068723816e-06,
+      "loss": 0.4811,
+      "step": 229
+    },
+    {
+      "epoch": 1.4110429447852761,
+      "grad_norm": 0.11967512965202332,
+      "learning_rate": 9.10791524264338e-06,
+      "loss": 0.4767,
+      "step": 230
+    },
+    {
+      "epoch": 1.4171779141104295,
+      "grad_norm": 0.11932694166898727,
+      "learning_rate": 9.099967608805152e-06,
+      "loss": 0.4567,
+      "step": 231
+    },
+    {
+      "epoch": 1.4233128834355828,
+      "grad_norm": 0.11423443257808685,
+      "learning_rate": 9.091988228747992e-06,
+      "loss": 0.4549,
+      "step": 232
+    },
+    {
+      "epoch": 1.4294478527607362,
+      "grad_norm": 0.11714685708284378,
+      "learning_rate": 9.08397716425657e-06,
+      "loss": 0.4259,
+      "step": 233
+    },
+    {
+      "epoch": 1.4355828220858895,
+      "grad_norm": 0.1315395087003708,
+      "learning_rate": 9.07593447736089e-06,
+      "loss": 0.502,
+      "step": 234
+    },
+    {
+      "epoch": 1.441717791411043,
+      "grad_norm": 0.1216348186135292,
+      "learning_rate": 9.06786023033581e-06,
+      "loss": 0.4622,
+      "step": 235
+    },
+    {
+      "epoch": 1.4478527607361964,
+      "grad_norm": 0.11694086343050003,
+      "learning_rate": 9.059754485700557e-06,
+      "loss": 0.4413,
+      "step": 236
+    },
+    {
+      "epoch": 1.4539877300613497,
+      "grad_norm": 0.12045416980981827,
+      "learning_rate": 9.05161730621825e-06,
+      "loss": 0.4579,
+      "step": 237
+    },
+    {
+      "epoch": 1.460122699386503,
+      "grad_norm": 0.11791419237852097,
+      "learning_rate": 9.043448754895405e-06,
+      "loss": 0.4712,
+      "step": 238
+    },
+    {
+      "epoch": 1.4662576687116564,
+      "grad_norm": 0.12033417075872421,
+      "learning_rate": 9.035248894981454e-06,
+      "loss": 0.4787,
+      "step": 239
+    },
+    {
+      "epoch": 1.4723926380368098,
+      "grad_norm": 0.1140502542257309,
+      "learning_rate": 9.02701778996825e-06,
+      "loss": 0.4909,
+      "step": 240
+    },
+    {
+      "epoch": 1.478527607361963,
+      "grad_norm": 0.1263529509305954,
+      "learning_rate": 9.018755503589582e-06,
+      "loss": 0.4606,
+      "step": 241
+    },
+    {
+      "epoch": 1.4846625766871164,
+      "grad_norm": 0.11217690259218216,
+      "learning_rate": 9.010462099820674e-06,
+      "loss": 0.4455,
+      "step": 242
+    },
+    {
+      "epoch": 1.49079754601227,
+      "grad_norm": 0.12041988223791122,
+      "learning_rate": 9.002137642877696e-06,
+      "loss": 0.4317,
+      "step": 243
+    },
+    {
+      "epoch": 1.4969325153374233,
+      "grad_norm": 0.128048375248909,
+      "learning_rate": 8.993782197217262e-06,
+      "loss": 0.4843,
+      "step": 244
+    },
+    {
+      "epoch": 1.5030674846625767,
+      "grad_norm": 0.12902449071407318,
+      "learning_rate": 8.985395827535934e-06,
+      "loss": 0.4224,
+      "step": 245
+    },
+    {
+      "epoch": 1.50920245398773,
+      "grad_norm": 0.1286553293466568,
+      "learning_rate": 8.976978598769719e-06,
+      "loss": 0.4613,
+      "step": 246
+    },
+    {
+      "epoch": 1.5153374233128836,
+      "grad_norm": 0.11924152076244354,
+      "learning_rate": 8.96853057609357e-06,
+      "loss": 0.4288,
+      "step": 247
+    },
+    {
+      "epoch": 1.521472392638037,
+      "grad_norm": 0.12467361986637115,
+      "learning_rate": 8.960051824920873e-06,
+      "loss": 0.4631,
+      "step": 248
+    },
+    {
+      "epoch": 1.5276073619631902,
+      "grad_norm": 0.12638317048549652,
+      "learning_rate": 8.951542410902949e-06,
+      "loss": 0.4331,
+      "step": 249
+    },
+    {
+      "epoch": 1.5337423312883436,
+      "grad_norm": 0.11676699668169022,
+      "learning_rate": 8.943002399928547e-06,
+      "loss": 0.4565,
+      "step": 250
+    },
+    {
+      "epoch": 1.539877300613497,
+      "grad_norm": 0.1417926698923111,
+      "learning_rate": 8.934431858123324e-06,
+      "loss": 0.4796,
+      "step": 251
+    },
+    {
+      "epoch": 1.5460122699386503,
+      "grad_norm": 0.1345888376235962,
+      "learning_rate": 8.925830851849338e-06,
+      "loss": 0.4658,
+      "step": 252
+    },
+    {
+      "epoch": 1.5521472392638036,
+      "grad_norm": 0.11821699142456055,
+      "learning_rate": 8.917199447704538e-06,
+      "loss": 0.4376,
+      "step": 253
+    },
+    {
+      "epoch": 1.558282208588957,
+      "grad_norm": 0.12282978743314743,
+      "learning_rate": 8.908537712522246e-06,
+      "loss": 0.4799,
+      "step": 254
+    },
+    {
+      "epoch": 1.5644171779141103,
+      "grad_norm": 0.1122986450791359,
+      "learning_rate": 8.899845713370632e-06,
+      "loss": 0.4423,
+      "step": 255
+    },
+    {
+      "epoch": 1.5705521472392638,
+      "grad_norm": 0.1364981234073639,
+      "learning_rate": 8.891123517552208e-06,
+      "loss": 0.4559,
+      "step": 256
+    },
+    {
+      "epoch": 1.5766871165644172,
+      "grad_norm": 0.14287355542182922,
+      "learning_rate": 8.882371192603297e-06,
+      "loss": 0.4446,
+      "step": 257
+    },
+    {
+      "epoch": 1.5828220858895705,
+      "grad_norm": 0.13158780336380005,
+      "learning_rate": 8.87358880629351e-06,
+      "loss": 0.4805,
+      "step": 258
+    },
+    {
+      "epoch": 1.588957055214724,
+      "grad_norm": 0.14378562569618225,
+      "learning_rate": 8.864776426625231e-06,
+      "loss": 0.4656,
+      "step": 259
+    },
+    {
+      "epoch": 1.5950920245398774,
+      "grad_norm": 0.1245376318693161,
+      "learning_rate": 8.855934121833083e-06,
+      "loss": 0.4425,
+      "step": 260
+    },
+    {
+      "epoch": 1.6012269938650308,
+      "grad_norm": 0.12470446527004242,
+      "learning_rate": 8.847061960383395e-06,
+      "loss": 0.4495,
+      "step": 261
+    },
+    {
+      "epoch": 1.607361963190184,
+      "grad_norm": 0.11994479596614838,
+      "learning_rate": 8.83816001097368e-06,
+      "loss": 0.4169,
+      "step": 262
+    },
+    {
+      "epoch": 1.6134969325153374,
+      "grad_norm": 0.13416039943695068,
+      "learning_rate": 8.8292283425321e-06,
+      "loss": 0.4749,
+      "step": 263
+    },
+    {
+      "epoch": 1.6196319018404908,
+      "grad_norm": 0.12816861271858215,
+      "learning_rate": 8.820267024216937e-06,
+      "loss": 0.4464,
+      "step": 264
+    },
+    {
+      "epoch": 1.6257668711656441,
+      "grad_norm": 0.1400822103023529,
+      "learning_rate": 8.811276125416048e-06,
+      "loss": 0.464,
+      "step": 265
+    },
+    {
+      "epoch": 1.6319018404907975,
+      "grad_norm": 0.1271728277206421,
+      "learning_rate": 8.802255715746333e-06,
+      "loss": 0.4798,
+      "step": 266
+    },
+    {
+      "epoch": 1.6380368098159508,
+      "grad_norm": 0.11546501517295837,
+      "learning_rate": 8.7932058650532e-06,
+      "loss": 0.4716,
+      "step": 267
+    },
+    {
+      "epoch": 1.6441717791411041,
+      "grad_norm": 0.12347917258739471,
+      "learning_rate": 8.784126643410015e-06,
+      "loss": 0.446,
+      "step": 268
+    },
+    {
+      "epoch": 1.6503067484662577,
+      "grad_norm": 0.1298205405473709,
+      "learning_rate": 8.775018121117569e-06,
+      "loss": 0.4569,
+      "step": 269
+    },
+    {
+      "epoch": 1.656441717791411,
+      "grad_norm": 0.12920261919498444,
+      "learning_rate": 8.765880368703527e-06,
+      "loss": 0.4946,
+      "step": 270
+    },
+    {
+      "epoch": 1.6625766871165644,
+      "grad_norm": 0.11939296126365662,
+      "learning_rate": 8.756713456921885e-06,
+      "loss": 0.4703,
+      "step": 271
+    },
+    {
+      "epoch": 1.668711656441718,
+      "grad_norm": 0.12618795037269592,
+      "learning_rate": 8.747517456752419e-06,
+      "loss": 0.4665,
+      "step": 272
+    },
+    {
+      "epoch": 1.6748466257668713,
+      "grad_norm": 0.1314852237701416,
+      "learning_rate": 8.73829243940014e-06,
+      "loss": 0.4577,
+      "step": 273
+    },
+    {
+      "epoch": 1.6809815950920246,
+      "grad_norm": 0.14323115348815918,
+      "learning_rate": 8.729038476294737e-06,
+      "loss": 0.428,
+      "step": 274
+    },
+    {
+      "epoch": 1.687116564417178,
+      "grad_norm": 0.1325852870941162,
+      "learning_rate": 8.719755639090032e-06,
+      "loss": 0.443,
+      "step": 275
+    },
+    {
+      "epoch": 1.6932515337423313,
+      "grad_norm": 0.12698835134506226,
+      "learning_rate": 8.710443999663417e-06,
+      "loss": 0.5045,
+      "step": 276
+    },
+    {
+      "epoch": 1.6993865030674846,
+      "grad_norm": 0.12905320525169373,
+      "learning_rate": 8.701103630115303e-06,
+      "loss": 0.4664,
+      "step": 277
+    },
+    {
+      "epoch": 1.705521472392638,
+      "grad_norm": 0.12309697270393372,
+      "learning_rate": 8.691734602768554e-06,
+      "loss": 0.4227,
+      "step": 278
+    },
+    {
+      "epoch": 1.7116564417177913,
+      "grad_norm": 0.12686602771282196,
+      "learning_rate": 8.68233699016794e-06,
+      "loss": 0.4878,
+      "step": 279
+    },
+    {
+      "epoch": 1.7177914110429446,
+      "grad_norm": 0.13946817815303802,
+      "learning_rate": 8.672910865079564e-06,
+      "loss": 0.4487,
+      "step": 280
+    },
+    {
+      "epoch": 1.7239263803680982,
+      "grad_norm": 0.13267400860786438,
+      "learning_rate": 8.663456300490302e-06,
+      "loss": 0.4481,
+      "step": 281
+    },
+    {
+      "epoch": 1.7300613496932515,
+      "grad_norm": 0.12559852004051208,
+      "learning_rate": 8.65397336960724e-06,
+      "loss": 0.449,
+      "step": 282
+    },
+    {
+      "epoch": 1.7361963190184049,
+      "grad_norm": 0.1440388262271881,
+      "learning_rate": 8.644462145857104e-06,
+      "loss": 0.4466,
+      "step": 283
+    },
+    {
+      "epoch": 1.7423312883435584,
+      "grad_norm": 0.1397363543510437,
+      "learning_rate": 8.634922702885693e-06,
+      "loss": 0.4113,
+      "step": 284
+    },
+    {
+      "epoch": 1.7484662576687118,
+      "grad_norm": 0.13909780979156494,
+      "learning_rate": 8.62535511455731e-06,
+      "loss": 0.4504,
+      "step": 285
+    },
+    {
+      "epoch": 1.7546012269938651,
+      "grad_norm": 0.128562331199646,
+      "learning_rate": 8.615759454954187e-06,
+      "loss": 0.4593,
+      "step": 286
+    },
+    {
+      "epoch": 1.7607361963190185,
+      "grad_norm": 0.12467021495103836,
+      "learning_rate": 8.60613579837591e-06,
+      "loss": 0.4728,
+      "step": 287
+    },
+    {
+      "epoch": 1.7668711656441718,
+      "grad_norm": 0.14602142572402954,
+      "learning_rate": 8.596484219338856e-06,
+      "loss": 0.4732,
+      "step": 288
+    },
+    {
+      "epoch": 1.7730061349693251,
+      "grad_norm": 0.14744794368743896,
+      "learning_rate": 8.586804792575596e-06,
+      "loss": 0.4308,
+      "step": 289
+    },
+    {
+      "epoch": 1.7791411042944785,
+      "grad_norm": 0.12893109023571014,
+      "learning_rate": 8.577097593034338e-06,
+      "loss": 0.4589,
+      "step": 290
+    },
+    {
+      "epoch": 1.7852760736196318,
+      "grad_norm": 0.1237974613904953,
+      "learning_rate": 8.567362695878325e-06,
+      "loss": 0.477,
+      "step": 291
+    },
+    {
+      "epoch": 1.7914110429447851,
+      "grad_norm": 0.12805530428886414,
+      "learning_rate": 8.55760017648527e-06,
+      "loss": 0.4544,
+      "step": 292
+    },
+    {
+      "epoch": 1.7975460122699385,
+      "grad_norm": 0.1318630576133728,
+      "learning_rate": 8.547810110446766e-06,
+      "loss": 0.4428,
+      "step": 293
+    },
+    {
+      "epoch": 1.803680981595092,
+      "grad_norm": 0.13845506310462952,
+      "learning_rate": 8.537992573567698e-06,
+      "loss": 0.4704,
+      "step": 294
+    },
+    {
+      "epoch": 1.8098159509202454,
+      "grad_norm": 0.14310558140277863,
+      "learning_rate": 8.528147641865661e-06,
+      "loss": 0.4534,
+      "step": 295
+    },
+    {
+      "epoch": 1.8159509202453987,
+      "grad_norm": 0.14174222946166992,
+      "learning_rate": 8.518275391570368e-06,
+      "loss": 0.435,
+      "step": 296
+    },
+    {
+      "epoch": 1.8220858895705523,
+      "grad_norm": 0.14490178227424622,
+      "learning_rate": 8.508375899123064e-06,
+      "loss": 0.4255,
+      "step": 297
+    },
+    {
+      "epoch": 1.8282208588957056,
+      "grad_norm": 0.13574454188346863,
+      "learning_rate": 8.498449241175927e-06,
+      "loss": 0.4384,
+      "step": 298
+    },
+    {
+      "epoch": 1.834355828220859,
+      "grad_norm": 0.12258850038051605,
+      "learning_rate": 8.488495494591482e-06,
+      "loss": 0.4114,
+      "step": 299
+    },
+    {
+      "epoch": 1.8404907975460123,
+      "grad_norm": 0.13615848124027252,
+      "learning_rate": 8.478514736441998e-06,
+      "loss": 0.4566,
+      "step": 300
+    },
+    {
+      "epoch": 1.8466257668711656,
+      "grad_norm": 0.1377963125705719,
+      "learning_rate": 8.468507044008902e-06,
+      "loss": 0.4452,
+      "step": 301
+    },
+    {
+      "epoch": 1.852760736196319,
+      "grad_norm": 0.13099685311317444,
+      "learning_rate": 8.458472494782169e-06,
+      "loss": 0.432,
+      "step": 302
+    },
+    {
+      "epoch": 1.8588957055214723,
+      "grad_norm": 0.13411100208759308,
+      "learning_rate": 8.44841116645973e-06,
+      "loss": 0.4551,
+      "step": 303
+    },
+    {
+      "epoch": 1.8650306748466257,
+      "grad_norm": 0.13292278349399567,
+      "learning_rate": 8.438323136946865e-06,
+      "loss": 0.4757,
+      "step": 304
+    },
+    {
+      "epoch": 1.871165644171779,
+      "grad_norm": 0.12634415924549103,
+      "learning_rate": 8.428208484355606e-06,
+      "loss": 0.4427,
+      "step": 305
+    },
+    {
+      "epoch": 1.8773006134969326,
+      "grad_norm": 0.147501140832901,
+      "learning_rate": 8.418067287004125e-06,
+      "loss": 0.4714,
+      "step": 306
+    },
+    {
+      "epoch": 1.883435582822086,
+      "grad_norm": 0.14923636615276337,
+      "learning_rate": 8.407899623416136e-06,
+      "loss": 0.4752,
+      "step": 307
+    },
+    {
+      "epoch": 1.8895705521472392,
+      "grad_norm": 0.126709446310997,
+      "learning_rate": 8.397705572320275e-06,
+      "loss": 0.476,
+      "step": 308
+    },
+    {
+      "epoch": 1.8957055214723928,
+      "grad_norm": 0.12852588295936584,
+      "learning_rate": 8.387485212649505e-06,
+      "loss": 0.4335,
+      "step": 309
+    },
+    {
+      "epoch": 1.9018404907975461,
+      "grad_norm": 0.13321667909622192,
+      "learning_rate": 8.377238623540491e-06,
+      "loss": 0.4517,
+      "step": 310
+    },
+    {
+      "epoch": 1.9079754601226995,
+      "grad_norm": 0.14450277388095856,
+      "learning_rate": 8.366965884333001e-06,
+      "loss": 0.4354,
+      "step": 311
+    },
+    {
+      "epoch": 1.9141104294478528,
+      "grad_norm": 0.12820766866207123,
+      "learning_rate": 8.356667074569274e-06,
+      "loss": 0.4088,
+      "step": 312
+    },
+    {
+      "epoch": 1.9202453987730062,
+      "grad_norm": 0.13799835741519928,
+      "learning_rate": 8.346342273993427e-06,
+      "loss": 0.47,
+      "step": 313
+    },
+    {
+      "epoch": 1.9263803680981595,
+      "grad_norm": 0.1558215469121933,
+      "learning_rate": 8.335991562550813e-06,
+      "loss": 0.4682,
+      "step": 314
+    },
+    {
+      "epoch": 1.9325153374233128,
+      "grad_norm": 0.14499524235725403,
+      "learning_rate": 8.325615020387422e-06,
+      "loss": 0.4896,
+      "step": 315
+    },
+    {
+      "epoch": 1.9386503067484662,
+      "grad_norm": 0.1416790634393692,
+      "learning_rate": 8.31521272784925e-06,
+      "loss": 0.4531,
+      "step": 316
+    },
+    {
+      "epoch": 1.9447852760736195,
+      "grad_norm": 0.1342548131942749,
+      "learning_rate": 8.304784765481676e-06,
+      "loss": 0.4403,
+      "step": 317
+    },
+    {
+      "epoch": 1.9509202453987728,
+      "grad_norm": 0.1562686264514923,
+      "learning_rate": 8.294331214028845e-06,
+      "loss": 0.4244,
+      "step": 318
+    },
+    {
+      "epoch": 1.9570552147239264,
+      "grad_norm": 0.13778996467590332,
+      "learning_rate": 8.283852154433042e-06,
+      "loss": 0.484,
+      "step": 319
+    },
+    {
+      "epoch": 1.9631901840490797,
+      "grad_norm": 0.14441350102424622,
+      "learning_rate": 8.273347667834057e-06,
+      "loss": 0.4534,
+      "step": 320
+    },
+    {
+      "epoch": 1.969325153374233,
+      "grad_norm": 0.15606369078159332,
+      "learning_rate": 8.262817835568563e-06,
+      "loss": 0.4403,
+      "step": 321
+    },
+    {
+      "epoch": 1.9754601226993866,
+      "grad_norm": 0.12662853300571442,
+      "learning_rate": 8.25226273916949e-06,
+      "loss": 0.4673,
+      "step": 322
+    },
+    {
+      "epoch": 1.98159509202454,
+      "grad_norm": 0.13923388719558716,
+      "learning_rate": 8.241682460365383e-06,
+      "loss": 0.4613,
+      "step": 323
+    },
+    {
+      "epoch": 1.9877300613496933,
+      "grad_norm": 0.13272419571876526,
+      "learning_rate": 8.231077081079781e-06,
+      "loss": 0.4716,
+      "step": 324
+    },
+    {
+      "epoch": 1.9938650306748467,
+      "grad_norm": 0.146401509642601,
+      "learning_rate": 8.220446683430577e-06,
+      "loss": 0.4427,
+      "step": 325
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.15548086166381836,
+      "learning_rate": 8.209791349729376e-06,
+      "loss": 0.4837,
+      "step": 326
+    },
+    {
+      "epoch": 2.0061349693251533,
+      "grad_norm": 0.14179112017154694,
+      "learning_rate": 8.199111162480871e-06,
+      "loss": 0.4326,
+      "step": 327
+    },
+    {
+      "epoch": 2.0122699386503067,
+      "grad_norm": 0.15060554444789886,
+      "learning_rate": 8.188406204382192e-06,
+      "loss": 0.4513,
+      "step": 328
+    },
+    {
+      "epoch": 2.01840490797546,
+      "grad_norm": 0.1416303813457489,
+      "learning_rate": 8.177676558322274e-06,
+      "loss": 0.4639,
+      "step": 329
+    },
+    {
+      "epoch": 2.0245398773006134,
+      "grad_norm": 0.15585218369960785,
+      "learning_rate": 8.16692230738121e-06,
+      "loss": 0.4315,
+      "step": 330
+    },
+    {
+      "epoch": 2.0306748466257667,
+      "grad_norm": 0.1324591487646103,
+      "learning_rate": 8.15614353482961e-06,
+      "loss": 0.4231,
+      "step": 331
+    },
+    {
+      "epoch": 2.03680981595092,
+      "grad_norm": 0.1416635811328888,
+      "learning_rate": 8.145340324127958e-06,
+      "loss": 0.4459,
+      "step": 332
+    },
+    {
+      "epoch": 2.042944785276074,
+      "grad_norm": 0.13037635385990143,
+      "learning_rate": 8.134512758925958e-06,
+      "loss": 0.4439,
+      "step": 333
+    },
+    {
+      "epoch": 2.049079754601227,
+      "grad_norm": 0.1375626176595688,
+      "learning_rate": 8.123660923061902e-06,
+      "loss": 0.4516,
+      "step": 334
+    },
+    {
+      "epoch": 2.0552147239263805,
+      "grad_norm": 0.1339251846075058,
+      "learning_rate": 8.112784900561997e-06,
+      "loss": 0.4357,
+      "step": 335
+    },
+    {
+      "epoch": 2.061349693251534,
+      "grad_norm": 0.14214186370372772,
+      "learning_rate": 8.10188477563974e-06,
+      "loss": 0.4511,
+      "step": 336
+    },
+    {
+      "epoch": 2.067484662576687,
+      "grad_norm": 0.1451665312051773,
+      "learning_rate": 8.090960632695246e-06,
+      "loss": 0.4122,
+      "step": 337
+    },
+    {
+      "epoch": 2.0736196319018405,
+      "grad_norm": 0.1382775753736496,
+      "learning_rate": 8.080012556314611e-06,
+      "loss": 0.4119,
+      "step": 338
+    },
+    {
+      "epoch": 2.079754601226994,
+      "grad_norm": 0.13937808573246002,
+      "learning_rate": 8.069040631269239e-06,
+      "loss": 0.4312,
+      "step": 339
+    },
+    {
+      "epoch": 2.085889570552147,
+      "grad_norm": 0.15605536103248596,
+      "learning_rate": 8.058044942515204e-06,
+      "loss": 0.4578,
+      "step": 340
+    },
+    {
+      "epoch": 2.0920245398773005,
+      "grad_norm": 0.13876931369304657,
+      "learning_rate": 8.047025575192576e-06,
+      "loss": 0.4444,
+      "step": 341
+    },
+    {
+      "epoch": 2.098159509202454,
+      "grad_norm": 0.1334417313337326,
+      "learning_rate": 8.035982614624774e-06,
+      "loss": 0.4576,
+      "step": 342
+    },
+    {
+      "epoch": 2.104294478527607,
+      "grad_norm": 0.14770197868347168,
+      "learning_rate": 8.024916146317896e-06,
+      "loss": 0.4584,
+      "step": 343
+    },
+    {
+      "epoch": 2.1104294478527605,
+      "grad_norm": 0.1419200599193573,
+      "learning_rate": 8.013826255960069e-06,
+      "loss": 0.4213,
+      "step": 344
+    },
+    {
+      "epoch": 2.116564417177914,
+      "grad_norm": 0.14547352492809296,
+      "learning_rate": 8.00271302942077e-06,
+      "loss": 0.4589,
+      "step": 345
+    },
+    {
+      "epoch": 2.1226993865030677,
+      "grad_norm": 0.15418393909931183,
+      "learning_rate": 7.991576552750173e-06,
+      "loss": 0.4393,
+      "step": 346
+    },
+    {
+      "epoch": 2.128834355828221,
+      "grad_norm": 0.1524142473936081,
+      "learning_rate": 7.980416912178478e-06,
+      "loss": 0.4594,
+      "step": 347
+    },
+    {
+      "epoch": 2.1349693251533743,
+      "grad_norm": 0.21241053938865662,
+      "learning_rate": 7.969234194115245e-06,
+      "loss": 0.4578,
+      "step": 348
+    },
+    {
+      "epoch": 2.1411042944785277,
+      "grad_norm": 0.15679003298282623,
+      "learning_rate": 7.95802848514872e-06,
+      "loss": 0.4427,
+      "step": 349
+    },
+    {
+      "epoch": 2.147239263803681,
+      "grad_norm": 0.14540907740592957,
+      "learning_rate": 7.946799872045173e-06,
+      "loss": 0.4388,
+      "step": 350
+    },
+    {
+      "epoch": 2.1533742331288344,
+      "grad_norm": 0.14929597079753876,
+      "learning_rate": 7.935548441748221e-06,
+      "loss": 0.4787,
+      "step": 351
+    },
+    {
+      "epoch": 2.1595092024539877,
+      "grad_norm": 0.1697789877653122,
+      "learning_rate": 7.924274281378153e-06,
+      "loss": 0.4433,
+      "step": 352
+    },
+    {
+      "epoch": 2.165644171779141,
+      "grad_norm": 0.15002626180648804,
+      "learning_rate": 7.912977478231262e-06,
+      "loss": 0.4689,
+      "step": 353
+    },
+    {
+      "epoch": 2.1717791411042944,
+      "grad_norm": 0.15629152953624725,
+      "learning_rate": 7.90165811977916e-06,
+      "loss": 0.445,
+      "step": 354
+    },
+    {
+      "epoch": 2.1779141104294477,
+      "grad_norm": 0.14952926337718964,
+      "learning_rate": 7.890316293668108e-06,
+      "loss": 0.452,
+      "step": 355
+    },
+    {
+      "epoch": 2.184049079754601,
+      "grad_norm": 0.18062762916088104,
+      "learning_rate": 7.878952087718336e-06,
+      "loss": 0.4675,
+      "step": 356
+    },
+    {
+      "epoch": 2.190184049079755,
+      "grad_norm": 0.13083772361278534,
+      "learning_rate": 7.867565589923364e-06,
+      "loss": 0.4529,
+      "step": 357
+    },
+    {
+      "epoch": 2.196319018404908,
+      "grad_norm": 0.3461189270019531,
+      "learning_rate": 7.856156888449312e-06,
+      "loss": 0.4596,
+      "step": 358
+    },
+    {
+      "epoch": 2.2024539877300615,
+      "grad_norm": 0.1565181016921997,
+      "learning_rate": 7.844726071634228e-06,
+      "loss": 0.4523,
+      "step": 359
+    },
+    {
+      "epoch": 2.208588957055215,
+      "grad_norm": 0.1486126184463501,
+      "learning_rate": 7.8332732279874e-06,
+      "loss": 0.4408,
+      "step": 360
+    },
+    {
+      "epoch": 2.214723926380368,
+      "grad_norm": 0.161222904920578,
+      "learning_rate": 7.82179844618867e-06,
+      "loss": 0.4455,
+      "step": 361
+    },
+    {
+      "epoch": 2.2208588957055215,
+      "grad_norm": 0.14370033144950867,
+      "learning_rate": 7.810301815087753e-06,
+      "loss": 0.4628,
+      "step": 362
+    },
+    {
+      "epoch": 2.226993865030675,
+      "grad_norm": 0.1402759701013565,
+      "learning_rate": 7.798783423703535e-06,
+      "loss": 0.4103,
+      "step": 363
+    },
+    {
+      "epoch": 2.233128834355828,
+      "grad_norm": 0.16180723905563354,
+      "learning_rate": 7.787243361223397e-06,
+      "loss": 0.4501,
+      "step": 364
+    },
+    {
+      "epoch": 2.2392638036809815,
+      "grad_norm": 0.14314652979373932,
+      "learning_rate": 7.775681717002523e-06,
+      "loss": 0.4504,
+      "step": 365
+    },
+    {
+      "epoch": 2.245398773006135,
+      "grad_norm": 0.14083176851272583,
+      "learning_rate": 7.764098580563203e-06,
+      "loss": 0.4505,
+      "step": 366
+    },
+    {
+      "epoch": 2.2515337423312882,
+      "grad_norm": 0.14769653975963593,
+      "learning_rate": 7.75249404159414e-06,
+      "loss": 0.4289,
+      "step": 367
+    },
+    {
+      "epoch": 2.2576687116564416,
+      "grad_norm": 0.17489399015903473,
+      "learning_rate": 7.740868189949762e-06,
+      "loss": 0.4586,
+      "step": 368
+    },
+    {
+      "epoch": 2.263803680981595,
+      "grad_norm": 0.15733091533184052,
+      "learning_rate": 7.729221115649516e-06,
+      "loss": 0.4499,
+      "step": 369
+    },
+    {
+      "epoch": 2.2699386503067487,
+      "grad_norm": 0.1377793848514557,
+      "learning_rate": 7.717552908877185e-06,
+      "loss": 0.441,
+      "step": 370
+    },
+    {
+      "epoch": 2.276073619631902,
+      "grad_norm": 0.13807876408100128,
+      "learning_rate": 7.705863659980175e-06,
+      "loss": 0.4457,
+      "step": 371
+    },
+    {
+      "epoch": 2.2822085889570554,
+      "grad_norm": 0.14367403090000153,
+      "learning_rate": 7.694153459468822e-06,
+      "loss": 0.4598,
+      "step": 372
+    },
+    {
+      "epoch": 2.2883435582822087,
+      "grad_norm": 0.1540713906288147,
+      "learning_rate": 7.682422398015696e-06,
+      "loss": 0.4124,
+      "step": 373
+    },
+    {
+      "epoch": 2.294478527607362,
+      "grad_norm": 0.15163351595401764,
+      "learning_rate": 7.67067056645489e-06,
+      "loss": 0.4608,
+      "step": 374
+    },
+    {
+      "epoch": 2.3006134969325154,
+      "grad_norm": 0.13935914635658264,
+      "learning_rate": 7.658898055781326e-06,
+      "loss": 0.4198,
+      "step": 375
+    },
+    {
+      "epoch": 2.3067484662576687,
+      "grad_norm": 0.15032264590263367,
+      "learning_rate": 7.647104957150037e-06,
+      "loss": 0.4414,
+      "step": 376
+    },
+    {
+      "epoch": 2.312883435582822,
+      "grad_norm": 0.16133847832679749,
+      "learning_rate": 7.635291361875474e-06,
+      "loss": 0.4228,
+      "step": 377
+    },
+    {
+      "epoch": 2.3190184049079754,
+      "grad_norm": 0.16057458519935608,
+      "learning_rate": 7.623457361430798e-06,
+      "loss": 0.4763,
+      "step": 378
+    },
+    {
+      "epoch": 2.3251533742331287,
+      "grad_norm": 0.1559583991765976,
+      "learning_rate": 7.611603047447161e-06,
+      "loss": 0.4191,
+      "step": 379
+    },
+    {
+      "epoch": 2.331288343558282,
+      "grad_norm": 0.16082173585891724,
+      "learning_rate": 7.5997285117130095e-06,
+      "loss": 0.4454,
+      "step": 380
+    },
+    {
+      "epoch": 2.3374233128834354,
+      "grad_norm": 0.14822718501091003,
+      "learning_rate": 7.587833846173363e-06,
+      "loss": 0.433,
+      "step": 381
+    },
+    {
+      "epoch": 2.3435582822085887,
+      "grad_norm": 0.16007304191589355,
+      "learning_rate": 7.57591914292911e-06,
+      "loss": 0.4177,
+      "step": 382
+    },
+    {
+      "epoch": 2.3496932515337425,
+      "grad_norm": 0.15621638298034668,
+      "learning_rate": 7.5639844942362915e-06,
+      "loss": 0.4429,
+      "step": 383
+    },
+    {
+      "epoch": 2.355828220858896,
+      "grad_norm": 0.1551215648651123,
+      "learning_rate": 7.552029992505385e-06,
+      "loss": 0.4545,
+      "step": 384
+    },
+    {
+      "epoch": 2.361963190184049,
+      "grad_norm": 0.15754824876785278,
+      "learning_rate": 7.540055730300595e-06,
+      "loss": 0.4292,
+      "step": 385
+    },
+    {
+      "epoch": 2.3680981595092025,
+      "grad_norm": 0.14232178032398224,
+      "learning_rate": 7.528061800339127e-06,
+      "loss": 0.4494,
+      "step": 386
+    },
+    {
+      "epoch": 2.374233128834356,
+      "grad_norm": 0.1486847698688507,
+      "learning_rate": 7.516048295490479e-06,
+      "loss": 0.4478,
+      "step": 387
+    },
+    {
+      "epoch": 2.3803680981595092,
+      "grad_norm": 0.1655314713716507,
+      "learning_rate": 7.504015308775714e-06,
+      "loss": 0.4309,
+      "step": 388
+    },
+    {
+      "epoch": 2.3865030674846626,
+      "grad_norm": 0.14602237939834595,
+      "learning_rate": 7.491962933366748e-06,
+      "loss": 0.4472,
+      "step": 389
+    },
+    {
+      "epoch": 2.392638036809816,
+      "grad_norm": 0.14621604979038239,
+      "learning_rate": 7.479891262585623e-06,
+      "loss": 0.4567,
+      "step": 390
+    },
+    {
+      "epoch": 2.3987730061349692,
+      "grad_norm": 0.1583264321088791,
+      "learning_rate": 7.467800389903786e-06,
+      "loss": 0.45,
+      "step": 391
+    },
+    {
+      "epoch": 2.4049079754601226,
+      "grad_norm": 0.17010322213172913,
+      "learning_rate": 7.455690408941363e-06,
+      "loss": 0.4426,
+      "step": 392
+    },
+    {
+      "epoch": 2.411042944785276,
+      "grad_norm": 0.16225053369998932,
+      "learning_rate": 7.443561413466439e-06,
+      "loss": 0.421,
+      "step": 393
+    },
+    {
+      "epoch": 2.4171779141104293,
+      "grad_norm": 0.20231634378433228,
+      "learning_rate": 7.431413497394328e-06,
+      "loss": 0.447,
+      "step": 394
+    },
+    {
+      "epoch": 2.4233128834355826,
+      "grad_norm": 0.1540592461824417,
+      "learning_rate": 7.419246754786847e-06,
+      "loss": 0.463,
+      "step": 395
+    },
+    {
+      "epoch": 2.4294478527607364,
+      "grad_norm": 0.15883290767669678,
+      "learning_rate": 7.407061279851589e-06,
+      "loss": 0.4636,
+      "step": 396
+    },
+    {
+      "epoch": 2.4355828220858897,
+      "grad_norm": 0.15072090923786163,
+      "learning_rate": 7.394857166941187e-06,
+      "loss": 0.4539,
+      "step": 397
+    },
+    {
+      "epoch": 2.441717791411043,
+      "grad_norm": 0.14263494312763214,
+      "learning_rate": 7.382634510552596e-06,
+      "loss": 0.44,
+      "step": 398
+    },
+    {
+      "epoch": 2.4478527607361964,
+      "grad_norm": 0.15457823872566223,
+      "learning_rate": 7.370393405326351e-06,
+      "loss": 0.4534,
+      "step": 399
+    },
+    {
+      "epoch": 2.4539877300613497,
+      "grad_norm": 0.1526721864938736,
+      "learning_rate": 7.358133946045834e-06,
+      "loss": 0.4551,
+      "step": 400
+    },
+    {
+      "epoch": 2.460122699386503,
+      "grad_norm": 0.14920799434185028,
+      "learning_rate": 7.345856227636548e-06,
+      "loss": 0.4593,
+      "step": 401
+    },
+    {
+      "epoch": 2.4662576687116564,
+      "grad_norm": 0.1798851191997528,
+      "learning_rate": 7.333560345165371e-06,
+      "loss": 0.4074,
+      "step": 402
+    },
+    {
+      "epoch": 2.4723926380368098,
+      "grad_norm": 0.1653115600347519,
+      "learning_rate": 7.321246393839836e-06,
+      "loss": 0.44,
+      "step": 403
+    },
+    {
+      "epoch": 2.478527607361963,
+      "grad_norm": 0.13865536451339722,
+      "learning_rate": 7.308914469007372e-06,
+      "loss": 0.4464,
+      "step": 404
+    },
+    {
+      "epoch": 2.4846625766871164,
+      "grad_norm": 0.16312679648399353,
+      "learning_rate": 7.296564666154589e-06,
+      "loss": 0.4617,
+      "step": 405
+    },
+    {
+      "epoch": 2.4907975460122698,
+      "grad_norm": 0.1640763282775879,
+      "learning_rate": 7.284197080906517e-06,
+      "loss": 0.4744,
+      "step": 406
+    },
+    {
+      "epoch": 2.4969325153374236,
+      "grad_norm": 0.17815761268138885,
+      "learning_rate": 7.271811809025882e-06,
+      "loss": 0.4404,
+      "step": 407
+    },
+    {
+      "epoch": 2.5030674846625764,
+      "grad_norm": 0.15390528738498688,
+      "learning_rate": 7.259408946412359e-06,
+      "loss": 0.4353,
+      "step": 408
+    },
+    {
+      "epoch": 2.5092024539877302,
+      "grad_norm": 0.15139776468276978,
+      "learning_rate": 7.246988589101825e-06,
+      "loss": 0.4763,
+      "step": 409
+    },
+    {
+      "epoch": 2.5153374233128836,
+      "grad_norm": 0.15306057035923004,
+      "learning_rate": 7.234550833265621e-06,
+      "loss": 0.4415,
+      "step": 410
+    },
+    {
+      "epoch": 2.521472392638037,
+      "grad_norm": 0.1511690318584442,
+      "learning_rate": 7.222095775209805e-06,
+      "loss": 0.4316,
+      "step": 411
+    },
+    {
+      "epoch": 2.5276073619631902,
+      "grad_norm": 0.15759505331516266,
+      "learning_rate": 7.209623511374407e-06,
+      "loss": 0.4321,
+      "step": 412
+    },
+    {
+      "epoch": 2.5337423312883436,
+      "grad_norm": 0.15876421332359314,
+      "learning_rate": 7.197134138332684e-06,
+      "loss": 0.4431,
+      "step": 413
+    },
+    {
+      "epoch": 2.539877300613497,
+      "grad_norm": 0.15190748870372772,
+      "learning_rate": 7.184627752790368e-06,
+      "loss": 0.4498,
+      "step": 414
+    },
+    {
+      "epoch": 2.5460122699386503,
+      "grad_norm": 0.1476791799068451,
+      "learning_rate": 7.1721044515849165e-06,
+      "loss": 0.4314,
+      "step": 415
+    },
+    {
+      "epoch": 2.5521472392638036,
+      "grad_norm": 0.1838875710964203,
+      "learning_rate": 7.159564331684774e-06,
+      "loss": 0.4257,
+      "step": 416
+    },
+    {
+      "epoch": 2.558282208588957,
+      "grad_norm": 0.15660858154296875,
+      "learning_rate": 7.1470074901886065e-06,
+      "loss": 0.4575,
+      "step": 417
+    },
+    {
+      "epoch": 2.5644171779141103,
+      "grad_norm": 0.15932102501392365,
+      "learning_rate": 7.134434024324557e-06,
+      "loss": 0.4338,
+      "step": 418
+    },
+    {
+      "epoch": 2.5705521472392636,
+      "grad_norm": 0.16551874577999115,
+      "learning_rate": 7.121844031449491e-06,
+      "loss": 0.4746,
+      "step": 419
+    },
+    {
+      "epoch": 2.5766871165644174,
+      "grad_norm": 0.16659101843833923,
+      "learning_rate": 7.109237609048247e-06,
+      "loss": 0.4983,
+      "step": 420
+    },
+    {
+      "epoch": 2.5828220858895703,
+      "grad_norm": 0.14108383655548096,
+      "learning_rate": 7.096614854732873e-06,
+      "loss": 0.4698,
+      "step": 421
+    },
+    {
+      "epoch": 2.588957055214724,
+      "grad_norm": 0.14742513000965118,
+      "learning_rate": 7.083975866241881e-06,
+      "loss": 0.4466,
+      "step": 422
+    },
+    {
+      "epoch": 2.5950920245398774,
+      "grad_norm": 0.15716223418712616,
+      "learning_rate": 7.071320741439481e-06,
+      "loss": 0.4671,
+      "step": 423
+    },
+    {
+      "epoch": 2.6012269938650308,
+      "grad_norm": 0.1822831928730011,
+      "learning_rate": 7.058649578314828e-06,
+      "loss": 0.4133,
+      "step": 424
+    },
+    {
+      "epoch": 2.607361963190184,
+      "grad_norm": 0.16716191172599792,
+      "learning_rate": 7.045962474981261e-06,
+      "loss": 0.4084,
+      "step": 425
+    },
+    {
+      "epoch": 2.6134969325153374,
+      "grad_norm": 0.1516677588224411,
+      "learning_rate": 7.03325952967555e-06,
+      "loss": 0.4342,
+      "step": 426
+    },
+    {
+      "epoch": 2.6196319018404908,
+      "grad_norm": 0.15387603640556335,
+      "learning_rate": 7.020540840757124e-06,
+      "loss": 0.4211,
+      "step": 427
+    },
+    {
+      "epoch": 2.625766871165644,
+      "grad_norm": 0.1543998122215271,
+      "learning_rate": 7.007806506707319e-06,
+      "loss": 0.4273,
+      "step": 428
+    },
+    {
+      "epoch": 2.6319018404907975,
+      "grad_norm": 0.15574420988559723,
+      "learning_rate": 6.995056626128609e-06,
+      "loss": 0.4458,
+      "step": 429
+    },
+    {
+      "epoch": 2.638036809815951,
+      "grad_norm": 0.1659959852695465,
+      "learning_rate": 6.982291297743848e-06,
+      "loss": 0.4592,
+      "step": 430
+    },
+    {
+      "epoch": 2.644171779141104,
+      "grad_norm": 0.18856996297836304,
+      "learning_rate": 6.969510620395503e-06,
+      "loss": 0.4552,
+      "step": 431
+    },
+    {
+      "epoch": 2.6503067484662575,
+      "grad_norm": 0.14291857182979584,
+      "learning_rate": 6.956714693044888e-06,
+      "loss": 0.4204,
+      "step": 432
+    },
+    {
+      "epoch": 2.6564417177914113,
+      "grad_norm": 0.17281392216682434,
+      "learning_rate": 6.943903614771397e-06,
+      "loss": 0.4571,
+      "step": 433
+    },
+    {
+      "epoch": 2.662576687116564,
+      "grad_norm": 0.16019487380981445,
+      "learning_rate": 6.931077484771739e-06,
+      "loss": 0.4563,
+      "step": 434
+    },
+    {
+      "epoch": 2.668711656441718,
+      "grad_norm": 0.16260547935962677,
+      "learning_rate": 6.9182364023591706e-06,
+      "loss": 0.425,
+      "step": 435
+    },
+    {
+      "epoch": 2.6748466257668713,
+      "grad_norm": 0.1593744456768036,
+      "learning_rate": 6.905380466962726e-06,
+      "loss": 0.4211,
+      "step": 436
+    },
+    {
+      "epoch": 2.6809815950920246,
+      "grad_norm": 0.16132207214832306,
+      "learning_rate": 6.892509778126442e-06,
+      "loss": 0.4377,
+      "step": 437
+    },
+    {
+      "epoch": 2.687116564417178,
+      "grad_norm": 0.15323421359062195,
+      "learning_rate": 6.879624435508596e-06,
+      "loss": 0.4284,
+      "step": 438
+    },
+    {
+      "epoch": 2.6932515337423313,
+      "grad_norm": 0.1697331666946411,
+      "learning_rate": 6.866724538880931e-06,
+      "loss": 0.4341,
+      "step": 439
+    },
+    {
+      "epoch": 2.6993865030674846,
+      "grad_norm": 0.1691737174987793,
+      "learning_rate": 6.85381018812788e-06,
+      "loss": 0.4907,
+      "step": 440
+    },
+    {
+      "epoch": 2.705521472392638,
+      "grad_norm": 0.1564476490020752,
+      "learning_rate": 6.840881483245797e-06,
+      "loss": 0.4133,
+      "step": 441
+    },
+    {
+      "epoch": 2.7116564417177913,
+      "grad_norm": 0.15074561536312103,
+      "learning_rate": 6.827938524342175e-06,
+      "loss": 0.4424,
+      "step": 442
+    },
+    {
+      "epoch": 2.7177914110429446,
+      "grad_norm": 0.15723863244056702,
+      "learning_rate": 6.814981411634885e-06,
+      "loss": 0.4469,
+      "step": 443
+    },
+    {
+      "epoch": 2.7239263803680984,
+      "grad_norm": 0.15343016386032104,
+      "learning_rate": 6.802010245451382e-06,
+      "loss": 0.4495,
+      "step": 444
+    },
+    {
+      "epoch": 2.7300613496932513,
+      "grad_norm": 0.14901430904865265,
+      "learning_rate": 6.789025126227948e-06,
+      "loss": 0.4332,
+      "step": 445
+    },
+    {
+      "epoch": 2.736196319018405,
+      "grad_norm": 0.18850690126419067,
+      "learning_rate": 6.7760261545088955e-06,
+      "loss": 0.4741,
+      "step": 446
+    },
+    {
+      "epoch": 2.7423312883435584,
+      "grad_norm": 0.18473871052265167,
+      "learning_rate": 6.763013430945803e-06,
+      "loss": 0.4579,
+      "step": 447
+    },
+    {
+      "epoch": 2.7484662576687118,
+      "grad_norm": 0.16771845519542694,
+      "learning_rate": 6.749987056296728e-06,
+      "loss": 0.4281,
+      "step": 448
+    },
+    {
+      "epoch": 2.754601226993865,
+      "grad_norm": 0.15645870566368103,
+      "learning_rate": 6.736947131425423e-06,
+      "loss": 0.4399,
+      "step": 449
+    },
+    {
+      "epoch": 2.7607361963190185,
+      "grad_norm": 0.15851353108882904,
+      "learning_rate": 6.723893757300572e-06,
+      "loss": 0.4245,
+      "step": 450
+    },
+    {
+      "epoch": 2.766871165644172,
+      "grad_norm": 0.15412819385528564,
+      "learning_rate": 6.710827034994991e-06,
+      "loss": 0.457,
+      "step": 451
+    },
+    {
+      "epoch": 2.773006134969325,
+      "grad_norm": 0.21733322739601135,
+      "learning_rate": 6.697747065684851e-06,
+      "loss": 0.4235,
+      "step": 452
+    },
+    {
+      "epoch": 2.7791411042944785,
+      "grad_norm": 0.1558465212583542,
+      "learning_rate": 6.684653950648893e-06,
+      "loss": 0.3966,
+      "step": 453
+    },
+    {
+      "epoch": 2.785276073619632,
+      "grad_norm": 0.1692306250333786,
+      "learning_rate": 6.671547791267652e-06,
+      "loss": 0.469,
+      "step": 454
+    },
+    {
+      "epoch": 2.791411042944785,
+      "grad_norm": 0.1569252610206604,
+      "learning_rate": 6.658428689022661e-06,
+      "loss": 0.4432,
+      "step": 455
+    },
+    {
+      "epoch": 2.7975460122699385,
+      "grad_norm": 0.17490547895431519,
+      "learning_rate": 6.6452967454956744e-06,
+      "loss": 0.454,
+      "step": 456
+    },
+    {
+      "epoch": 2.8036809815950923,
+      "grad_norm": 0.16125169396400452,
+      "learning_rate": 6.632152062367871e-06,
+      "loss": 0.4699,
+      "step": 457
+    },
+    {
+      "epoch": 2.809815950920245,
+      "grad_norm": 0.16368430852890015,
+      "learning_rate": 6.618994741419078e-06,
+      "loss": 0.4194,
+      "step": 458
+    },
+    {
+      "epoch": 2.815950920245399,
+      "grad_norm": 0.1626632660627365,
+      "learning_rate": 6.605824884526978e-06,
+      "loss": 0.4289,
+      "step": 459
+    },
+    {
+      "epoch": 2.8220858895705523,
+      "grad_norm": 0.17192265391349792,
+      "learning_rate": 6.592642593666316e-06,
+      "loss": 0.4629,
+      "step": 460
+    },
+    {
+      "epoch": 2.8282208588957056,
+      "grad_norm": 0.15331393480300903,
+      "learning_rate": 6.579447970908115e-06,
+      "loss": 0.4261,
+      "step": 461
+    },
+    {
+      "epoch": 2.834355828220859,
+      "grad_norm": 0.16751109063625336,
+      "learning_rate": 6.566241118418888e-06,
+      "loss": 0.4398,
+      "step": 462
+    },
+    {
+      "epoch": 2.8404907975460123,
+      "grad_norm": 0.20704393088817596,
+      "learning_rate": 6.553022138459839e-06,
+      "loss": 0.4196,
+      "step": 463
+    },
+    {
+      "epoch": 2.8466257668711656,
+      "grad_norm": 0.17874981462955475,
+      "learning_rate": 6.539791133386077e-06,
+      "loss": 0.4542,
+      "step": 464
+    },
+    {
+      "epoch": 2.852760736196319,
+      "grad_norm": 0.1628667265176773,
+      "learning_rate": 6.526548205645823e-06,
+      "loss": 0.4278,
+      "step": 465
+    },
+    {
+      "epoch": 2.8588957055214723,
+      "grad_norm": 0.18325841426849365,
+      "learning_rate": 6.513293457779614e-06,
+      "loss": 0.4448,
+      "step": 466
+    },
+    {
+      "epoch": 2.8650306748466257,
+      "grad_norm": 0.17128658294677734,
+      "learning_rate": 6.50002699241951e-06,
+      "loss": 0.4778,
+      "step": 467
+    },
+    {
+      "epoch": 2.871165644171779,
+      "grad_norm": 0.18761861324310303,
+      "learning_rate": 6.486748912288305e-06,
+      "loss": 0.4313,
+      "step": 468
+    },
+    {
+      "epoch": 2.8773006134969323,
+      "grad_norm": 0.16829295456409454,
+      "learning_rate": 6.4734593201987205e-06,
+      "loss": 0.4589,
+      "step": 469
+    },
+    {
+      "epoch": 2.883435582822086,
+      "grad_norm": 0.16941910982131958,
+      "learning_rate": 6.46015831905262e-06,
+      "loss": 0.4844,
+      "step": 470
+    },
+    {
+      "epoch": 2.889570552147239,
+      "grad_norm": 0.18299958109855652,
+      "learning_rate": 6.446846011840204e-06,
+      "loss": 0.4204,
+      "step": 471
+    },
+    {
+      "epoch": 2.895705521472393,
+      "grad_norm": 0.1755492389202118,
+      "learning_rate": 6.43352250163922e-06,
+      "loss": 0.4265,
+      "step": 472
+    },
+    {
+      "epoch": 2.901840490797546,
+      "grad_norm": 0.17342810332775116,
+      "learning_rate": 6.420187891614158e-06,
+      "loss": 0.4314,
+      "step": 473
+    },
+    {
+      "epoch": 2.9079754601226995,
+      "grad_norm": 0.15891548991203308,
+      "learning_rate": 6.406842285015455e-06,
+      "loss": 0.4188,
+      "step": 474
+    },
+    {
+      "epoch": 2.914110429447853,
+      "grad_norm": 0.17201867699623108,
+      "learning_rate": 6.393485785178699e-06,
+      "loss": 0.4459,
+      "step": 475
+    },
+    {
+      "epoch": 2.920245398773006,
+      "grad_norm": 0.16666850447654724,
+      "learning_rate": 6.380118495523816e-06,
+      "loss": 0.437,
+      "step": 476
+    },
+    {
+      "epoch": 2.9263803680981595,
+      "grad_norm": 0.17189852893352509,
+      "learning_rate": 6.366740519554286e-06,
+      "loss": 0.4263,
+      "step": 477
+    },
+    {
+      "epoch": 2.932515337423313,
+      "grad_norm": 0.16016732156276703,
+      "learning_rate": 6.353351960856332e-06,
+      "loss": 0.4358,
+      "step": 478
+    },
+    {
+      "epoch": 2.938650306748466,
+      "grad_norm": 0.15281671285629272,
+      "learning_rate": 6.339952923098117e-06,
+      "loss": 0.4314,
+      "step": 479
+    },
+    {
+      "epoch": 2.9447852760736195,
+      "grad_norm": 0.19255416095256805,
+      "learning_rate": 6.326543510028943e-06,
+      "loss": 0.42,
+      "step": 480
+    },
+    {
+      "epoch": 2.950920245398773,
+      "grad_norm": 0.1635408252477646,
+      "learning_rate": 6.3131238254784534e-06,
+      "loss": 0.4247,
+      "step": 481
+    },
+    {
+      "epoch": 2.957055214723926,
+      "grad_norm": 0.16226807236671448,
+      "learning_rate": 6.299693973355821e-06,
+      "loss": 0.4348,
+      "step": 482
+    },
+    {
+      "epoch": 2.96319018404908,
+      "grad_norm": 0.17341448366641998,
+      "learning_rate": 6.286254057648945e-06,
+      "loss": 0.4174,
+      "step": 483
+    },
+    {
+      "epoch": 2.969325153374233,
+      "grad_norm": 0.16853678226470947,
+      "learning_rate": 6.27280418242365e-06,
+      "loss": 0.4529,
+      "step": 484
+    },
+    {
+      "epoch": 2.9754601226993866,
+      "grad_norm": 0.16225695610046387,
+      "learning_rate": 6.259344451822877e-06,
+      "loss": 0.4685,
+      "step": 485
+    },
+    {
+      "epoch": 2.98159509202454,
+      "grad_norm": 0.16942046582698822,
+      "learning_rate": 6.245874970065877e-06,
+      "loss": 0.4335,
+      "step": 486
+    },
+    {
+      "epoch": 2.9877300613496933,
+      "grad_norm": 0.1746375411748886,
+      "learning_rate": 6.2323958414474065e-06,
+      "loss": 0.4304,
+      "step": 487
+    },
+    {
+      "epoch": 2.9938650306748467,
+      "grad_norm": 0.19197514653205872,
+      "learning_rate": 6.218907170336912e-06,
+      "loss": 0.4651,
+      "step": 488
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.16628479957580566,
+      "learning_rate": 6.2054090611777385e-06,
+      "loss": 0.4484,
+      "step": 489
+    },
+    {
+      "epoch": 3.0061349693251533,
+      "grad_norm": 0.1807141751050949,
+      "learning_rate": 6.191901618486299e-06,
+      "loss": 0.4403,
+      "step": 490
+    },
+    {
+      "epoch": 3.0122699386503067,
+      "grad_norm": 0.18243259191513062,
+      "learning_rate": 6.178384946851284e-06,
+      "loss": 0.4427,
+      "step": 491
+    },
+    {
+      "epoch": 3.01840490797546,
+      "grad_norm": 0.17769359052181244,
+      "learning_rate": 6.164859150932839e-06,
+      "loss": 0.4177,
+      "step": 492
+    },
+    {
+      "epoch": 3.0245398773006134,
+      "grad_norm": 0.17457637190818787,
+      "learning_rate": 6.151324335461766e-06,
+      "loss": 0.4623,
+      "step": 493
+    },
+    {
+      "epoch": 3.0306748466257667,
+      "grad_norm": 0.1725994348526001,
+      "learning_rate": 6.137780605238698e-06,
+      "loss": 0.4638,
+      "step": 494
+    },
+    {
+      "epoch": 3.03680981595092,
+      "grad_norm": 0.17166265845298767,
+      "learning_rate": 6.1242280651332995e-06,
+      "loss": 0.446,
+      "step": 495
+    },
+    {
+      "epoch": 3.042944785276074,
+      "grad_norm": 0.17501436173915863,
+      "learning_rate": 6.11066682008345e-06,
+      "loss": 0.4405,
+      "step": 496
+    },
+    {
+      "epoch": 3.049079754601227,
+      "grad_norm": 0.1676449030637741,
+      "learning_rate": 6.097096975094432e-06,
+      "loss": 0.4718,
+      "step": 497
+    },
+    {
+      "epoch": 3.0552147239263805,
+      "grad_norm": 0.1620589643716812,
+      "learning_rate": 6.083518635238117e-06,
+      "loss": 0.4353,
+      "step": 498
+    },
+    {
+      "epoch": 3.061349693251534,
+      "grad_norm": 0.17251572012901306,
+      "learning_rate": 6.069931905652151e-06,
+      "loss": 0.4409,
+      "step": 499
+    },
+    {
+      "epoch": 3.067484662576687,
+      "grad_norm": 0.1733006089925766,
+      "learning_rate": 6.056336891539144e-06,
+      "loss": 0.4323,
+      "step": 500
+    },
+    {
+      "epoch": 3.0736196319018405,
+      "grad_norm": 0.1714855581521988,
+      "learning_rate": 6.042733698165855e-06,
+      "loss": 0.4297,
+      "step": 501
+    },
+    {
+      "epoch": 3.079754601226994,
+      "grad_norm": 0.15556511282920837,
+      "learning_rate": 6.029122430862373e-06,
+      "loss": 0.4546,
+      "step": 502
+    },
+    {
+      "epoch": 3.085889570552147,
+      "grad_norm": 0.18687503039836884,
+      "learning_rate": 6.015503195021303e-06,
+      "loss": 0.4586,
+      "step": 503
+    },
+    {
+      "epoch": 3.0920245398773005,
+      "grad_norm": 0.1955818086862564,
+      "learning_rate": 6.001876096096951e-06,
+      "loss": 0.4405,
+      "step": 504
+    },
+    {
+      "epoch": 3.098159509202454,
+      "grad_norm": 0.18032221496105194,
+      "learning_rate": 5.988241239604511e-06,
+      "loss": 0.4621,
+      "step": 505
+    },
+    {
+      "epoch": 3.104294478527607,
+      "grad_norm": 0.1692107915878296,
+      "learning_rate": 5.97459873111924e-06,
+      "loss": 0.4557,
+      "step": 506
+    },
+    {
+      "epoch": 3.1104294478527605,
+      "grad_norm": 0.18335744738578796,
+      "learning_rate": 5.9609486762756465e-06,
+      "loss": 0.43,
+      "step": 507
+    },
+    {
+      "epoch": 3.116564417177914,
+      "grad_norm": 0.16377264261245728,
+      "learning_rate": 5.947291180766668e-06,
+      "loss": 0.4473,
+      "step": 508
+    },
+    {
+      "epoch": 3.1226993865030677,
+      "grad_norm": 0.1665698140859604,
+      "learning_rate": 5.933626350342858e-06,
+      "loss": 0.4262,
+      "step": 509
+    },
+    {
+      "epoch": 3.128834355828221,
+      "grad_norm": 0.1706293821334839,
+      "learning_rate": 5.9199542908115694e-06,
+      "loss": 0.4118,
+      "step": 510
+    },
+    {
+      "epoch": 3.1349693251533743,
+      "grad_norm": 0.17646320164203644,
+      "learning_rate": 5.906275108036119e-06,
+      "loss": 0.4503,
+      "step": 511
+    },
+    {
+      "epoch": 3.1411042944785277,
+      "grad_norm": 0.18442606925964355,
+      "learning_rate": 5.892588907934988e-06,
+      "loss": 0.448,
+      "step": 512
+    },
+    {
+      "epoch": 3.147239263803681,
+      "grad_norm": 0.16845177114009857,
+      "learning_rate": 5.87889579648099e-06,
+      "loss": 0.4533,
+      "step": 513
+    },
+    {
+      "epoch": 3.1533742331288344,
+      "grad_norm": 0.17161841690540314,
+      "learning_rate": 5.865195879700454e-06,
+      "loss": 0.4358,
+      "step": 514
+    },
+    {
+      "epoch": 3.1595092024539877,
+      "grad_norm": 0.1597752720117569,
+      "learning_rate": 5.8514892636724005e-06,
+      "loss": 0.4444,
+      "step": 515
+    },
+    {
+      "epoch": 3.165644171779141,
+      "grad_norm": 0.1626145988702774,
+      "learning_rate": 5.83777605452773e-06,
+      "loss": 0.4317,
+      "step": 516
+    },
+    {
+      "epoch": 3.1717791411042944,
+      "grad_norm": 0.17902852594852448,
+      "learning_rate": 5.8240563584483855e-06,
+      "loss": 0.4604,
+      "step": 517
+    },
+    {
+      "epoch": 3.1779141104294477,
+      "grad_norm": 0.1681181639432907,
+      "learning_rate": 5.810330281666542e-06,
+      "loss": 0.4577,
+      "step": 518
+    },
+    {
+      "epoch": 3.184049079754601,
+      "grad_norm": 0.18032066524028778,
+      "learning_rate": 5.796597930463776e-06,
+      "loss": 0.4214,
+      "step": 519
+    },
+    {
+      "epoch": 3.190184049079755,
+      "grad_norm": 0.18553897738456726,
+      "learning_rate": 5.782859411170261e-06,
+      "loss": 0.3931,
+      "step": 520
+    },
+    {
+      "epoch": 3.196319018404908,
+      "grad_norm": 0.16626308858394623,
+      "learning_rate": 5.769114830163913e-06,
+      "loss": 0.4307,
+      "step": 521
+    },
+    {
+      "epoch": 3.2024539877300615,
+      "grad_norm": 0.17122527956962585,
+      "learning_rate": 5.7553642938695945e-06,
+      "loss": 0.4582,
+      "step": 522
+    },
+    {
+      "epoch": 3.208588957055215,
+      "grad_norm": 0.1699482649564743,
+      "learning_rate": 5.741607908758275e-06,
+      "loss": 0.4575,
+      "step": 523
+    },
+    {
+      "epoch": 3.214723926380368,
+      "grad_norm": 0.185217946767807,
+      "learning_rate": 5.727845781346217e-06,
+      "loss": 0.4063,
+      "step": 524
+    },
+    {
+      "epoch": 3.2208588957055215,
+      "grad_norm": 0.17612098157405853,
+      "learning_rate": 5.714078018194141e-06,
+      "loss": 0.4555,
+      "step": 525
+    },
+    {
+      "epoch": 3.226993865030675,
+      "grad_norm": 0.17292365431785583,
+      "learning_rate": 5.7003047259064095e-06,
+      "loss": 0.4402,
+      "step": 526
+    },
+    {
+      "epoch": 3.233128834355828,
+      "grad_norm": 0.18183457851409912,
+      "learning_rate": 5.68652601113019e-06,
+      "loss": 0.4694,
+      "step": 527
+    },
+    {
+      "epoch": 3.2392638036809815,
+      "grad_norm": 0.1628892868757248,
+      "learning_rate": 5.672741980554646e-06,
+      "loss": 0.4472,
+      "step": 528
+    },
+    {
+      "epoch": 3.245398773006135,
+      "grad_norm": 0.17642362415790558,
+      "learning_rate": 5.658952740910094e-06,
+      "loss": 0.4239,
+      "step": 529
+    },
+    {
+      "epoch": 3.2515337423312882,
+      "grad_norm": 0.1806386113166809,
+      "learning_rate": 5.645158398967191e-06,
+      "loss": 0.4282,
+      "step": 530
+    },
+    {
+      "epoch": 3.2576687116564416,
+      "grad_norm": 0.1763850599527359,
+      "learning_rate": 5.6313590615360935e-06,
+      "loss": 0.4193,
+      "step": 531
+    },
+    {
+      "epoch": 3.263803680981595,
+      "grad_norm": 0.16797596216201782,
+      "learning_rate": 5.617554835465646e-06,
+      "loss": 0.4358,
+      "step": 532
+    },
+    {
+      "epoch": 3.2699386503067487,
+      "grad_norm": 0.16996163129806519,
+      "learning_rate": 5.6037458276425394e-06,
+      "loss": 0.4048,
+      "step": 533
+    },
+    {
+      "epoch": 3.276073619631902,
+      "grad_norm": 0.16143372654914856,
+      "learning_rate": 5.589932144990495e-06,
+      "loss": 0.4421,
+      "step": 534
+    },
+    {
+      "epoch": 3.2822085889570554,
+      "grad_norm": 0.16983363032341003,
+      "learning_rate": 5.5761138944694295e-06,
+      "loss": 0.4253,
+      "step": 535
+    },
+    {
+      "epoch": 3.2883435582822087,
+      "grad_norm": 0.17256614565849304,
+      "learning_rate": 5.562291183074627e-06,
+      "loss": 0.4216,
+      "step": 536
+    },
+    {
+      "epoch": 3.294478527607362,
+      "grad_norm": 0.18795952200889587,
+      "learning_rate": 5.548464117835917e-06,
+      "loss": 0.4246,
+      "step": 537
+    },
+    {
+      "epoch": 3.3006134969325154,
+      "grad_norm": 0.18537187576293945,
+      "learning_rate": 5.534632805816835e-06,
+      "loss": 0.4124,
+      "step": 538
+    },
+    {
+      "epoch": 3.3067484662576687,
+      "grad_norm": 0.19903261959552765,
+      "learning_rate": 5.520797354113804e-06,
+      "loss": 0.46,
+      "step": 539
+    },
+    {
+      "epoch": 3.312883435582822,
+      "grad_norm": 0.1933709681034088,
+      "learning_rate": 5.5069578698553e-06,
+      "loss": 0.431,
+      "step": 540
+    },
+    {
+      "epoch": 3.3190184049079754,
+      "grad_norm": 0.17167270183563232,
+      "learning_rate": 5.4931144602010224e-06,
+      "loss": 0.4773,
+      "step": 541
+    },
+    {
+      "epoch": 3.3251533742331287,
+      "grad_norm": 0.16057482361793518,
+      "learning_rate": 5.479267232341064e-06,
+      "loss": 0.4089,
+      "step": 542
+    },
+    {
+      "epoch": 3.331288343558282,
+      "grad_norm": 0.17433694005012512,
+      "learning_rate": 5.465416293495083e-06,
+      "loss": 0.4502,
+      "step": 543
+    },
+    {
+      "epoch": 3.3374233128834354,
+      "grad_norm": 0.1779199242591858,
+      "learning_rate": 5.451561750911475e-06,
+      "loss": 0.4048,
+      "step": 544
+    },
+    {
+      "epoch": 3.3435582822085887,
+      "grad_norm": 0.17335927486419678,
+      "learning_rate": 5.437703711866534e-06,
+      "loss": 0.4341,
+      "step": 545
+    },
+    {
+      "epoch": 3.3496932515337425,
+      "grad_norm": 0.17468775808811188,
+      "learning_rate": 5.4238422836636315e-06,
+      "loss": 0.4448,
+      "step": 546
+    },
+    {
+      "epoch": 3.355828220858896,
+      "grad_norm": 0.18919560313224792,
+      "learning_rate": 5.40997757363238e-06,
+      "loss": 0.4489,
+      "step": 547
+    },
+    {
+      "epoch": 3.361963190184049,
+      "grad_norm": 0.17396590113639832,
+      "learning_rate": 5.3961096891278035e-06,
+      "loss": 0.4325,
+      "step": 548
+    },
+    {
+      "epoch": 3.3680981595092025,
+      "grad_norm": 0.16672684252262115,
+      "learning_rate": 5.382238737529505e-06,
+      "loss": 0.4077,
+      "step": 549
+    },
+    {
+      "epoch": 3.374233128834356,
+      "grad_norm": 0.1844366043806076,
+      "learning_rate": 5.368364826240836e-06,
+      "loss": 0.4196,
+      "step": 550
+    },
+    {
+      "epoch": 3.3803680981595092,
+      "grad_norm": 0.18350937962532043,
+      "learning_rate": 5.354488062688068e-06,
+      "loss": 0.4523,
+      "step": 551
+    },
+    {
+      "epoch": 3.3865030674846626,
+      "grad_norm": 0.26154661178588867,
+      "learning_rate": 5.3406085543195555e-06,
+      "loss": 0.4761,
+      "step": 552
+    },
+    {
+      "epoch": 3.392638036809816,
+      "grad_norm": 0.1603960394859314,
+      "learning_rate": 5.3267264086049054e-06,
+      "loss": 0.4433,
+      "step": 553
+    },
+    {
+      "epoch": 3.3987730061349692,
+      "grad_norm": 0.19381873309612274,
+      "learning_rate": 5.312841733034147e-06,
+      "loss": 0.4031,
+      "step": 554
+    },
+    {
+      "epoch": 3.4049079754601226,
+      "grad_norm": 0.21427270770072937,
+      "learning_rate": 5.2989546351168985e-06,
+      "loss": 0.4391,
+      "step": 555
+    },
+    {
+      "epoch": 3.411042944785276,
+      "grad_norm": 0.1873149424791336,
+      "learning_rate": 5.285065222381533e-06,
+      "loss": 0.4409,
+      "step": 556
+    },
+    {
+      "epoch": 3.4171779141104293,
+      "grad_norm": 0.1799328476190567,
+      "learning_rate": 5.27117360237435e-06,
+      "loss": 0.4381,
+      "step": 557
+    },
+    {
+      "epoch": 3.4233128834355826,
+      "grad_norm": 0.1705305278301239,
+      "learning_rate": 5.257279882658737e-06,
+      "loss": 0.3905,
+      "step": 558
+    },
+    {
+      "epoch": 3.4294478527607364,
+      "grad_norm": 0.1969294250011444,
+      "learning_rate": 5.2433841708143405e-06,
+      "loss": 0.4264,
+      "step": 559
+    },
+    {
+      "epoch": 3.4355828220858897,
+      "grad_norm": 0.17147500813007355,
+      "learning_rate": 5.229486574436236e-06,
+      "loss": 0.4592,
+      "step": 560
+    },
+    {
+      "epoch": 3.441717791411043,
+      "grad_norm": 0.1796891838312149,
+      "learning_rate": 5.215587201134081e-06,
+      "loss": 0.4223,
+      "step": 561
+    },
+    {
+      "epoch": 3.4478527607361964,
+      "grad_norm": 0.17482995986938477,
+      "learning_rate": 5.201686158531304e-06,
+      "loss": 0.4143,
+      "step": 562
+    },
+    {
+      "epoch": 3.4539877300613497,
+      "grad_norm": 0.18087799847126007,
+      "learning_rate": 5.187783554264253e-06,
+      "loss": 0.4611,
+      "step": 563
+    },
+    {
+      "epoch": 3.460122699386503,
+      "grad_norm": 0.1888713240623474,
+      "learning_rate": 5.173879495981367e-06,
+      "loss": 0.4569,
+      "step": 564
+    },
+    {
+      "epoch": 3.4662576687116564,
+      "grad_norm": 0.20657221972942352,
+      "learning_rate": 5.1599740913423435e-06,
+      "loss": 0.4762,
+      "step": 565
+    },
+    {
+      "epoch": 3.4723926380368098,
+      "grad_norm": 0.18362171947956085,
+      "learning_rate": 5.146067448017308e-06,
+      "loss": 0.4268,
+      "step": 566
+    },
+    {
+      "epoch": 3.478527607361963,
+      "grad_norm": 0.18593919277191162,
+      "learning_rate": 5.132159673685976e-06,
+      "loss": 0.4506,
+      "step": 567
+    },
+    {
+      "epoch": 3.4846625766871164,
+      "grad_norm": 0.1854875087738037,
+      "learning_rate": 5.1182508760368195e-06,
+      "loss": 0.4209,
+      "step": 568
+    },
+    {
+      "epoch": 3.4907975460122698,
+      "grad_norm": 0.18463847041130066,
+      "learning_rate": 5.104341162766234e-06,
+      "loss": 0.4349,
+      "step": 569
+    },
+    {
+      "epoch": 3.4969325153374236,
+      "grad_norm": 0.18153201043605804,
+      "learning_rate": 5.090430641577705e-06,
+      "loss": 0.459,
+      "step": 570
+    },
+    {
+      "epoch": 3.5030674846625764,
+      "grad_norm": 0.20472465455532074,
+      "learning_rate": 5.0765194201809755e-06,
+      "loss": 0.4294,
+      "step": 571
+    },
+    {
+      "epoch": 3.5092024539877302,
+      "grad_norm": 0.16046775877475739,
+      "learning_rate": 5.062607606291208e-06,
+      "loss": 0.4442,
+      "step": 572
+    },
+    {
+      "epoch": 3.5153374233128836,
+      "grad_norm": 0.17173658311367035,
+      "learning_rate": 5.048695307628152e-06,
+      "loss": 0.4155,
+      "step": 573
+    },
+    {
+      "epoch": 3.521472392638037,
+      "grad_norm": 0.18764469027519226,
+      "learning_rate": 5.034782631915314e-06,
+      "loss": 0.4081,
+      "step": 574
+    },
+    {
+      "epoch": 3.5276073619631902,
+      "grad_norm": 0.18143120408058167,
+      "learning_rate": 5.020869686879115e-06,
+      "loss": 0.4273,
+      "step": 575
+    },
+    {
+      "epoch": 3.5337423312883436,
+      "grad_norm": 0.16679701209068298,
+      "learning_rate": 5.006956580248069e-06,
+      "loss": 0.445,
+      "step": 576
+    },
+    {
+      "epoch": 3.539877300613497,
+      "grad_norm": 0.1785566657781601,
+      "learning_rate": 4.993043419751933e-06,
+      "loss": 0.4325,
+      "step": 577
+    },
+    {
+      "epoch": 3.5460122699386503,
+      "grad_norm": 0.18561996519565582,
+      "learning_rate": 4.979130313120885e-06,
+      "loss": 0.4432,
+      "step": 578
+    },
+    {
+      "epoch": 3.5521472392638036,
+      "grad_norm": 0.17507800459861755,
+      "learning_rate": 4.965217368084688e-06,
+      "loss": 0.4358,
+      "step": 579
+    },
+    {
+      "epoch": 3.558282208588957,
+      "grad_norm": 0.17406027019023895,
+      "learning_rate": 4.95130469237185e-06,
+      "loss": 0.4551,
+      "step": 580
+    },
+    {
+      "epoch": 3.5644171779141103,
+      "grad_norm": 0.18630944192409515,
+      "learning_rate": 4.937392393708794e-06,
+      "loss": 0.4398,
+      "step": 581
+    },
+    {
+      "epoch": 3.5705521472392636,
+      "grad_norm": 0.15710729360580444,
+      "learning_rate": 4.923480579819025e-06,
+      "loss": 0.428,
+      "step": 582
+    },
+    {
+      "epoch": 3.5766871165644174,
+      "grad_norm": 0.17362196743488312,
+      "learning_rate": 4.909569358422296e-06,
+      "loss": 0.4388,
+      "step": 583
+    },
+    {
+      "epoch": 3.5828220858895703,
+      "grad_norm": 0.1887211948633194,
+      "learning_rate": 4.895658837233767e-06,
+      "loss": 0.4241,
+      "step": 584
+    },
+    {
+      "epoch": 3.588957055214724,
+      "grad_norm": 0.18522702157497406,
+      "learning_rate": 4.881749123963183e-06,
+      "loss": 0.443,
+      "step": 585
+    },
+    {
+      "epoch": 3.5950920245398774,
+      "grad_norm": 0.17625677585601807,
+      "learning_rate": 4.867840326314024e-06,
+      "loss": 0.4345,
+      "step": 586
+    },
+    {
+      "epoch": 3.6012269938650308,
+      "grad_norm": 0.17641369998455048,
+      "learning_rate": 4.853932551982692e-06,
+      "loss": 0.4018,
+      "step": 587
+    },
+    {
+      "epoch": 3.607361963190184,
+      "grad_norm": 0.18613062798976898,
+      "learning_rate": 4.840025908657658e-06,
+      "loss": 0.4787,
+      "step": 588
+    },
+    {
+      "epoch": 3.6134969325153374,
+      "grad_norm": 0.17531706392765045,
+      "learning_rate": 4.826120504018635e-06,
+      "loss": 0.4197,
+      "step": 589
+    },
+    {
+      "epoch": 3.6196319018404908,
+      "grad_norm": 0.18549470603466034,
+      "learning_rate": 4.812216445735749e-06,
+      "loss": 0.4228,
+      "step": 590
+    },
+    {
+      "epoch": 3.625766871165644,
+      "grad_norm": 0.17408756911754608,
+      "learning_rate": 4.798313841468697e-06,
+      "loss": 0.4054,
+      "step": 591
+    },
+    {
+      "epoch": 3.6319018404907975,
+      "grad_norm": 0.19779986143112183,
+      "learning_rate": 4.7844127988659204e-06,
+      "loss": 0.4277,
+      "step": 592
+    },
+    {
+      "epoch": 3.638036809815951,
+      "grad_norm": 0.16971039772033691,
+      "learning_rate": 4.7705134255637676e-06,
+      "loss": 0.4098,
+      "step": 593
+    },
+    {
+      "epoch": 3.644171779141104,
+      "grad_norm": 0.18103785812854767,
+      "learning_rate": 4.756615829185661e-06,
+      "loss": 0.4254,
+      "step": 594
+    },
+    {
+      "epoch": 3.6503067484662575,
+      "grad_norm": 0.18030977249145508,
+      "learning_rate": 4.742720117341265e-06,
+      "loss": 0.4203,
+      "step": 595
+    },
+    {
+      "epoch": 3.6564417177914113,
+      "grad_norm": 0.17600463330745697,
+      "learning_rate": 4.728826397625651e-06,
+      "loss": 0.429,
+      "step": 596
+    },
+    {
+      "epoch": 3.662576687116564,
+      "grad_norm": 0.18706922233104706,
+      "learning_rate": 4.714934777618468e-06,
+      "loss": 0.4562,
+      "step": 597
+    },
+    {
+      "epoch": 3.668711656441718,
+      "grad_norm": 0.18239618837833405,
+      "learning_rate": 4.701045364883103e-06,
+      "loss": 0.4327,
+      "step": 598
+    },
+    {
+      "epoch": 3.6748466257668713,
+      "grad_norm": 0.20267638564109802,
+      "learning_rate": 4.6871582669658545e-06,
+      "loss": 0.4167,
+      "step": 599
+    },
+    {
+      "epoch": 3.6809815950920246,
+      "grad_norm": 0.1839340478181839,
+      "learning_rate": 4.673273591395095e-06,
+      "loss": 0.4224,
+      "step": 600
+    },
+    {
+      "epoch": 3.687116564417178,
+      "grad_norm": 0.18033233284950256,
+      "learning_rate": 4.659391445680446e-06,
+      "loss": 0.4043,
+      "step": 601
+    },
+    {
+      "epoch": 3.6932515337423313,
+      "grad_norm": 0.20073480904102325,
+      "learning_rate": 4.645511937311934e-06,
+      "loss": 0.4511,
+      "step": 602
+    },
+    {
+      "epoch": 3.6993865030674846,
+      "grad_norm": 0.20318840444087982,
+      "learning_rate": 4.631635173759165e-06,
+      "loss": 0.4279,
+      "step": 603
+    },
+    {
+      "epoch": 3.705521472392638,
+      "grad_norm": 0.17758074402809143,
+      "learning_rate": 4.6177612624704975e-06,
+      "loss": 0.4735,
+      "step": 604
+    },
+    {
+      "epoch": 3.7116564417177913,
+      "grad_norm": 0.17162683606147766,
+      "learning_rate": 4.603890310872197e-06,
+      "loss": 0.4514,
+      "step": 605
+    },
+    {
+      "epoch": 3.7177914110429446,
+      "grad_norm": 0.19262997806072235,
+      "learning_rate": 4.590022426367621e-06,
+      "loss": 0.4313,
+      "step": 606
+    },
+    {
+      "epoch": 3.7239263803680984,
+      "grad_norm": 0.19909274578094482,
+      "learning_rate": 4.576157716336369e-06,
+      "loss": 0.4347,
+      "step": 607
+    },
+    {
+      "epoch": 3.7300613496932513,
+      "grad_norm": 0.17881116271018982,
+      "learning_rate": 4.5622962881334666e-06,
+      "loss": 0.4581,
+      "step": 608
+    },
+    {
+      "epoch": 3.736196319018405,
+      "grad_norm": 0.18698081374168396,
+      "learning_rate": 4.5484382490885265e-06,
+      "loss": 0.4273,
+      "step": 609
+    },
+    {
+      "epoch": 3.7423312883435584,
+      "grad_norm": 0.17165279388427734,
+      "learning_rate": 4.534583706504919e-06,
+      "loss": 0.4009,
+      "step": 610
+    },
+    {
+      "epoch": 3.7484662576687118,
+      "grad_norm": 0.17795486748218536,
+      "learning_rate": 4.520732767658938e-06,
+      "loss": 0.4287,
+      "step": 611
+    },
+    {
+      "epoch": 3.754601226993865,
+      "grad_norm": 0.18492695689201355,
+      "learning_rate": 4.50688553979898e-06,
+      "loss": 0.4405,
+      "step": 612
+    },
+    {
+      "epoch": 3.7607361963190185,
+      "grad_norm": 0.20490054786205292,
+      "learning_rate": 4.493042130144702e-06,
+      "loss": 0.4344,
+      "step": 613
+    },
+    {
+      "epoch": 3.766871165644172,
+      "grad_norm": 0.1859641969203949,
+      "learning_rate": 4.479202645886196e-06,
+      "loss": 0.4283,
+      "step": 614
+    },
+    {
+      "epoch": 3.773006134969325,
+      "grad_norm": 0.20046532154083252,
+      "learning_rate": 4.4653671941831665e-06,
+      "loss": 0.444,
+      "step": 615
+    },
+    {
+      "epoch": 3.7791411042944785,
+      "grad_norm": 0.1931602507829666,
+      "learning_rate": 4.451535882164084e-06,
+      "loss": 0.4278,
+      "step": 616
+    },
+    {
+      "epoch": 3.785276073619632,
+      "grad_norm": 0.19650715589523315,
+      "learning_rate": 4.437708816925374e-06,
+      "loss": 0.4441,
+      "step": 617
+    },
+    {
+      "epoch": 3.791411042944785,
+      "grad_norm": 0.18395523726940155,
+      "learning_rate": 4.423886105530573e-06,
+      "loss": 0.4286,
+      "step": 618
+    },
+    {
+      "epoch": 3.7975460122699385,
+      "grad_norm": 0.19610300660133362,
+      "learning_rate": 4.410067855009506e-06,
+      "loss": 0.4051,
+      "step": 619
+    },
+    {
+      "epoch": 3.8036809815950923,
+      "grad_norm": 0.18188579380512238,
+      "learning_rate": 4.396254172357462e-06,
+      "loss": 0.4446,
+      "step": 620
+    },
+    {
+      "epoch": 3.809815950920245,
+      "grad_norm": 0.19300203025341034,
+      "learning_rate": 4.382445164534357e-06,
+      "loss": 0.4438,
+      "step": 621
+    },
+    {
+      "epoch": 3.815950920245399,
+      "grad_norm": 0.1908637285232544,
+      "learning_rate": 4.368640938463909e-06,
+      "loss": 0.4448,
+      "step": 622
+    },
+    {
+      "epoch": 3.8220858895705523,
+      "grad_norm": 0.17132696509361267,
+      "learning_rate": 4.354841601032811e-06,
+      "loss": 0.4108,
+      "step": 623
+    },
+    {
+      "epoch": 3.8282208588957056,
+      "grad_norm": 0.16772066056728363,
+      "learning_rate": 4.341047259089906e-06,
+      "loss": 0.4376,
+      "step": 624
+    },
+    {
+      "epoch": 3.834355828220859,
+      "grad_norm": 0.20349860191345215,
+      "learning_rate": 4.327258019445355e-06,
+      "loss": 0.4809,
+      "step": 625
+    },
+    {
+      "epoch": 3.8404907975460123,
+      "grad_norm": 0.1808754950761795,
+      "learning_rate": 4.313473988869811e-06,
+      "loss": 0.4217,
+      "step": 626
+    },
+    {
+      "epoch": 3.8466257668711656,
+      "grad_norm": 0.17573867738246918,
+      "learning_rate": 4.299695274093593e-06,
+      "loss": 0.4149,
+      "step": 627
+    },
+    {
+      "epoch": 3.852760736196319,
+      "grad_norm": 0.1589432954788208,
+      "learning_rate": 4.28592198180586e-06,
+      "loss": 0.4216,
+      "step": 628
+    },
+    {
+      "epoch": 3.8588957055214723,
+      "grad_norm": 0.17543958127498627,
+      "learning_rate": 4.272154218653784e-06,
+      "loss": 0.474,
+      "step": 629
+    },
+    {
+      "epoch": 3.8650306748466257,
+      "grad_norm": 0.19895213842391968,
+      "learning_rate": 4.258392091241727e-06,
+      "loss": 0.3786,
+      "step": 630
+    },
+    {
+      "epoch": 3.871165644171779,
+      "grad_norm": 0.2045029103755951,
+      "learning_rate": 4.244635706130408e-06,
+      "loss": 0.4593,
+      "step": 631
+    },
+    {
+      "epoch": 3.8773006134969323,
+      "grad_norm": 0.2224407196044922,
+      "learning_rate": 4.23088516983609e-06,
+      "loss": 0.4171,
+      "step": 632
+    },
+    {
+      "epoch": 3.883435582822086,
+      "grad_norm": 0.17593076825141907,
+      "learning_rate": 4.21714058882974e-06,
+      "loss": 0.4178,
+      "step": 633
+    },
+    {
+      "epoch": 3.889570552147239,
+      "grad_norm": 0.18816006183624268,
+      "learning_rate": 4.203402069536224e-06,
+      "loss": 0.4484,
+      "step": 634
+    },
+    {
+      "epoch": 3.895705521472393,
+      "grad_norm": 0.18676427006721497,
+      "learning_rate": 4.18966971833346e-06,
+      "loss": 0.4609,
+      "step": 635
+    },
+    {
+      "epoch": 3.901840490797546,
+      "grad_norm": 0.18539687991142273,
+      "learning_rate": 4.175943641551616e-06,
+      "loss": 0.4407,
+      "step": 636
+    },
+    {
+      "epoch": 3.9079754601226995,
+      "grad_norm": 0.18633869290351868,
+      "learning_rate": 4.162223945472271e-06,
+      "loss": 0.4665,
+      "step": 637
+    },
+    {
+      "epoch": 3.914110429447853,
+      "grad_norm": 0.1992420107126236,
+      "learning_rate": 4.1485107363276e-06,
+      "loss": 0.4493,
+      "step": 638
+    },
+    {
+      "epoch": 3.920245398773006,
+      "grad_norm": 0.1835312396287918,
+      "learning_rate": 4.1348041202995484e-06,
+      "loss": 0.4172,
+      "step": 639
+    },
+    {
+      "epoch": 3.9263803680981595,
+      "grad_norm": 0.1969442069530487,
+      "learning_rate": 4.121104203519012e-06,
+      "loss": 0.4374,
+      "step": 640
+    },
+    {
+      "epoch": 3.932515337423313,
+      "grad_norm": 0.20216475427150726,
+      "learning_rate": 4.107411092065015e-06,
+      "loss": 0.4207,
+      "step": 641
+    },
+    {
+      "epoch": 3.938650306748466,
+      "grad_norm": 0.17508967220783234,
+      "learning_rate": 4.093724891963882e-06,
+      "loss": 0.4164,
+      "step": 642
+    },
+    {
+      "epoch": 3.9447852760736195,
+      "grad_norm": 0.19727467000484467,
+      "learning_rate": 4.080045709188431e-06,
+      "loss": 0.4335,
+      "step": 643
+    },
+    {
+      "epoch": 3.950920245398773,
+      "grad_norm": 0.1708287000656128,
+      "learning_rate": 4.066373649657142e-06,
+      "loss": 0.4187,
+      "step": 644
+    },
+    {
+      "epoch": 3.957055214723926,
+      "grad_norm": 0.21632403135299683,
+      "learning_rate": 4.052708819233334e-06,
+      "loss": 0.4237,
+      "step": 645
+    },
+    {
+      "epoch": 3.96319018404908,
+      "grad_norm": 0.17434976994991302,
+      "learning_rate": 4.039051323724355e-06,
+      "loss": 0.4238,
+      "step": 646
+    },
+    {
+      "epoch": 3.969325153374233,
+      "grad_norm": 0.18432791531085968,
+      "learning_rate": 4.025401268880762e-06,
+      "loss": 0.4483,
+      "step": 647
+    },
+    {
+      "epoch": 3.9754601226993866,
+      "grad_norm": 0.1773991733789444,
+      "learning_rate": 4.011758760395491e-06,
+      "loss": 0.3934,
+      "step": 648
+    },
+    {
+      "epoch": 3.98159509202454,
+      "grad_norm": 0.19982725381851196,
+      "learning_rate": 3.998123903903051e-06,
+      "loss": 0.4381,
+      "step": 649
+    },
+    {
+      "epoch": 3.9877300613496933,
+      "grad_norm": 0.21029439568519592,
+      "learning_rate": 3.9844968049786995e-06,
+      "loss": 0.4267,
+      "step": 650
+    },
+    {
+      "epoch": 3.9938650306748467,
+      "grad_norm": 0.21180294454097748,
+      "learning_rate": 3.97087756913763e-06,
+      "loss": 0.4185,
+      "step": 651
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.17993085086345673,
+      "learning_rate": 3.957266301834145e-06,
+      "loss": 0.4382,
+      "step": 652
+    },
+    {
+      "epoch": 4.006134969325154,
+      "grad_norm": 0.2000538408756256,
+      "learning_rate": 3.943663108460857e-06,
+      "loss": 0.4189,
+      "step": 653
+    },
+    {
+      "epoch": 4.012269938650307,
+      "grad_norm": 0.1886722296476364,
+      "learning_rate": 3.93006809434785e-06,
+      "loss": 0.4464,
+      "step": 654
+    },
+    {
+      "epoch": 4.0184049079754605,
+      "grad_norm": 0.1883607655763626,
+      "learning_rate": 3.916481364761885e-06,
+      "loss": 0.417,
+      "step": 655
+    },
+    {
+      "epoch": 4.024539877300613,
+      "grad_norm": 0.1941026896238327,
+      "learning_rate": 3.90290302490557e-06,
+      "loss": 0.4515,
+      "step": 656
+    },
+    {
+      "epoch": 4.030674846625767,
+      "grad_norm": 0.1868050992488861,
+      "learning_rate": 3.889333179916552e-06,
+      "loss": 0.415,
+      "step": 657
+    },
+    {
+      "epoch": 4.03680981595092,
+      "grad_norm": 0.21741467714309692,
+      "learning_rate": 3.875771934866702e-06,
+      "loss": 0.3978,
+      "step": 658
+    },
+    {
+      "epoch": 4.042944785276074,
+      "grad_norm": 0.18438977003097534,
+      "learning_rate": 3.862219394761305e-06,
+      "loss": 0.4197,
+      "step": 659
+    },
+    {
+      "epoch": 4.049079754601227,
+      "grad_norm": 0.21055185794830322,
+      "learning_rate": 3.848675664538238e-06,
+      "loss": 0.4022,
+      "step": 660
+    },
+    {
+      "epoch": 4.0552147239263805,
+      "grad_norm": 0.1822698414325714,
+      "learning_rate": 3.8351408490671614e-06,
+      "loss": 0.4707,
+      "step": 661
+    },
+    {
+      "epoch": 4.061349693251533,
+      "grad_norm": 0.19663125276565552,
+      "learning_rate": 3.821615053148717e-06,
+      "loss": 0.4332,
+      "step": 662
+    },
+    {
+      "epoch": 4.067484662576687,
+      "grad_norm": 0.20445053279399872,
+      "learning_rate": 3.8080983815137017e-06,
+      "loss": 0.4197,
+      "step": 663
+    },
+    {
+      "epoch": 4.07361963190184,
+      "grad_norm": 0.17627210915088654,
+      "learning_rate": 3.7945909388222636e-06,
+      "loss": 0.4325,
+      "step": 664
+    },
+    {
+      "epoch": 4.079754601226994,
+      "grad_norm": 0.17431975901126862,
+      "learning_rate": 3.781092829663089e-06,
+      "loss": 0.4354,
+      "step": 665
+    },
+    {
+      "epoch": 4.085889570552148,
+      "grad_norm": 0.1871926486492157,
+      "learning_rate": 3.7676041585525956e-06,
+      "loss": 0.4155,
+      "step": 666
+    },
+    {
+      "epoch": 4.0920245398773005,
+      "grad_norm": 0.19517648220062256,
+      "learning_rate": 3.7541250299341243e-06,
+      "loss": 0.4327,
+      "step": 667
+    },
+    {
+      "epoch": 4.098159509202454,
+      "grad_norm": 0.1866348683834076,
+      "learning_rate": 3.740655548177125e-06,
+      "loss": 0.4042,
+      "step": 668
+    },
+    {
+      "epoch": 4.104294478527607,
+      "grad_norm": 0.17768587172031403,
+      "learning_rate": 3.7271958175763518e-06,
+      "loss": 0.4258,
+      "step": 669
+    },
+    {
+      "epoch": 4.110429447852761,
+      "grad_norm": 0.17544397711753845,
+      "learning_rate": 3.713745942351056e-06,
+      "loss": 0.4033,
+      "step": 670
+    },
+    {
+      "epoch": 4.116564417177914,
+      "grad_norm": 0.18574115633964539,
+      "learning_rate": 3.7003060266441804e-06,
+      "loss": 0.4229,
+      "step": 671
+    },
+    {
+      "epoch": 4.122699386503068,
+      "grad_norm": 0.1867496371269226,
+      "learning_rate": 3.6868761745215474e-06,
+      "loss": 0.432,
+      "step": 672
+    },
+    {
+      "epoch": 4.128834355828221,
+      "grad_norm": 0.19182687997817993,
+      "learning_rate": 3.6734564899710577e-06,
+      "loss": 0.4408,
+      "step": 673
+    },
+    {
+      "epoch": 4.134969325153374,
+      "grad_norm": 0.18049779534339905,
+      "learning_rate": 3.660047076901885e-06,
+      "loss": 0.4351,
+      "step": 674
+    },
+    {
+      "epoch": 4.141104294478527,
+      "grad_norm": 0.19350336492061615,
+      "learning_rate": 3.646648039143669e-06,
+      "loss": 0.4494,
+      "step": 675
+    },
+    {
+      "epoch": 4.147239263803681,
+      "grad_norm": 0.20113109052181244,
+      "learning_rate": 3.633259480445715e-06,
+      "loss": 0.4387,
+      "step": 676
+    },
+    {
+      "epoch": 4.153374233128835,
+      "grad_norm": 0.18117064237594604,
+      "learning_rate": 3.6198815044761847e-06,
+      "loss": 0.4792,
+      "step": 677
+    },
+    {
+      "epoch": 4.159509202453988,
+      "grad_norm": 0.1871654987335205,
+      "learning_rate": 3.6065142148213033e-06,
+      "loss": 0.4623,
+      "step": 678
+    },
+    {
+      "epoch": 4.1656441717791415,
+      "grad_norm": 0.17341911792755127,
+      "learning_rate": 3.5931577149845465e-06,
+      "loss": 0.4431,
+      "step": 679
+    },
+    {
+      "epoch": 4.171779141104294,
+      "grad_norm": 0.1755349338054657,
+      "learning_rate": 3.579812108385843e-06,
+      "loss": 0.4163,
+      "step": 680
+    },
+    {
+      "epoch": 4.177914110429448,
+      "grad_norm": 0.20305219292640686,
+      "learning_rate": 3.566477498360782e-06,
+      "loss": 0.4365,
+      "step": 681
+    },
+    {
+      "epoch": 4.184049079754601,
+      "grad_norm": 0.1874455362558365,
+      "learning_rate": 3.5531539881597967e-06,
+      "loss": 0.4727,
+      "step": 682
+    },
+    {
+      "epoch": 4.190184049079755,
+      "grad_norm": 0.20131394267082214,
+      "learning_rate": 3.5398416809473813e-06,
+      "loss": 0.4145,
+      "step": 683
+    },
+    {
+      "epoch": 4.196319018404908,
+      "grad_norm": 0.18988259136676788,
+      "learning_rate": 3.5265406798012804e-06,
+      "loss": 0.4457,
+      "step": 684
+    },
+    {
+      "epoch": 4.2024539877300615,
+      "grad_norm": 0.1912226378917694,
+      "learning_rate": 3.5132510877116953e-06,
+      "loss": 0.4358,
+      "step": 685
+    },
+    {
+      "epoch": 4.208588957055214,
+      "grad_norm": 0.17754784226417542,
+      "learning_rate": 3.4999730075804907e-06,
+      "loss": 0.4116,
+      "step": 686
+    },
+    {
+      "epoch": 4.214723926380368,
+      "grad_norm": 0.1884073168039322,
+      "learning_rate": 3.4867065422203885e-06,
+      "loss": 0.4366,
+      "step": 687
+    },
+    {
+      "epoch": 4.220858895705521,
+      "grad_norm": 0.18211857974529266,
+      "learning_rate": 3.473451794354179e-06,
+      "loss": 0.4276,
+      "step": 688
+    },
+    {
+      "epoch": 4.226993865030675,
+      "grad_norm": 0.20465627312660217,
+      "learning_rate": 3.460208866613923e-06,
+      "loss": 0.4266,
+      "step": 689
+    },
+    {
+      "epoch": 4.233128834355828,
+      "grad_norm": 0.26603659987449646,
+      "learning_rate": 3.4469778615401616e-06,
+      "loss": 0.4693,
+      "step": 690
+    },
+    {
+      "epoch": 4.2392638036809815,
+      "grad_norm": 0.19085724651813507,
+      "learning_rate": 3.4337588815811128e-06,
+      "loss": 0.4322,
+      "step": 691
+    },
+    {
+      "epoch": 4.245398773006135,
+      "grad_norm": 0.19165652990341187,
+      "learning_rate": 3.420552029091886e-06,
+      "loss": 0.4401,
+      "step": 692
+    },
+    {
+      "epoch": 4.251533742331288,
+      "grad_norm": 0.21447667479515076,
+      "learning_rate": 3.4073574063336857e-06,
+      "loss": 0.4327,
+      "step": 693
+    },
+    {
+      "epoch": 4.257668711656442,
+      "grad_norm": 0.21556028723716736,
+      "learning_rate": 3.394175115473024e-06,
+      "loss": 0.4507,
+      "step": 694
+    },
+    {
+      "epoch": 4.263803680981595,
+      "grad_norm": 0.16888567805290222,
+      "learning_rate": 3.3810052585809233e-06,
+      "loss": 0.4292,
+      "step": 695
+    },
+    {
+      "epoch": 4.269938650306749,
+      "grad_norm": 0.19774934649467468,
+      "learning_rate": 3.3678479376321304e-06,
+      "loss": 0.4345,
+      "step": 696
+    },
+    {
+      "epoch": 4.276073619631902,
+      "grad_norm": 0.18974056839942932,
+      "learning_rate": 3.354703254504328e-06,
+      "loss": 0.401,
+      "step": 697
+    },
+    {
+      "epoch": 4.282208588957055,
+      "grad_norm": 0.18433569371700287,
+      "learning_rate": 3.3415713109773386e-06,
+      "loss": 0.4327,
+      "step": 698
+    },
+    {
+      "epoch": 4.288343558282208,
+      "grad_norm": 0.1922151893377304,
+      "learning_rate": 3.328452208732349e-06,
+      "loss": 0.4239,
+      "step": 699
+    },
+    {
+      "epoch": 4.294478527607362,
+      "grad_norm": 0.18833968043327332,
+      "learning_rate": 3.3153460493511086e-06,
+      "loss": 0.4287,
+      "step": 700
+    },
+    {
+      "epoch": 4.300613496932515,
+      "grad_norm": 0.18338987231254578,
+      "learning_rate": 3.302252934315151e-06,
+      "loss": 0.4467,
+      "step": 701
+    },
+    {
+      "epoch": 4.306748466257669,
+      "grad_norm": 0.19972403347492218,
+      "learning_rate": 3.2891729650050096e-06,
+      "loss": 0.4153,
+      "step": 702
+    },
+    {
+      "epoch": 4.3128834355828225,
+      "grad_norm": 0.19376607239246368,
+      "learning_rate": 3.276106242699429e-06,
+      "loss": 0.4117,
+      "step": 703
+    },
+    {
+      "epoch": 4.319018404907975,
+      "grad_norm": 0.18517455458641052,
+      "learning_rate": 3.263052868574578e-06,
+      "loss": 0.4168,
+      "step": 704
+    },
+    {
+      "epoch": 4.325153374233129,
+      "grad_norm": 0.18251436948776245,
+      "learning_rate": 3.2500129437032756e-06,
+      "loss": 0.4552,
+      "step": 705
+    },
+    {
+      "epoch": 4.331288343558282,
+      "grad_norm": 0.1943109780550003,
+      "learning_rate": 3.236986569054199e-06,
+      "loss": 0.431,
+      "step": 706
+    },
+    {
+      "epoch": 4.337423312883436,
+      "grad_norm": 0.18778099119663239,
+      "learning_rate": 3.2239738454911057e-06,
+      "loss": 0.413,
+      "step": 707
+    },
+    {
+      "epoch": 4.343558282208589,
+      "grad_norm": 0.18466182053089142,
+      "learning_rate": 3.2109748737720537e-06,
+      "loss": 0.4332,
+      "step": 708
+    },
+    {
+      "epoch": 4.3496932515337425,
+      "grad_norm": 0.19483985006809235,
+      "learning_rate": 3.197989754548618e-06,
+      "loss": 0.436,
+      "step": 709
+    },
+    {
+      "epoch": 4.355828220858895,
+      "grad_norm": 0.17666223645210266,
+      "learning_rate": 3.1850185883651175e-06,
+      "loss": 0.4102,
+      "step": 710
+    },
+    {
+      "epoch": 4.361963190184049,
+      "grad_norm": 0.19580869376659393,
+      "learning_rate": 3.1720614756578267e-06,
+      "loss": 0.4151,
+      "step": 711
+    },
+    {
+      "epoch": 4.368098159509202,
+      "grad_norm": 0.20442509651184082,
+      "learning_rate": 3.1591185167542047e-06,
+      "loss": 0.4353,
+      "step": 712
+    },
+    {
+      "epoch": 4.374233128834356,
+      "grad_norm": 0.19967563450336456,
+      "learning_rate": 3.14618981187212e-06,
+      "loss": 0.4239,
+      "step": 713
+    },
+    {
+      "epoch": 4.38036809815951,
+      "grad_norm": 0.18093548715114594,
+      "learning_rate": 3.1332754611190695e-06,
+      "loss": 0.4312,
+      "step": 714
+    },
+    {
+      "epoch": 4.386503067484663,
+      "grad_norm": 0.2043561488389969,
+      "learning_rate": 3.1203755644914046e-06,
+      "loss": 0.4263,
+      "step": 715
+    },
+    {
+      "epoch": 4.392638036809816,
+      "grad_norm": 0.18950863182544708,
+      "learning_rate": 3.1074902218735602e-06,
+      "loss": 0.4453,
+      "step": 716
+    },
+    {
+      "epoch": 4.398773006134969,
+      "grad_norm": 0.19419826567173004,
+      "learning_rate": 3.0946195330372754e-06,
+      "loss": 0.4252,
+      "step": 717
+    },
+    {
+      "epoch": 4.404907975460123,
+      "grad_norm": 0.1931648850440979,
+      "learning_rate": 3.08176359764083e-06,
+      "loss": 0.422,
+      "step": 718
+    },
+    {
+      "epoch": 4.411042944785276,
+      "grad_norm": 0.19961263239383698,
+      "learning_rate": 3.0689225152282627e-06,
+      "loss": 0.4541,
+      "step": 719
+    },
+    {
+      "epoch": 4.41717791411043,
+      "grad_norm": 0.1944529414176941,
+      "learning_rate": 3.0560963852286046e-06,
+      "loss": 0.424,
+      "step": 720
+    },
+    {
+      "epoch": 4.423312883435583,
+      "grad_norm": 0.19636866450309753,
+      "learning_rate": 3.043285306955114e-06,
+      "loss": 0.4994,
+      "step": 721
+    },
+    {
+      "epoch": 4.429447852760736,
+      "grad_norm": 0.18336038291454315,
+      "learning_rate": 3.0304893796044988e-06,
+      "loss": 0.4109,
+      "step": 722
+    },
+    {
+      "epoch": 4.435582822085889,
+      "grad_norm": 0.18142063915729523,
+      "learning_rate": 3.017708702256153e-06,
+      "loss": 0.4464,
+      "step": 723
+    },
+    {
+      "epoch": 4.441717791411043,
+      "grad_norm": 0.17481929063796997,
+      "learning_rate": 3.004943373871393e-06,
+      "loss": 0.4341,
+      "step": 724
+    },
+    {
+      "epoch": 4.447852760736196,
+      "grad_norm": 0.19967781007289886,
+      "learning_rate": 2.9921934932926837e-06,
+      "loss": 0.3999,
+      "step": 725
+    },
+    {
+      "epoch": 4.45398773006135,
+      "grad_norm": 0.19884563982486725,
+      "learning_rate": 2.9794591592428767e-06,
+      "loss": 0.4417,
+      "step": 726
+    },
+    {
+      "epoch": 4.460122699386503,
+      "grad_norm": 0.1929159313440323,
+      "learning_rate": 2.966740470324451e-06,
+      "loss": 0.4371,
+      "step": 727
+    },
+    {
+      "epoch": 4.466257668711656,
+      "grad_norm": 0.19332391023635864,
+      "learning_rate": 2.954037525018739e-06,
+      "loss": 0.4668,
+      "step": 728
+    },
+    {
+      "epoch": 4.47239263803681,
+      "grad_norm": 0.18918125331401825,
+      "learning_rate": 2.9413504216851742e-06,
+      "loss": 0.4179,
+      "step": 729
+    },
+    {
+      "epoch": 4.478527607361963,
+      "grad_norm": 0.19312560558319092,
+      "learning_rate": 2.9286792585605206e-06,
+      "loss": 0.4045,
+      "step": 730
+    },
+    {
+      "epoch": 4.484662576687117,
+      "grad_norm": 0.19809098541736603,
+      "learning_rate": 2.9160241337581198e-06,
+      "loss": 0.4417,
+      "step": 731
+    },
+    {
+      "epoch": 4.49079754601227,
+      "grad_norm": 0.19697564840316772,
+      "learning_rate": 2.903385145267129e-06,
+      "loss": 0.4272,
+      "step": 732
+    },
+    {
+      "epoch": 4.4969325153374236,
+      "grad_norm": 0.2234029769897461,
+      "learning_rate": 2.8907623909517555e-06,
+      "loss": 0.4365,
+      "step": 733
+    },
+    {
+      "epoch": 4.5030674846625764,
+      "grad_norm": 0.2001408040523529,
+      "learning_rate": 2.8781559685505106e-06,
+      "loss": 0.4427,
+      "step": 734
+    },
+    {
+      "epoch": 4.50920245398773,
+      "grad_norm": 0.18446412682533264,
+      "learning_rate": 2.8655659756754474e-06,
+      "loss": 0.4511,
+      "step": 735
+    },
+    {
+      "epoch": 4.515337423312883,
+      "grad_norm": 0.20659488439559937,
+      "learning_rate": 2.8529925098113943e-06,
+      "loss": 0.4249,
+      "step": 736
+    },
+    {
+      "epoch": 4.521472392638037,
+      "grad_norm": 0.20220941305160522,
+      "learning_rate": 2.8404356683152256e-06,
+      "loss": 0.4228,
+      "step": 737
+    },
+    {
+      "epoch": 4.52760736196319,
+      "grad_norm": 0.21185429394245148,
+      "learning_rate": 2.827895548415084e-06,
+      "loss": 0.402,
+      "step": 738
+    },
+    {
+      "epoch": 4.533742331288344,
+      "grad_norm": 0.189244344830513,
+      "learning_rate": 2.8153722472096334e-06,
+      "loss": 0.4236,
+      "step": 739
+    },
+    {
+      "epoch": 4.539877300613497,
+      "grad_norm": 0.1980522722005844,
+      "learning_rate": 2.8028658616673184e-06,
+      "loss": 0.4302,
+      "step": 740
+    },
+    {
+      "epoch": 4.54601226993865,
+      "grad_norm": 0.19940122961997986,
+      "learning_rate": 2.7903764886255942e-06,
+      "loss": 0.408,
+      "step": 741
+    },
+    {
+      "epoch": 4.552147239263804,
+      "grad_norm": 0.19964726269245148,
+      "learning_rate": 2.777904224790197e-06,
+      "loss": 0.4316,
+      "step": 742
+    },
+    {
+      "epoch": 4.558282208588957,
+      "grad_norm": 0.20940682291984558,
+      "learning_rate": 2.765449166734382e-06,
+      "loss": 0.4208,
+      "step": 743
+    },
+    {
+      "epoch": 4.564417177914111,
+      "grad_norm": 0.1969623565673828,
+      "learning_rate": 2.7530114108981775e-06,
+      "loss": 0.4035,
+      "step": 744
+    },
+    {
+      "epoch": 4.570552147239264,
+      "grad_norm": 0.20081190764904022,
+      "learning_rate": 2.7405910535876407e-06,
+      "loss": 0.4469,
+      "step": 745
+    },
+    {
+      "epoch": 4.576687116564417,
+      "grad_norm": 0.1978064477443695,
+      "learning_rate": 2.7281881909741185e-06,
+      "loss": 0.4336,
+      "step": 746
+    },
+    {
+      "epoch": 4.58282208588957,
+      "grad_norm": 0.2098206877708435,
+      "learning_rate": 2.715802919093484e-06,
+      "loss": 0.4359,
+      "step": 747
+    },
+    {
+      "epoch": 4.588957055214724,
+      "grad_norm": 0.20042365789413452,
+      "learning_rate": 2.7034353338454142e-06,
+      "loss": 0.417,
+      "step": 748
+    },
+    {
+      "epoch": 4.595092024539877,
+      "grad_norm": 0.2288142442703247,
+      "learning_rate": 2.691085530992629e-06,
+      "loss": 0.4272,
+      "step": 749
+    },
+    {
+      "epoch": 4.601226993865031,
+      "grad_norm": 0.2015533745288849,
+      "learning_rate": 2.678753606160166e-06,
+      "loss": 0.407,
+      "step": 750
+    },
+    {
+      "epoch": 4.6073619631901845,
+      "grad_norm": 0.19977515935897827,
+      "learning_rate": 2.6664396548346303e-06,
+      "loss": 0.4324,
+      "step": 751
+    },
+    {
+      "epoch": 4.613496932515337,
+      "grad_norm": 0.1808634102344513,
+      "learning_rate": 2.654143772363455e-06,
+      "loss": 0.3987,
+      "step": 752
+    },
+    {
+      "epoch": 4.61963190184049,
+      "grad_norm": 0.20292384922504425,
+      "learning_rate": 2.6418660539541674e-06,
+      "loss": 0.405,
+      "step": 753
+    },
+    {
+      "epoch": 4.625766871165644,
+      "grad_norm": 0.19244877994060516,
+      "learning_rate": 2.6296065946736506e-06,
+      "loss": 0.4307,
+      "step": 754
+    },
+    {
+      "epoch": 4.631901840490798,
+      "grad_norm": 0.20017580687999725,
+      "learning_rate": 2.617365489447404e-06,
+      "loss": 0.4162,
+      "step": 755
+    },
+    {
+      "epoch": 4.638036809815951,
+      "grad_norm": 0.18756996095180511,
+      "learning_rate": 2.6051428330588147e-06,
+      "loss": 0.4307,
+      "step": 756
+    },
+    {
+      "epoch": 4.644171779141105,
+      "grad_norm": 0.1955227255821228,
+      "learning_rate": 2.5929387201484133e-06,
+      "loss": 0.4546,
+      "step": 757
+    },
+    {
+      "epoch": 4.6503067484662575,
+      "grad_norm": 0.1984601467847824,
+      "learning_rate": 2.5807532452131533e-06,
+      "loss": 0.4017,
+      "step": 758
+    },
+    {
+      "epoch": 4.656441717791411,
+      "grad_norm": 0.19388411939144135,
+      "learning_rate": 2.5685865026056745e-06,
+      "loss": 0.4706,
+      "step": 759
+    },
+    {
+      "epoch": 4.662576687116564,
+      "grad_norm": 0.18463459610939026,
+      "learning_rate": 2.5564385865335628e-06,
+      "loss": 0.4108,
+      "step": 760
+    },
+    {
+      "epoch": 4.668711656441718,
+      "grad_norm": 0.18664699792861938,
+      "learning_rate": 2.544309591058638e-06,
+      "loss": 0.4299,
+      "step": 761
+    },
+    {
+      "epoch": 4.674846625766871,
+      "grad_norm": 0.18623848259449005,
+      "learning_rate": 2.5321996100962163e-06,
+      "loss": 0.4277,
+      "step": 762
+    },
+    {
+      "epoch": 4.680981595092025,
+      "grad_norm": 0.20391374826431274,
+      "learning_rate": 2.5201087374143783e-06,
+      "loss": 0.4436,
+      "step": 763
+    },
+    {
+      "epoch": 4.6871165644171775,
+      "grad_norm": 0.20646578073501587,
+      "learning_rate": 2.5080370666332532e-06,
+      "loss": 0.4502,
+      "step": 764
+    },
+    {
+      "epoch": 4.693251533742331,
+      "grad_norm": 0.20507745444774628,
+      "learning_rate": 2.495984691224287e-06,
+      "loss": 0.4164,
+      "step": 765
+    },
+    {
+      "epoch": 4.699386503067485,
+      "grad_norm": 0.19352522492408752,
+      "learning_rate": 2.4839517045095225e-06,
+      "loss": 0.412,
+      "step": 766
+    },
+    {
+      "epoch": 4.705521472392638,
+      "grad_norm": 0.19516626000404358,
+      "learning_rate": 2.4719381996608748e-06,
+      "loss": 0.4253,
+      "step": 767
+    },
+    {
+      "epoch": 4.711656441717792,
+      "grad_norm": 0.19005592167377472,
+      "learning_rate": 2.459944269699407e-06,
+      "loss": 0.4253,
+      "step": 768
+    },
+    {
+      "epoch": 4.717791411042945,
+      "grad_norm": 0.18863846361637115,
+      "learning_rate": 2.4479700074946154e-06,
+      "loss": 0.4063,
+      "step": 769
+    },
+    {
+      "epoch": 4.723926380368098,
+      "grad_norm": 0.2115408331155777,
+      "learning_rate": 2.4360155057637115e-06,
+      "loss": 0.439,
+      "step": 770
+    },
+    {
+      "epoch": 4.730061349693251,
+      "grad_norm": 0.1989760547876358,
+      "learning_rate": 2.4240808570708926e-06,
+      "loss": 0.4358,
+      "step": 771
+    },
+    {
+      "epoch": 4.736196319018405,
+      "grad_norm": 0.18984295427799225,
+      "learning_rate": 2.412166153826639e-06,
+      "loss": 0.4262,
+      "step": 772
+    },
+    {
+      "epoch": 4.742331288343558,
+      "grad_norm": 0.17575299739837646,
+      "learning_rate": 2.400271488286992e-06,
+      "loss": 0.4263,
+      "step": 773
+    },
+    {
+      "epoch": 4.748466257668712,
+      "grad_norm": 0.18755771219730377,
+      "learning_rate": 2.3883969525528396e-06,
+      "loss": 0.4089,
+      "step": 774
+    },
+    {
+      "epoch": 4.754601226993865,
+      "grad_norm": 0.2058447301387787,
+      "learning_rate": 2.3765426385692044e-06,
+      "loss": 0.4427,
+      "step": 775
+    },
+    {
+      "epoch": 4.7607361963190185,
+      "grad_norm": 0.20019108057022095,
+      "learning_rate": 2.3647086381245267e-06,
+      "loss": 0.4548,
+      "step": 776
+    },
+    {
+      "epoch": 4.766871165644172,
+      "grad_norm": 0.17807462811470032,
+      "learning_rate": 2.352895042849965e-06,
+      "loss": 0.4172,
+      "step": 777
+    },
+    {
+      "epoch": 4.773006134969325,
+      "grad_norm": 0.18490706384181976,
+      "learning_rate": 2.3411019442186766e-06,
+      "loss": 0.4319,
+      "step": 778
+    },
+    {
+      "epoch": 4.779141104294479,
+      "grad_norm": 0.1955891102552414,
+      "learning_rate": 2.3293294335451104e-06,
+      "loss": 0.4516,
+      "step": 779
+    },
+    {
+      "epoch": 4.785276073619632,
+      "grad_norm": 0.19553668797016144,
+      "learning_rate": 2.317577601984305e-06,
+      "loss": 0.4211,
+      "step": 780
+    },
+    {
+      "epoch": 4.791411042944786,
+      "grad_norm": 0.1864728480577469,
+      "learning_rate": 2.30584654053118e-06,
+      "loss": 0.3994,
+      "step": 781
+    },
+    {
+      "epoch": 4.7975460122699385,
+      "grad_norm": 0.1869369000196457,
+      "learning_rate": 2.294136340019826e-06,
+      "loss": 0.4526,
+      "step": 782
+    },
+    {
+      "epoch": 4.803680981595092,
+      "grad_norm": 0.19105935096740723,
+      "learning_rate": 2.282447091122816e-06,
+      "loss": 0.473,
+      "step": 783
+    },
+    {
+      "epoch": 4.809815950920245,
+      "grad_norm": 0.18605674803256989,
+      "learning_rate": 2.2707788843504836e-06,
+      "loss": 0.4634,
+      "step": 784
+    },
+    {
+      "epoch": 4.815950920245399,
+      "grad_norm": 0.18209435045719147,
+      "learning_rate": 2.2591318100502385e-06,
+      "loss": 0.4566,
+      "step": 785
+    },
+    {
+      "epoch": 4.822085889570552,
+      "grad_norm": 0.19514133036136627,
+      "learning_rate": 2.2475059584058612e-06,
+      "loss": 0.4082,
+      "step": 786
+    },
+    {
+      "epoch": 4.828220858895706,
+      "grad_norm": 0.192851722240448,
+      "learning_rate": 2.2359014194367986e-06,
+      "loss": 0.4739,
+      "step": 787
+    },
+    {
+      "epoch": 4.8343558282208585,
+      "grad_norm": 0.2023897022008896,
+      "learning_rate": 2.2243182829974775e-06,
+      "loss": 0.4416,
+      "step": 788
+    },
+    {
+      "epoch": 4.840490797546012,
+      "grad_norm": 0.18871212005615234,
+      "learning_rate": 2.2127566387766045e-06,
+      "loss": 0.4408,
+      "step": 789
+    },
+    {
+      "epoch": 4.846625766871165,
+      "grad_norm": 0.20089678466320038,
+      "learning_rate": 2.2012165762964677e-06,
+      "loss": 0.4236,
+      "step": 790
+    },
+    {
+      "epoch": 4.852760736196319,
+      "grad_norm": 0.2131776064634323,
+      "learning_rate": 2.189698184912249e-06,
+      "loss": 0.4368,
+      "step": 791
+    },
+    {
+      "epoch": 4.858895705521473,
+      "grad_norm": 0.2049061506986618,
+      "learning_rate": 2.17820155381133e-06,
+      "loss": 0.4444,
+      "step": 792
+    },
+    {
+      "epoch": 4.865030674846626,
+      "grad_norm": 0.18744398653507233,
+      "learning_rate": 2.1667267720126014e-06,
+      "loss": 0.4292,
+      "step": 793
+    },
+    {
+      "epoch": 4.871165644171779,
+      "grad_norm": 0.20928369462490082,
+      "learning_rate": 2.1552739283657753e-06,
+      "loss": 0.4321,
+      "step": 794
+    },
+    {
+      "epoch": 4.877300613496932,
+      "grad_norm": 0.18477670848369598,
+      "learning_rate": 2.1438431115506908e-06,
+      "loss": 0.4629,
+      "step": 795
+    },
+    {
+      "epoch": 4.883435582822086,
+      "grad_norm": 0.17708535492420197,
+      "learning_rate": 2.1324344100766376e-06,
+      "loss": 0.4647,
+      "step": 796
+    },
+    {
+      "epoch": 4.889570552147239,
+      "grad_norm": 0.1869410127401352,
+      "learning_rate": 2.1210479122816646e-06,
+      "loss": 0.4141,
+      "step": 797
+    },
+    {
+      "epoch": 4.895705521472393,
+      "grad_norm": 0.18549315631389618,
+      "learning_rate": 2.109683706331893e-06,
+      "loss": 0.4649,
+      "step": 798
+    },
+    {
+      "epoch": 4.901840490797546,
+      "grad_norm": 0.21219757199287415,
+      "learning_rate": 2.0983418802208416e-06,
+      "loss": 0.4252,
+      "step": 799
+    },
+    {
+      "epoch": 4.9079754601226995,
+      "grad_norm": 0.1829107105731964,
+      "learning_rate": 2.0870225217687408e-06,
+      "loss": 0.4095,
+      "step": 800
+    },
+    {
+      "epoch": 4.914110429447852,
+      "grad_norm": 0.2121373414993286,
+      "learning_rate": 2.0757257186218465e-06,
+      "loss": 0.3889,
+      "step": 801
+    },
+    {
+      "epoch": 4.920245398773006,
+      "grad_norm": 0.19195012748241425,
+      "learning_rate": 2.0644515582517803e-06,
+      "loss": 0.4324,
+      "step": 802
+    },
+    {
+      "epoch": 4.92638036809816,
+      "grad_norm": 0.20068559050559998,
+      "learning_rate": 2.053200127954828e-06,
+      "loss": 0.4484,
+      "step": 803
+    },
+    {
+      "epoch": 4.932515337423313,
+      "grad_norm": 0.18420425057411194,
+      "learning_rate": 2.0419715148512807e-06,
+      "loss": 0.4237,
+      "step": 804
+    },
+    {
+      "epoch": 4.938650306748467,
+      "grad_norm": 0.2045467048883438,
+      "learning_rate": 2.0307658058847577e-06,
+      "loss": 0.4114,
+      "step": 805
+    },
+    {
+      "epoch": 4.9447852760736195,
+      "grad_norm": 0.19473262131214142,
+      "learning_rate": 2.0195830878215236e-06,
+      "loss": 0.4305,
+      "step": 806
+    },
+    {
+      "epoch": 4.950920245398773,
+      "grad_norm": 0.1820230484008789,
+      "learning_rate": 2.0084234472498274e-06,
+      "loss": 0.4232,
+      "step": 807
+    },
+    {
+      "epoch": 4.957055214723926,
+      "grad_norm": 0.20162828266620636,
+      "learning_rate": 1.997286970579232e-06,
+      "loss": 0.3802,
+      "step": 808
+    },
+    {
+      "epoch": 4.96319018404908,
+      "grad_norm": 0.194149449467659,
+      "learning_rate": 1.9861737440399327e-06,
+      "loss": 0.4432,
+      "step": 809
+    },
+    {
+      "epoch": 4.969325153374233,
+      "grad_norm": 0.21044090390205383,
+      "learning_rate": 1.9750838536821048e-06,
+      "loss": 0.4031,
+      "step": 810
+    },
+    {
+      "epoch": 4.975460122699387,
+      "grad_norm": 0.18757662177085876,
+      "learning_rate": 1.964017385375228e-06,
+      "loss": 0.4047,
+      "step": 811
+    },
+    {
+      "epoch": 4.9815950920245395,
+      "grad_norm": 0.2228487730026245,
+      "learning_rate": 1.952974424807425e-06,
+      "loss": 0.432,
+      "step": 812
+    },
+    {
+      "epoch": 4.987730061349693,
+      "grad_norm": 0.21380779147148132,
+      "learning_rate": 1.9419550574847986e-06,
+      "loss": 0.4277,
+      "step": 813
+    },
+    {
+      "epoch": 4.993865030674847,
+      "grad_norm": 0.21621406078338623,
+      "learning_rate": 1.9309593687307622e-06,
+      "loss": 0.4616,
+      "step": 814
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.19357597827911377,
+      "learning_rate": 1.9199874436853904e-06,
+      "loss": 0.4202,
+      "step": 815
+    },
+    {
+      "epoch": 5.006134969325154,
+      "grad_norm": 0.18243946135044098,
+      "learning_rate": 1.9090393673047557e-06,
+      "loss": 0.4184,
+      "step": 816
+    },
+    {
+      "epoch": 5.012269938650307,
+      "grad_norm": 0.21097031235694885,
+      "learning_rate": 1.898115224360263e-06,
+      "loss": 0.4323,
+      "step": 817
+    },
+    {
+      "epoch": 5.0184049079754605,
+      "grad_norm": 0.19240233302116394,
+      "learning_rate": 1.8872150994380045e-06,
+      "loss": 0.4077,
+      "step": 818
+    },
+    {
+      "epoch": 5.024539877300613,
+      "grad_norm": 0.20225709676742554,
+      "learning_rate": 1.8763390769381017e-06,
+      "loss": 0.4164,
+      "step": 819
+    },
+    {
+      "epoch": 5.030674846625767,
+      "grad_norm": 0.22095705568790436,
+      "learning_rate": 1.865487241074041e-06,
+      "loss": 0.417,
+      "step": 820
+    },
+    {
+      "epoch": 5.03680981595092,
+      "grad_norm": 0.18732763826847076,
+      "learning_rate": 1.8546596758720437e-06,
+      "loss": 0.406,
+      "step": 821
+    },
+    {
+      "epoch": 5.042944785276074,
+      "grad_norm": 0.19568726420402527,
+      "learning_rate": 1.84385646517039e-06,
+      "loss": 0.4335,
+      "step": 822
+    },
+    {
+      "epoch": 5.049079754601227,
+      "grad_norm": 0.22073768079280853,
+      "learning_rate": 1.8330776926187904e-06,
+      "loss": 0.4485,
+      "step": 823
+    },
+    {
+      "epoch": 5.0552147239263805,
+      "grad_norm": 0.22986707091331482,
+      "learning_rate": 1.8223234416777275e-06,
+      "loss": 0.4347,
+      "step": 824
+    },
+    {
+      "epoch": 5.061349693251533,
+      "grad_norm": 0.17984724044799805,
+      "learning_rate": 1.8115937956178093e-06,
+      "loss": 0.4037,
+      "step": 825
+    },
+    {
+      "epoch": 5.067484662576687,
+      "grad_norm": 0.2089032679796219,
+      "learning_rate": 1.8008888375191302e-06,
+      "loss": 0.4223,
+      "step": 826
+    },
+    {
+      "epoch": 5.07361963190184,
+      "grad_norm": 0.18969732522964478,
+      "learning_rate": 1.7902086502706256e-06,
+      "loss": 0.4375,
+      "step": 827
+    },
+    {
+      "epoch": 5.079754601226994,
+      "grad_norm": 0.203078955411911,
+      "learning_rate": 1.779553316569425e-06,
+      "loss": 0.4424,
+      "step": 828
+    },
+    {
+      "epoch": 5.085889570552148,
+      "grad_norm": 0.20583704113960266,
+      "learning_rate": 1.7689229189202196e-06,
+      "loss": 0.4161,
+      "step": 829
+    },
+    {
+      "epoch": 5.0920245398773005,
+      "grad_norm": 0.20547014474868774,
+      "learning_rate": 1.758317539634618e-06,
+      "loss": 0.4291,
+      "step": 830
+    },
+    {
+      "epoch": 5.098159509202454,
+      "grad_norm": 0.20702853798866272,
+      "learning_rate": 1.747737260830512e-06,
+      "loss": 0.3902,
+      "step": 831
+    },
+    {
+      "epoch": 5.104294478527607,
+      "grad_norm": 0.20404289662837982,
+      "learning_rate": 1.7371821644314392e-06,
+      "loss": 0.4249,
+      "step": 832
+    },
+    {
+      "epoch": 5.110429447852761,
+      "grad_norm": 0.19528424739837646,
+      "learning_rate": 1.726652332165945e-06,
+      "loss": 0.4356,
+      "step": 833
+    },
+    {
+      "epoch": 5.116564417177914,
+      "grad_norm": 0.18722812831401825,
+      "learning_rate": 1.716147845566959e-06,
+      "loss": 0.4146,
+      "step": 834
+    },
+    {
+      "epoch": 5.122699386503068,
+      "grad_norm": 0.20289267599582672,
+      "learning_rate": 1.7056687859711563e-06,
+      "loss": 0.4343,
+      "step": 835
+    },
+    {
+      "epoch": 5.128834355828221,
+      "grad_norm": 0.19126549363136292,
+      "learning_rate": 1.695215234518326e-06,
+      "loss": 0.4834,
+      "step": 836
+    },
+    {
+      "epoch": 5.134969325153374,
+      "grad_norm": 0.20215460658073425,
+      "learning_rate": 1.6847872721507525e-06,
+      "loss": 0.4272,
+      "step": 837
+    },
+    {
+      "epoch": 5.141104294478527,
+      "grad_norm": 0.200180321931839,
+      "learning_rate": 1.674384979612579e-06,
+      "loss": 0.4332,
+      "step": 838
+    },
+    {
+      "epoch": 5.147239263803681,
+      "grad_norm": 0.20032553374767303,
+      "learning_rate": 1.6640084374491872e-06,
+      "loss": 0.459,
+      "step": 839
+    },
+    {
+      "epoch": 5.153374233128835,
+      "grad_norm": 0.20815366506576538,
+      "learning_rate": 1.653657726006575e-06,
+      "loss": 0.4564,
+      "step": 840
+    },
+    {
+      "epoch": 5.159509202453988,
+      "grad_norm": 0.19371084868907928,
+      "learning_rate": 1.6433329254307261e-06,
+      "loss": 0.4595,
+      "step": 841
+    },
+    {
+      "epoch": 5.1656441717791415,
+      "grad_norm": 0.179207444190979,
+      "learning_rate": 1.633034115667001e-06,
+      "loss": 0.4564,
+      "step": 842
+    },
+    {
+      "epoch": 5.171779141104294,
+      "grad_norm": 0.19510361552238464,
+      "learning_rate": 1.6227613764595107e-06,
+      "loss": 0.4207,
+      "step": 843
+    },
+    {
+      "epoch": 5.177914110429448,
+      "grad_norm": 0.19818256795406342,
+      "learning_rate": 1.6125147873504971e-06,
+      "loss": 0.4274,
+      "step": 844
+    },
+    {
+      "epoch": 5.184049079754601,
+      "grad_norm": 0.18541868031024933,
+      "learning_rate": 1.6022944276797265e-06,
+      "loss": 0.4103,
+      "step": 845
+    },
+    {
+      "epoch": 5.190184049079755,
+      "grad_norm": 0.19620756804943085,
+      "learning_rate": 1.5921003765838673e-06,
+      "loss": 0.4019,
+      "step": 846
+    },
+    {
+      "epoch": 5.196319018404908,
+      "grad_norm": 0.19282202422618866,
+      "learning_rate": 1.5819327129958762e-06,
+      "loss": 0.4135,
+      "step": 847
+    },
+    {
+      "epoch": 5.2024539877300615,
+      "grad_norm": 0.20788726210594177,
+      "learning_rate": 1.5717915156443953e-06,
+      "loss": 0.3979,
+      "step": 848
+    },
+    {
+      "epoch": 5.208588957055214,
+      "grad_norm": 0.20662984251976013,
+      "learning_rate": 1.5616768630531353e-06,
+      "loss": 0.4236,
+      "step": 849
+    },
+    {
+      "epoch": 5.214723926380368,
+      "grad_norm": 0.1801910698413849,
+      "learning_rate": 1.5515888335402706e-06,
+      "loss": 0.4094,
+      "step": 850
+    },
+    {
+      "epoch": 5.220858895705521,
+      "grad_norm": 0.21069598197937012,
+      "learning_rate": 1.5415275052178318e-06,
+      "loss": 0.4229,
+      "step": 851
+    },
+    {
+      "epoch": 5.226993865030675,
+      "grad_norm": 0.18792353570461273,
+      "learning_rate": 1.5314929559910985e-06,
+      "loss": 0.3851,
+      "step": 852
+    },
+    {
+      "epoch": 5.233128834355828,
+      "grad_norm": 0.21205176413059235,
+      "learning_rate": 1.5214852635580018e-06,
+      "loss": 0.4524,
+      "step": 853
+    },
+    {
+      "epoch": 5.2392638036809815,
+      "grad_norm": 0.20375634729862213,
+      "learning_rate": 1.5115045054085204e-06,
+      "loss": 0.4293,
+      "step": 854
+    },
+    {
+      "epoch": 5.245398773006135,
+      "grad_norm": 0.19179388880729675,
+      "learning_rate": 1.5015507588240742e-06,
+      "loss": 0.4438,
+      "step": 855
+    },
+    {
+      "epoch": 5.251533742331288,
+      "grad_norm": 0.1827925592660904,
+      "learning_rate": 1.4916241008769372e-06,
+      "loss": 0.4183,
+      "step": 856
+    },
+    {
+      "epoch": 5.257668711656442,
+      "grad_norm": 0.21141961216926575,
+      "learning_rate": 1.4817246084296327e-06,
+      "loss": 0.4388,
+      "step": 857
+    },
+    {
+      "epoch": 5.263803680981595,
+      "grad_norm": 0.18689335882663727,
+      "learning_rate": 1.4718523581343403e-06,
+      "loss": 0.4312,
+      "step": 858
+    },
+    {
+      "epoch": 5.269938650306749,
+      "grad_norm": 0.195315420627594,
+      "learning_rate": 1.4620074264323048e-06,
+      "loss": 0.4287,
+      "step": 859
+    },
+    {
+      "epoch": 5.276073619631902,
+      "grad_norm": 0.1855771541595459,
+      "learning_rate": 1.452189889553236e-06,
+      "loss": 0.4252,
+      "step": 860
+    },
+    {
+      "epoch": 5.282208588957055,
+      "grad_norm": 0.18047569692134857,
+      "learning_rate": 1.4423998235147307e-06,
+      "loss": 0.3753,
+      "step": 861
+    },
+    {
+      "epoch": 5.288343558282208,
+      "grad_norm": 0.19961078464984894,
+      "learning_rate": 1.4326373041216774e-06,
+      "loss": 0.4277,
+      "step": 862
+    },
+    {
+      "epoch": 5.294478527607362,
+      "grad_norm": 0.19070571660995483,
+      "learning_rate": 1.422902406965664e-06,
+      "loss": 0.4258,
+      "step": 863
+    },
+    {
+      "epoch": 5.300613496932515,
+      "grad_norm": 0.18950672447681427,
+      "learning_rate": 1.4131952074244037e-06,
+      "loss": 0.4223,
+      "step": 864
+    },
+    {
+      "epoch": 5.306748466257669,
+      "grad_norm": 0.1713619977235794,
+      "learning_rate": 1.4035157806611465e-06,
+      "loss": 0.4282,
+      "step": 865
+    },
+    {
+      "epoch": 5.3128834355828225,
+      "grad_norm": 0.18861691653728485,
+      "learning_rate": 1.39386420162409e-06,
+      "loss": 0.4332,
+      "step": 866
+    },
+    {
+      "epoch": 5.319018404907975,
+      "grad_norm": 0.20134612917900085,
+      "learning_rate": 1.3842405450458158e-06,
+      "loss": 0.431,
+      "step": 867
+    },
+    {
+      "epoch": 5.325153374233129,
+      "grad_norm": 0.19348467886447906,
+      "learning_rate": 1.3746448854426908e-06,
+      "loss": 0.3873,
+      "step": 868
+    },
+    {
+      "epoch": 5.331288343558282,
+      "grad_norm": 0.20825737714767456,
+      "learning_rate": 1.3650772971143067e-06,
+      "loss": 0.409,
+      "step": 869
+    },
+    {
+      "epoch": 5.337423312883436,
+      "grad_norm": 0.20485500991344452,
+      "learning_rate": 1.355537854142897e-06,
+      "loss": 0.4514,
+      "step": 870
+    },
+    {
+      "epoch": 5.343558282208589,
+      "grad_norm": 0.19826869666576385,
+      "learning_rate": 1.3460266303927604e-06,
+      "loss": 0.4406,
+      "step": 871
+    },
+    {
+      "epoch": 5.3496932515337425,
+      "grad_norm": 0.1876692771911621,
+      "learning_rate": 1.336543699509698e-06,
+      "loss": 0.445,
+      "step": 872
+    },
+    {
+      "epoch": 5.355828220858895,
+      "grad_norm": 0.19540998339653015,
+      "learning_rate": 1.3270891349204378e-06,
+      "loss": 0.4144,
+      "step": 873
+    },
+    {
+      "epoch": 5.361963190184049,
+      "grad_norm": 0.20348355174064636,
+      "learning_rate": 1.3176630098320615e-06,
+      "loss": 0.4201,
+      "step": 874
+    },
+    {
+      "epoch": 5.368098159509202,
+      "grad_norm": 0.1816159337759018,
+      "learning_rate": 1.3082653972314475e-06,
+      "loss": 0.4181,
+      "step": 875
+    },
+    {
+      "epoch": 5.374233128834356,
+      "grad_norm": 0.19366130232810974,
+      "learning_rate": 1.2988963698846997e-06,
+      "loss": 0.4246,
+      "step": 876
+    },
+    {
+      "epoch": 5.38036809815951,
+      "grad_norm": 0.1955079287290573,
+      "learning_rate": 1.2895560003365837e-06,
+      "loss": 0.4378,
+      "step": 877
+    },
+    {
+      "epoch": 5.386503067484663,
+      "grad_norm": 0.20214815437793732,
+      "learning_rate": 1.2802443609099696e-06,
+      "loss": 0.4306,
+      "step": 878
+    },
+    {
+      "epoch": 5.392638036809816,
+      "grad_norm": 0.1942247748374939,
+      "learning_rate": 1.270961523705264e-06,
+      "loss": 0.4158,
+      "step": 879
+    },
+    {
+      "epoch": 5.398773006134969,
+      "grad_norm": 0.19706179201602936,
+      "learning_rate": 1.2617075605998618e-06,
+      "loss": 0.4227,
+      "step": 880
+    },
+    {
+      "epoch": 5.404907975460123,
+      "grad_norm": 0.19230398535728455,
+      "learning_rate": 1.2524825432475828e-06,
+      "loss": 0.4289,
+      "step": 881
+    },
+    {
+      "epoch": 5.411042944785276,
+      "grad_norm": 0.1848897784948349,
+      "learning_rate": 1.2432865430781166e-06,
+      "loss": 0.4216,
+      "step": 882
+    },
+    {
+      "epoch": 5.41717791411043,
+      "grad_norm": 0.20822712779045105,
+      "learning_rate": 1.234119631296473e-06,
+      "loss": 0.4419,
+      "step": 883
+    },
+    {
+      "epoch": 5.423312883435583,
+      "grad_norm": 0.20798738300800323,
+      "learning_rate": 1.2249818788824324e-06,
+      "loss": 0.4567,
+      "step": 884
+    },
+    {
+      "epoch": 5.429447852760736,
+      "grad_norm": 0.2089272290468216,
+      "learning_rate": 1.2158733565899855e-06,
+      "loss": 0.4499,
+      "step": 885
+    },
+    {
+      "epoch": 5.435582822085889,
+      "grad_norm": 0.18762832880020142,
+      "learning_rate": 1.2067941349468021e-06,
+      "loss": 0.4365,
+      "step": 886
+    },
+    {
+      "epoch": 5.441717791411043,
+      "grad_norm": 0.22340965270996094,
+      "learning_rate": 1.1977442842536685e-06,
+      "loss": 0.4296,
+      "step": 887
+    },
+    {
+      "epoch": 5.447852760736196,
+      "grad_norm": 0.18473917245864868,
+      "learning_rate": 1.1887238745839536e-06,
+      "loss": 0.4347,
+      "step": 888
+    },
+    {
+      "epoch": 5.45398773006135,
+      "grad_norm": 0.2267259657382965,
+      "learning_rate": 1.179732975783065e-06,
+      "loss": 0.4137,
+      "step": 889
+    },
+    {
+      "epoch": 5.460122699386503,
+      "grad_norm": 0.20931611955165863,
+      "learning_rate": 1.1707716574679007e-06,
+      "loss": 0.3982,
+      "step": 890
+    },
+    {
+      "epoch": 5.466257668711656,
+      "grad_norm": 0.19619382917881012,
+      "learning_rate": 1.1618399890263215e-06,
+      "loss": 0.4455,
+      "step": 891
+    },
+    {
+      "epoch": 5.47239263803681,
+      "grad_norm": 0.18555289506912231,
+      "learning_rate": 1.1529380396166074e-06,
+      "loss": 0.4423,
+      "step": 892
+    },
+    {
+      "epoch": 5.478527607361963,
+      "grad_norm": 0.18679551780223846,
+      "learning_rate": 1.1440658781669179e-06,
+      "loss": 0.4392,
+      "step": 893
+    },
+    {
+      "epoch": 5.484662576687117,
+      "grad_norm": 0.19644379615783691,
+      "learning_rate": 1.1352235733747685e-06,
+      "loss": 0.4744,
+      "step": 894
+    },
+    {
+      "epoch": 5.49079754601227,
+      "grad_norm": 0.18453319370746613,
+      "learning_rate": 1.1264111937064902e-06,
+      "loss": 0.4172,
+      "step": 895
+    },
+    {
+      "epoch": 5.4969325153374236,
+      "grad_norm": 0.19223150610923767,
+      "learning_rate": 1.117628807396705e-06,
+      "loss": 0.435,
+      "step": 896
+    },
+    {
+      "epoch": 5.5030674846625764,
+      "grad_norm": 0.1897895187139511,
+      "learning_rate": 1.1088764824477938e-06,
+      "loss": 0.4312,
+      "step": 897
+    },
+    {
+      "epoch": 5.50920245398773,
+      "grad_norm": 0.20507152378559113,
+      "learning_rate": 1.100154286629369e-06,
+      "loss": 0.4133,
+      "step": 898
+    },
+    {
+      "epoch": 5.515337423312883,
+      "grad_norm": 0.183609277009964,
+      "learning_rate": 1.0914622874777547e-06,
+      "loss": 0.4109,
+      "step": 899
+    },
+    {
+      "epoch": 5.521472392638037,
+      "grad_norm": 0.18945349752902985,
+      "learning_rate": 1.0828005522954626e-06,
+      "loss": 0.4336,
+      "step": 900
+    },
+    {
+      "epoch": 5.52760736196319,
+      "grad_norm": 0.22751018404960632,
+      "learning_rate": 1.0741691481506627e-06,
+      "loss": 0.4164,
+      "step": 901
+    },
+    {
+      "epoch": 5.533742331288344,
+      "grad_norm": 0.21165773272514343,
+      "learning_rate": 1.0655681418766772e-06,
+      "loss": 0.4538,
+      "step": 902
+    },
+    {
+      "epoch": 5.539877300613497,
+      "grad_norm": 0.18855099380016327,
+      "learning_rate": 1.0569976000714544e-06,
+      "loss": 0.4405,
+      "step": 903
+    },
+    {
+      "epoch": 5.54601226993865,
+      "grad_norm": 0.19720859825611115,
+      "learning_rate": 1.0484575890970505e-06,
+      "loss": 0.4143,
+      "step": 904
+    },
+    {
+      "epoch": 5.552147239263804,
+      "grad_norm": 0.19412867724895477,
+      "learning_rate": 1.0399481750791291e-06,
+      "loss": 0.4443,
+      "step": 905
+    },
+    {
+      "epoch": 5.558282208588957,
+      "grad_norm": 0.19512735307216644,
+      "learning_rate": 1.0314694239064315e-06,
+      "loss": 0.4435,
+      "step": 906
+    },
+    {
+      "epoch": 5.564417177914111,
+      "grad_norm": 0.19326624274253845,
+      "learning_rate": 1.0230214012302807e-06,
+      "loss": 0.4268,
+      "step": 907
+    },
+    {
+      "epoch": 5.570552147239264,
+      "grad_norm": 0.18657195568084717,
+      "learning_rate": 1.014604172464067e-06,
+      "loss": 0.4158,
+      "step": 908
+    },
+    {
+      "epoch": 5.576687116564417,
+      "grad_norm": 0.20302285254001617,
+      "learning_rate": 1.0062178027827385e-06,
+      "loss": 0.4361,
+      "step": 909
+    },
+    {
+      "epoch": 5.58282208588957,
+      "grad_norm": 0.18828590214252472,
+      "learning_rate": 9.978623571223045e-07,
+      "loss": 0.439,
+      "step": 910
+    },
+    {
+      "epoch": 5.588957055214724,
+      "grad_norm": 0.1946876496076584,
+      "learning_rate": 9.89537900179327e-07,
+      "loss": 0.4277,
+      "step": 911
+    },
+    {
+      "epoch": 5.595092024539877,
+      "grad_norm": 0.19324910640716553,
+      "learning_rate": 9.812444964104195e-07,
+      "loss": 0.4745,
+      "step": 912
+    },
+    {
+      "epoch": 5.601226993865031,
+      "grad_norm": 0.21109642088413239,
+      "learning_rate": 9.72982210031751e-07,
+      "loss": 0.4066,
+      "step": 913
+    },
+    {
+      "epoch": 5.6073619631901845,
+      "grad_norm": 0.2006876915693283,
+      "learning_rate": 9.647511050185475e-07,
+      "loss": 0.459,
+      "step": 914
+    },
+    {
+      "epoch": 5.613496932515337,
+      "grad_norm": 0.21445654332637787,
+      "learning_rate": 9.56551245104596e-07,
+      "loss": 0.4202,
+      "step": 915
+    },
+    {
+      "epoch": 5.61963190184049,
+      "grad_norm": 0.2019462138414383,
+      "learning_rate": 9.48382693781752e-07,
+      "loss": 0.4487,
+      "step": 916
+    },
+    {
+      "epoch": 5.625766871165644,
+      "grad_norm": 0.21154631674289703,
+      "learning_rate": 9.402455142994443e-07,
+      "loss": 0.4411,
+      "step": 917
+    },
+    {
+      "epoch": 5.631901840490798,
+      "grad_norm": 0.2066999077796936,
+      "learning_rate": 9.321397696641916e-07,
+      "loss": 0.4351,
+      "step": 918
+    },
+    {
+      "epoch": 5.638036809815951,
+      "grad_norm": 0.21491585671901703,
+      "learning_rate": 9.240655226391121e-07,
+      "loss": 0.4178,
+      "step": 919
+    },
+    {
+      "epoch": 5.644171779141105,
+      "grad_norm": 0.20312045514583588,
+      "learning_rate": 9.160228357434314e-07,
+      "loss": 0.4321,
+      "step": 920
+    },
+    {
+      "epoch": 5.6503067484662575,
+      "grad_norm": 0.20171020925045013,
+      "learning_rate": 9.080117712520087e-07,
+      "loss": 0.4243,
+      "step": 921
+    },
+    {
+      "epoch": 5.656441717791411,
+      "grad_norm": 0.19424133002758026,
+      "learning_rate": 9.000323911948483e-07,
+      "loss": 0.4096,
+      "step": 922
+    },
+    {
+      "epoch": 5.662576687116564,
+      "grad_norm": 0.20163863897323608,
+      "learning_rate": 8.920847573566204e-07,
+      "loss": 0.4381,
+      "step": 923
+    },
+    {
+      "epoch": 5.668711656441718,
+      "grad_norm": 0.1946740597486496,
+      "learning_rate": 8.841689312761837e-07,
+      "loss": 0.4318,
+      "step": 924
+    },
+    {
+      "epoch": 5.674846625766871,
+      "grad_norm": 0.19821976125240326,
+      "learning_rate": 8.762849742461044e-07,
+      "loss": 0.4143,
+      "step": 925
+    },
+    {
+      "epoch": 5.680981595092025,
+      "grad_norm": 0.1881324201822281,
+      "learning_rate": 8.6843294731219e-07,
+      "loss": 0.4343,
+      "step": 926
+    },
+    {
+      "epoch": 5.6871165644171775,
+      "grad_norm": 0.1956545114517212,
+      "learning_rate": 8.60612911273011e-07,
+      "loss": 0.4057,
+      "step": 927
+    },
+    {
+      "epoch": 5.693251533742331,
+      "grad_norm": 0.17811179161071777,
+      "learning_rate": 8.528249266794286e-07,
+      "loss": 0.4045,
+      "step": 928
+    },
+    {
+      "epoch": 5.699386503067485,
+      "grad_norm": 0.21344789862632751,
+      "learning_rate": 8.450690538341299e-07,
+      "loss": 0.3894,
+      "step": 929
+    },
+    {
+      "epoch": 5.705521472392638,
+      "grad_norm": 0.20089948177337646,
+      "learning_rate": 8.373453527911618e-07,
+      "loss": 0.4302,
+      "step": 930
+    },
+    {
+      "epoch": 5.711656441717792,
+      "grad_norm": 0.19986341893672943,
+      "learning_rate": 8.29653883355458e-07,
+      "loss": 0.4502,
+      "step": 931
+    },
+    {
+      "epoch": 5.717791411042945,
+      "grad_norm": 0.21868270635604858,
+      "learning_rate": 8.219947050823862e-07,
+      "loss": 0.4134,
+      "step": 932
+    },
+    {
+      "epoch": 5.723926380368098,
+      "grad_norm": 0.17949172854423523,
+      "learning_rate": 8.143678772772811e-07,
+      "loss": 0.4315,
+      "step": 933
+    },
+    {
+      "epoch": 5.730061349693251,
+      "grad_norm": 0.20468908548355103,
+      "learning_rate": 8.06773458994986e-07,
+      "loss": 0.4446,
+      "step": 934
+    },
+    {
+      "epoch": 5.736196319018405,
+      "grad_norm": 0.21158595383167267,
+      "learning_rate": 7.99211509039397e-07,
+      "loss": 0.4719,
+      "step": 935
+    },
+    {
+      "epoch": 5.742331288343558,
+      "grad_norm": 0.22054874897003174,
+      "learning_rate": 7.916820859630031e-07,
+      "loss": 0.4244,
+      "step": 936
+    },
+    {
+      "epoch": 5.748466257668712,
+      "grad_norm": 0.2146049439907074,
+      "learning_rate": 7.841852480664414e-07,
+      "loss": 0.419,
+      "step": 937
+    },
+    {
+      "epoch": 5.754601226993865,
+      "grad_norm": 0.20754463970661163,
+      "learning_rate": 7.767210533980373e-07,
+      "loss": 0.4288,
+      "step": 938
+    },
+    {
+      "epoch": 5.7607361963190185,
+      "grad_norm": 0.2121291607618332,
+      "learning_rate": 7.692895597533584e-07,
+      "loss": 0.4149,
+      "step": 939
+    },
+    {
+      "epoch": 5.766871165644172,
+      "grad_norm": 0.21436981856822968,
+      "learning_rate": 7.618908246747686e-07,
+      "loss": 0.4888,
+      "step": 940
+    },
+    {
+      "epoch": 5.773006134969325,
+      "grad_norm": 0.20958930253982544,
+      "learning_rate": 7.54524905450979e-07,
+      "loss": 0.423,
+      "step": 941
+    },
+    {
+      "epoch": 5.779141104294479,
+      "grad_norm": 0.20395760238170624,
+      "learning_rate": 7.471918591166078e-07,
+      "loss": 0.4115,
+      "step": 942
+    },
+    {
+      "epoch": 5.785276073619632,
+      "grad_norm": 0.18619965016841888,
+      "learning_rate": 7.398917424517377e-07,
+      "loss": 0.4395,
+      "step": 943
+    },
+    {
+      "epoch": 5.791411042944786,
+      "grad_norm": 0.18316420912742615,
+      "learning_rate": 7.326246119814712e-07,
+      "loss": 0.4201,
+      "step": 944
+    },
+    {
+      "epoch": 5.7975460122699385,
+      "grad_norm": 0.18815039098262787,
+      "learning_rate": 7.253905239755021e-07,
+      "loss": 0.4224,
+      "step": 945
+    },
+    {
+      "epoch": 5.803680981595092,
+      "grad_norm": 0.2220267653465271,
+      "learning_rate": 7.181895344476747e-07,
+      "loss": 0.4289,
+      "step": 946
+    },
+    {
+      "epoch": 5.809815950920245,
+      "grad_norm": 0.20367726683616638,
+      "learning_rate": 7.110216991555457e-07,
+      "loss": 0.4518,
+      "step": 947
+    },
+    {
+      "epoch": 5.815950920245399,
+      "grad_norm": 0.2034081518650055,
+      "learning_rate": 7.038870735999631e-07,
+      "loss": 0.4344,
+      "step": 948
+    },
+    {
+      "epoch": 5.822085889570552,
+      "grad_norm": 0.1995425820350647,
+      "learning_rate": 6.96785713024628e-07,
+      "loss": 0.4201,
+      "step": 949
+    },
+    {
+      "epoch": 5.828220858895706,
+      "grad_norm": 0.2018972784280777,
+      "learning_rate": 6.897176724156663e-07,
+      "loss": 0.4358,
+      "step": 950
+    },
+    {
+      "epoch": 5.8343558282208585,
+      "grad_norm": 0.20849905908107758,
+      "learning_rate": 6.82683006501213e-07,
+      "loss": 0.4192,
+      "step": 951
+    },
+    {
+      "epoch": 5.840490797546012,
+      "grad_norm": 0.18639759719371796,
+      "learning_rate": 6.756817697509755e-07,
+      "loss": 0.4372,
+      "step": 952
+    },
+    {
+      "epoch": 5.846625766871165,
+      "grad_norm": 0.19128410518169403,
+      "learning_rate": 6.687140163758194e-07,
+      "loss": 0.4123,
+      "step": 953
+    },
+    {
+      "epoch": 5.852760736196319,
+      "grad_norm": 0.21019329130649567,
+      "learning_rate": 6.617798003273496e-07,
+      "loss": 0.4193,
+      "step": 954
+    },
+    {
+      "epoch": 5.858895705521473,
+      "grad_norm": 0.17338204383850098,
+      "learning_rate": 6.548791752974853e-07,
+      "loss": 0.4051,
+      "step": 955
+    },
+    {
+      "epoch": 5.865030674846626,
+      "grad_norm": 0.20529429614543915,
+      "learning_rate": 6.480121947180534e-07,
+      "loss": 0.4616,
+      "step": 956
+    },
+    {
+      "epoch": 5.871165644171779,
+      "grad_norm": 0.19057132303714752,
+      "learning_rate": 6.411789117603701e-07,
+      "loss": 0.423,
+      "step": 957
+    },
+    {
+      "epoch": 5.877300613496932,
+      "grad_norm": 0.1914457529783249,
+      "learning_rate": 6.343793793348247e-07,
+      "loss": 0.4017,
+      "step": 958
+    },
+    {
+      "epoch": 5.883435582822086,
+      "grad_norm": 0.20379380881786346,
+      "learning_rate": 6.276136500904823e-07,
+      "loss": 0.4191,
+      "step": 959
+    },
+    {
+      "epoch": 5.889570552147239,
+      "grad_norm": 0.19549894332885742,
+      "learning_rate": 6.208817764146596e-07,
+      "loss": 0.4201,
+      "step": 960
+    },
+    {
+      "epoch": 5.895705521472393,
+      "grad_norm": 0.2176375687122345,
+      "learning_rate": 6.141838104325376e-07,
+      "loss": 0.443,
+      "step": 961
+    },
+    {
+      "epoch": 5.901840490797546,
+      "grad_norm": 0.18703949451446533,
+      "learning_rate": 6.075198040067432e-07,
+      "loss": 0.4184,
+      "step": 962
+    },
+    {
+      "epoch": 5.9079754601226995,
+      "grad_norm": 0.19391953945159912,
+      "learning_rate": 6.00889808736953e-07,
+      "loss": 0.4547,
+      "step": 963
+    },
+    {
+      "epoch": 5.914110429447852,
+      "grad_norm": 0.21159960329532623,
+      "learning_rate": 5.942938759594952e-07,
+      "loss": 0.4461,
+      "step": 964
+    },
+    {
+      "epoch": 5.920245398773006,
+      "grad_norm": 0.17359314858913422,
+      "learning_rate": 5.877320567469514e-07,
+      "loss": 0.3967,
+      "step": 965
+    },
+    {
+      "epoch": 5.92638036809816,
+      "grad_norm": 0.2088865041732788,
+      "learning_rate": 5.812044019077578e-07,
+      "loss": 0.4448,
+      "step": 966
+    },
+    {
+      "epoch": 5.932515337423313,
+      "grad_norm": 0.20267169177532196,
+      "learning_rate": 5.747109619858176e-07,
+      "loss": 0.4304,
+      "step": 967
+    },
+    {
+      "epoch": 5.938650306748467,
+      "grad_norm": 0.22262024879455566,
+      "learning_rate": 5.682517872601034e-07,
+      "loss": 0.44,
+      "step": 968
+    },
+    {
+      "epoch": 5.9447852760736195,
+      "grad_norm": 0.20060719549655914,
+      "learning_rate": 5.618269277442723e-07,
+      "loss": 0.4505,
+      "step": 969
+    },
+    {
+      "epoch": 5.950920245398773,
+      "grad_norm": 0.20760180056095123,
+      "learning_rate": 5.554364331862799e-07,
+      "loss": 0.4116,
+      "step": 970
+    },
+    {
+      "epoch": 5.957055214723926,
+      "grad_norm": 0.19337566196918488,
+      "learning_rate": 5.490803530679883e-07,
+      "loss": 0.4193,
+      "step": 971
+    },
+    {
+      "epoch": 5.96319018404908,
+      "grad_norm": 0.17379286885261536,
+      "learning_rate": 5.427587366047893e-07,
+      "loss": 0.428,
+      "step": 972
+    },
+    {
+      "epoch": 5.969325153374233,
+      "grad_norm": 0.19079700112342834,
+      "learning_rate": 5.364716327452219e-07,
+      "loss": 0.4439,
+      "step": 973
+    },
+    {
+      "epoch": 5.975460122699387,
+      "grad_norm": 0.21255679428577423,
+      "learning_rate": 5.3021909017059e-07,
+      "loss": 0.4143,
+      "step": 974
+    },
+    {
+      "epoch": 5.9815950920245395,
+      "grad_norm": 0.2059745490550995,
+      "learning_rate": 5.240011572945896e-07,
+      "loss": 0.4334,
+      "step": 975
+    },
+    {
+      "epoch": 5.987730061349693,
+      "grad_norm": 0.19311602413654327,
+      "learning_rate": 5.178178822629348e-07,
+      "loss": 0.421,
+      "step": 976
+    },
+    {
+      "epoch": 5.993865030674847,
+      "grad_norm": 0.20050852000713348,
+      "learning_rate": 5.11669312952977e-07,
+      "loss": 0.4219,
+      "step": 977
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.19952666759490967,
+      "learning_rate": 5.05555496973344e-07,
+      "loss": 0.4253,
+      "step": 978
+    },
+    {
+      "epoch": 6.006134969325154,
+      "grad_norm": 0.19700153172016144,
+      "learning_rate": 4.994764816635666e-07,
+      "loss": 0.4002,
+      "step": 979
+    },
+    {
+      "epoch": 6.012269938650307,
+      "grad_norm": 0.19736388325691223,
+      "learning_rate": 4.934323140937125e-07,
+      "loss": 0.4436,
+      "step": 980
+    },
+    {
+      "epoch": 6.0184049079754605,
+      "grad_norm": 0.20198658108711243,
+      "learning_rate": 4.874230410640207e-07,
+      "loss": 0.4454,
+      "step": 981
+    },
+    {
+      "epoch": 6.024539877300613,
+      "grad_norm": 0.18840666115283966,
+      "learning_rate": 4.814487091045405e-07,
+      "loss": 0.448,
+      "step": 982
+    },
+    {
+      "epoch": 6.030674846625767,
+      "grad_norm": 0.18371936678886414,
+      "learning_rate": 4.7550936447477215e-07,
+      "loss": 0.3951,
+      "step": 983
+    },
+    {
+      "epoch": 6.03680981595092,
+      "grad_norm": 0.20922499895095825,
+      "learning_rate": 4.6960505316330783e-07,
+      "loss": 0.4124,
+      "step": 984
+    },
+    {
+      "epoch": 6.042944785276074,
+      "grad_norm": 0.1890077441930771,
+      "learning_rate": 4.637358208874726e-07,
+      "loss": 0.4318,
+      "step": 985
+    },
+    {
+      "epoch": 6.049079754601227,
+      "grad_norm": 0.1936153620481491,
+      "learning_rate": 4.579017130929775e-07,
+      "loss": 0.4279,
+      "step": 986
+    },
+    {
+      "epoch": 6.0552147239263805,
+      "grad_norm": 0.19301314651966095,
+      "learning_rate": 4.521027749535578e-07,
+      "loss": 0.4398,
+      "step": 987
+    },
+    {
+      "epoch": 6.061349693251533,
+      "grad_norm": 0.18920563161373138,
+      "learning_rate": 4.463390513706317e-07,
+      "loss": 0.4367,
+      "step": 988
+    },
+    {
+      "epoch": 6.067484662576687,
+      "grad_norm": 0.20188719034194946,
+      "learning_rate": 4.406105869729532e-07,
+      "loss": 0.4287,
+      "step": 989
+    },
+    {
+      "epoch": 6.07361963190184,
+      "grad_norm": 0.19266493618488312,
+      "learning_rate": 4.3491742611625587e-07,
+      "loss": 0.4405,
+      "step": 990
+    },
+    {
+      "epoch": 6.079754601226994,
+      "grad_norm": 0.1787448674440384,
+      "learning_rate": 4.292596128829207e-07,
+      "loss": 0.3901,
+      "step": 991
+    },
+    {
+      "epoch": 6.085889570552148,
+      "grad_norm": 0.20086398720741272,
+      "learning_rate": 4.2363719108163113e-07,
+      "loss": 0.3748,
+      "step": 992
+    },
+    {
+      "epoch": 6.0920245398773005,
+      "grad_norm": 0.2082991749048233,
+      "learning_rate": 4.1805020424703024e-07,
+      "loss": 0.4626,
+      "step": 993
+    },
+    {
+      "epoch": 6.098159509202454,
+      "grad_norm": 0.20025362074375153,
+      "learning_rate": 4.124986956393895e-07,
+      "loss": 0.466,
+      "step": 994
+    },
+    {
+      "epoch": 6.104294478527607,
+      "grad_norm": 0.18952858448028564,
+      "learning_rate": 4.0698270824426846e-07,
+      "loss": 0.4022,
+      "step": 995
+    },
+    {
+      "epoch": 6.110429447852761,
+      "grad_norm": 0.19780947268009186,
+      "learning_rate": 4.0150228477218664e-07,
+      "loss": 0.4298,
+      "step": 996
+    },
+    {
+      "epoch": 6.116564417177914,
+      "grad_norm": 0.19643534719944,
+      "learning_rate": 3.960574676582901e-07,
+      "loss": 0.4248,
+      "step": 997
+    },
+    {
+      "epoch": 6.122699386503068,
+      "grad_norm": 0.1984051913022995,
+      "learning_rate": 3.906482990620236e-07,
+      "loss": 0.4615,
+      "step": 998
+    },
+    {
+      "epoch": 6.128834355828221,
+      "grad_norm": 0.20450861752033234,
+      "learning_rate": 3.8527482086680277e-07,
+      "loss": 0.4138,
+      "step": 999
+    },
+    {
+      "epoch": 6.134969325153374,
+      "grad_norm": 0.1938575804233551,
+      "learning_rate": 3.7993707467969267e-07,
+      "loss": 0.4107,
+      "step": 1000
+    },
+    {
+      "epoch": 6.141104294478527,
+      "grad_norm": 0.19877435266971588,
+      "learning_rate": 3.746351018310812e-07,
+      "loss": 0.4304,
+      "step": 1001
+    },
+    {
+      "epoch": 6.147239263803681,
+      "grad_norm": 0.19607821106910706,
+      "learning_rate": 3.693689433743658e-07,
+      "loss": 0.4031,
+      "step": 1002
+    },
+    {
+      "epoch": 6.153374233128835,
+      "grad_norm": 0.208278626203537,
+      "learning_rate": 3.6413864008562785e-07,
+      "loss": 0.4144,
+      "step": 1003
+    },
+    {
+      "epoch": 6.159509202453988,
+      "grad_norm": 0.19486194849014282,
+      "learning_rate": 3.589442324633224e-07,
+      "loss": 0.4729,
+      "step": 1004
+    },
+    {
+      "epoch": 6.1656441717791415,
+      "grad_norm": 0.1959027498960495,
+      "learning_rate": 3.537857607279638e-07,
+      "loss": 0.4195,
+      "step": 1005
+    },
+    {
+      "epoch": 6.171779141104294,
+      "grad_norm": 0.2013474851846695,
+      "learning_rate": 3.4866326482181025e-07,
+      "loss": 0.4364,
+      "step": 1006
+    },
+    {
+      "epoch": 6.177914110429448,
+      "grad_norm": 0.19486162066459656,
+      "learning_rate": 3.4357678440856136e-07,
+      "loss": 0.413,
+      "step": 1007
+    },
+    {
+      "epoch": 6.184049079754601,
+      "grad_norm": 0.18348956108093262,
+      "learning_rate": 3.385263588730464e-07,
+      "loss": 0.427,
+      "step": 1008
+    },
+    {
+      "epoch": 6.190184049079755,
+      "grad_norm": 0.18834654986858368,
+      "learning_rate": 3.3351202732091693e-07,
+      "loss": 0.4004,
+      "step": 1009
+    },
+    {
+      "epoch": 6.196319018404908,
+      "grad_norm": 0.22560544312000275,
+      "learning_rate": 3.2853382857835124e-07,
+      "loss": 0.382,
+      "step": 1010
+    },
+    {
+      "epoch": 6.2024539877300615,
+      "grad_norm": 0.19283907115459442,
+      "learning_rate": 3.235918011917477e-07,
+      "loss": 0.4918,
+      "step": 1011
+    },
+    {
+      "epoch": 6.208588957055214,
+      "grad_norm": 0.21347446739673615,
+      "learning_rate": 3.186859834274292e-07,
+      "loss": 0.4492,
+      "step": 1012
+    },
+    {
+      "epoch": 6.214723926380368,
+      "grad_norm": 0.2025686800479889,
+      "learning_rate": 3.1381641327134616e-07,
+      "loss": 0.4489,
+      "step": 1013
+    },
+    {
+      "epoch": 6.220858895705521,
+      "grad_norm": 0.23214280605316162,
+      "learning_rate": 3.0898312842878144e-07,
+      "loss": 0.4648,
+      "step": 1014
+    },
+    {
+      "epoch": 6.226993865030675,
+      "grad_norm": 0.19760558009147644,
+      "learning_rate": 3.041861663240592e-07,
+      "loss": 0.4261,
+      "step": 1015
+    },
+    {
+      "epoch": 6.233128834355828,
+      "grad_norm": 0.18486978113651276,
+      "learning_rate": 2.994255641002564e-07,
+      "loss": 0.443,
+      "step": 1016
+    },
+    {
+      "epoch": 6.2392638036809815,
+      "grad_norm": 0.2192317545413971,
+      "learning_rate": 2.947013586189124e-07,
+      "loss": 0.4466,
+      "step": 1017
+    },
+    {
+      "epoch": 6.245398773006135,
+      "grad_norm": 0.22236081957817078,
+      "learning_rate": 2.9001358645974696e-07,
+      "loss": 0.401,
+      "step": 1018
+    },
+    {
+      "epoch": 6.251533742331288,
+      "grad_norm": 0.1953679621219635,
+      "learning_rate": 2.85362283920374e-07,
+      "loss": 0.403,
+      "step": 1019
+    },
+    {
+      "epoch": 6.257668711656442,
+      "grad_norm": 0.19780360162258148,
+      "learning_rate": 2.8074748701601984e-07,
+      "loss": 0.4384,
+      "step": 1020
+    },
+    {
+      "epoch": 6.263803680981595,
+      "grad_norm": 0.20466431975364685,
+      "learning_rate": 2.761692314792502e-07,
+      "loss": 0.4312,
+      "step": 1021
+    },
+    {
+      "epoch": 6.269938650306749,
+      "grad_norm": 0.18782320618629456,
+      "learning_rate": 2.7162755275968513e-07,
+      "loss": 0.4042,
+      "step": 1022
+    },
+    {
+      "epoch": 6.276073619631902,
+      "grad_norm": 0.1981232464313507,
+      "learning_rate": 2.6712248602373205e-07,
+      "loss": 0.4474,
+      "step": 1023
+    },
+    {
+      "epoch": 6.282208588957055,
+      "grad_norm": 0.2031874656677246,
+      "learning_rate": 2.626540661543103e-07,
+      "loss": 0.4364,
+      "step": 1024
+    },
+    {
+      "epoch": 6.288343558282208,
+      "grad_norm": 0.18200956284999847,
+      "learning_rate": 2.582223277505769e-07,
+      "loss": 0.4289,
+      "step": 1025
+    },
+    {
+      "epoch": 6.294478527607362,
+      "grad_norm": 0.18988868594169617,
+      "learning_rate": 2.538273051276685e-07,
+      "loss": 0.4263,
+      "step": 1026
+    },
+    {
+      "epoch": 6.300613496932515,
+      "grad_norm": 0.1987670212984085,
+      "learning_rate": 2.4946903231642727e-07,
+      "loss": 0.4316,
+      "step": 1027
+    },
+    {
+      "epoch": 6.306748466257669,
+      "grad_norm": 0.21598701179027557,
+      "learning_rate": 2.451475430631384e-07,
+      "loss": 0.4416,
+      "step": 1028
+    },
+    {
+      "epoch": 6.3128834355828225,
+      "grad_norm": 0.20041204988956451,
+      "learning_rate": 2.408628708292732e-07,
+      "loss": 0.4592,
+      "step": 1029
+    },
+    {
+      "epoch": 6.319018404907975,
+      "grad_norm": 0.22675997018814087,
+      "learning_rate": 2.3661504879122554e-07,
+      "loss": 0.4562,
+      "step": 1030
+    },
+    {
+      "epoch": 6.325153374233129,
+      "grad_norm": 0.2256935089826584,
+      "learning_rate": 2.3240410984005701e-07,
+      "loss": 0.4044,
+      "step": 1031
+    },
+    {
+      "epoch": 6.331288343558282,
+      "grad_norm": 0.21728554368019104,
+      "learning_rate": 2.2823008658124425e-07,
+      "loss": 0.3976,
+      "step": 1032
+    },
+    {
+      "epoch": 6.337423312883436,
+      "grad_norm": 0.19935831427574158,
+      "learning_rate": 2.2409301133441918e-07,
+      "loss": 0.4648,
+      "step": 1033
+    },
+    {
+      "epoch": 6.343558282208589,
+      "grad_norm": 0.1918664127588272,
+      "learning_rate": 2.1999291613312824e-07,
+      "loss": 0.4083,
+      "step": 1034
+    },
+    {
+      "epoch": 6.3496932515337425,
+      "grad_norm": 0.18631911277770996,
+      "learning_rate": 2.15929832724579e-07,
+      "loss": 0.4585,
+      "step": 1035
+    },
+    {
+      "epoch": 6.355828220858895,
+      "grad_norm": 0.20234107971191406,
+      "learning_rate": 2.1190379256939342e-07,
+      "loss": 0.4341,
+      "step": 1036
+    },
+    {
+      "epoch": 6.361963190184049,
+      "grad_norm": 0.20566298067569733,
+      "learning_rate": 2.0791482684136833e-07,
+      "loss": 0.4054,
+      "step": 1037
+    },
+    {
+      "epoch": 6.368098159509202,
+      "grad_norm": 0.19845980405807495,
+      "learning_rate": 2.0396296642722856e-07,
+      "loss": 0.4592,
+      "step": 1038
+    },
+    {
+      "epoch": 6.374233128834356,
+      "grad_norm": 0.19033271074295044,
+      "learning_rate": 2.0004824192639437e-07,
+      "loss": 0.4201,
+      "step": 1039
+    },
+    {
+      "epoch": 6.38036809815951,
+      "grad_norm": 0.20748300850391388,
+      "learning_rate": 1.9617068365073987e-07,
+      "loss": 0.4349,
+      "step": 1040
+    },
+    {
+      "epoch": 6.386503067484663,
+      "grad_norm": 0.19087129831314087,
+      "learning_rate": 1.9233032162435828e-07,
+      "loss": 0.4122,
+      "step": 1041
+    },
+    {
+      "epoch": 6.392638036809816,
+      "grad_norm": 0.2218639850616455,
+      "learning_rate": 1.885271855833315e-07,
+      "loss": 0.4133,
+      "step": 1042
+    },
+    {
+      "epoch": 6.398773006134969,
+      "grad_norm": 0.19954140484333038,
+      "learning_rate": 1.847613049755015e-07,
+      "loss": 0.4433,
+      "step": 1043
+    },
+    {
+      "epoch": 6.404907975460123,
+      "grad_norm": 0.18886399269104004,
+      "learning_rate": 1.8103270896023427e-07,
+      "loss": 0.3899,
+      "step": 1044
+    },
+    {
+      "epoch": 6.411042944785276,
+      "grad_norm": 0.20953959226608276,
+      "learning_rate": 1.7734142640820684e-07,
+      "loss": 0.4357,
+      "step": 1045
+    },
+    {
+      "epoch": 6.41717791411043,
+      "grad_norm": 0.20415045320987701,
+      "learning_rate": 1.7368748590117058e-07,
+      "loss": 0.434,
+      "step": 1046
+    },
+    {
+      "epoch": 6.423312883435583,
+      "grad_norm": 0.2486049383878708,
+      "learning_rate": 1.7007091573173818e-07,
+      "loss": 0.4239,
+      "step": 1047
+    },
+    {
+      "epoch": 6.429447852760736,
+      "grad_norm": 0.19526998698711395,
+      "learning_rate": 1.6649174390316325e-07,
+      "loss": 0.3995,
+      "step": 1048
+    },
+    {
+      "epoch": 6.435582822085889,
+      "grad_norm": 0.19858838617801666,
+      "learning_rate": 1.629499981291205e-07,
+      "loss": 0.4451,
+      "step": 1049
+    },
+    {
+      "epoch": 6.441717791411043,
+      "grad_norm": 0.1742226481437683,
+      "learning_rate": 1.5944570583349416e-07,
+      "loss": 0.4165,
+      "step": 1050
+    },
+    {
+      "epoch": 6.447852760736196,
+      "grad_norm": 0.20162968337535858,
+      "learning_rate": 1.5597889415016609e-07,
+      "loss": 0.4152,
+      "step": 1051
+    },
+    {
+      "epoch": 6.45398773006135,
+      "grad_norm": 0.22722604870796204,
+      "learning_rate": 1.5254958992280022e-07,
+      "loss": 0.4472,
+      "step": 1052
+    },
+    {
+      "epoch": 6.460122699386503,
+      "grad_norm": 0.20080749690532684,
+      "learning_rate": 1.4915781970464226e-07,
+      "loss": 0.4206,
+      "step": 1053
+    },
+    {
+      "epoch": 6.466257668711656,
+      "grad_norm": 0.20352134108543396,
+      "learning_rate": 1.4580360975830988e-07,
+      "loss": 0.4739,
+      "step": 1054
+    },
+    {
+      "epoch": 6.47239263803681,
+      "grad_norm": 0.2271064966917038,
+      "learning_rate": 1.4248698605558887e-07,
+      "loss": 0.4051,
+      "step": 1055
+    },
+    {
+      "epoch": 6.478527607361963,
+      "grad_norm": 0.21113221347332,
+      "learning_rate": 1.3920797427723454e-07,
+      "loss": 0.4652,
+      "step": 1056
+    },
+    {
+      "epoch": 6.484662576687117,
+      "grad_norm": 0.19622337818145752,
+      "learning_rate": 1.3596659981277016e-07,
+      "loss": 0.4079,
+      "step": 1057
+    },
+    {
+      "epoch": 6.49079754601227,
+      "grad_norm": 0.20004568994045258,
+      "learning_rate": 1.3276288776029267e-07,
+      "loss": 0.437,
+      "step": 1058
+    },
+    {
+      "epoch": 6.4969325153374236,
+      "grad_norm": 0.19560450315475464,
+      "learning_rate": 1.2959686292627782e-07,
+      "loss": 0.4172,
+      "step": 1059
+    },
+    {
+      "epoch": 6.5030674846625764,
+      "grad_norm": 0.20478709042072296,
+      "learning_rate": 1.2646854982538593e-07,
+      "loss": 0.4426,
+      "step": 1060
+    },
+    {
+      "epoch": 6.50920245398773,
+      "grad_norm": 0.20099417865276337,
+      "learning_rate": 1.2337797268027475e-07,
+      "loss": 0.4246,
+      "step": 1061
+    },
+    {
+      "epoch": 6.515337423312883,
+      "grad_norm": 0.19599393010139465,
+      "learning_rate": 1.2032515542141188e-07,
+      "loss": 0.4089,
+      "step": 1062
+    },
+    {
+      "epoch": 6.521472392638037,
+      "grad_norm": 0.20901450514793396,
+      "learning_rate": 1.1731012168688715e-07,
+      "loss": 0.4031,
+      "step": 1063
+    },
+    {
+      "epoch": 6.52760736196319,
+      "grad_norm": 0.20133835077285767,
+      "learning_rate": 1.1433289482223276e-07,
+      "loss": 0.429,
+      "step": 1064
+    },
+    {
+      "epoch": 6.533742331288344,
+      "grad_norm": 0.2047884315252304,
+      "learning_rate": 1.1139349788023779e-07,
+      "loss": 0.4024,
+      "step": 1065
+    },
+    {
+      "epoch": 6.539877300613497,
+      "grad_norm": 0.18399380147457123,
+      "learning_rate": 1.084919536207757e-07,
+      "loss": 0.4377,
+      "step": 1066
+    },
+    {
+      "epoch": 6.54601226993865,
+      "grad_norm": 0.18939992785453796,
+      "learning_rate": 1.0562828451062323e-07,
+      "loss": 0.4141,
+      "step": 1067
+    },
+    {
+      "epoch": 6.552147239263804,
+      "grad_norm": 0.21095708012580872,
+      "learning_rate": 1.0280251272328956e-07,
+      "loss": 0.4187,
+      "step": 1068
+    },
+    {
+      "epoch": 6.558282208588957,
+      "grad_norm": 0.20305517315864563,
+      "learning_rate": 1.0001466013884131e-07,
+      "loss": 0.4238,
+      "step": 1069
+    },
+    {
+      "epoch": 6.564417177914111,
+      "grad_norm": 0.20672884583473206,
+      "learning_rate": 9.726474834373833e-08,
+      "loss": 0.3993,
+      "step": 1070
+    },
+    {
+      "epoch": 6.570552147239264,
+      "grad_norm": 0.19539766013622284,
+      "learning_rate": 9.455279863065936e-08,
+      "loss": 0.422,
+      "step": 1071
+    },
+    {
+      "epoch": 6.576687116564417,
+      "grad_norm": 0.2007475346326828,
+      "learning_rate": 9.187883199834491e-08,
+      "loss": 0.3905,
+      "step": 1072
+    },
+    {
+      "epoch": 6.58282208588957,
+      "grad_norm": 0.2838771641254425,
+      "learning_rate": 8.924286915142854e-08,
+      "loss": 0.4672,
+      "step": 1073
+    },
+    {
+      "epoch": 6.588957055214724,
+      "grad_norm": 0.19780975580215454,
+      "learning_rate": 8.664493050028033e-08,
+      "loss": 0.4534,
+      "step": 1074
+    },
+    {
+      "epoch": 6.595092024539877,
+      "grad_norm": 0.18463049829006195,
+      "learning_rate": 8.40850361608464e-08,
+      "loss": 0.4187,
+      "step": 1075
+    },
+    {
+      "epoch": 6.601226993865031,
+      "grad_norm": 0.20436058938503265,
+      "learning_rate": 8.156320595449463e-08,
+      "loss": 0.4615,
+      "step": 1076
+    },
+    {
+      "epoch": 6.6073619631901845,
+      "grad_norm": 0.1944579929113388,
+      "learning_rate": 7.907945940786033e-08,
+      "loss": 0.4576,
+      "step": 1077
+    },
+    {
+      "epoch": 6.613496932515337,
+      "grad_norm": 0.20546551048755646,
+      "learning_rate": 7.66338157526969e-08,
+      "loss": 0.4644,
+      "step": 1078
+    },
+    {
+      "epoch": 6.61963190184049,
+      "grad_norm": 0.1846093386411667,
+      "learning_rate": 7.422629392572323e-08,
+      "loss": 0.4321,
+      "step": 1079
+    },
+    {
+      "epoch": 6.625766871165644,
+      "grad_norm": 0.1991308182477951,
+      "learning_rate": 7.185691256848093e-08,
+      "loss": 0.441,
+      "step": 1080
+    },
+    {
+      "epoch": 6.631901840490798,
+      "grad_norm": 0.1854790449142456,
+      "learning_rate": 6.95256900271879e-08,
+      "loss": 0.4384,
+      "step": 1081
+    },
+    {
+      "epoch": 6.638036809815951,
+      "grad_norm": 0.20065605640411377,
+      "learning_rate": 6.723264435259725e-08,
+      "loss": 0.4157,
+      "step": 1082
+    },
+    {
+      "epoch": 6.644171779141105,
+      "grad_norm": 0.1937716007232666,
+      "learning_rate": 6.497779329985631e-08,
+      "loss": 0.4157,
+      "step": 1083
+    },
+    {
+      "epoch": 6.6503067484662575,
+      "grad_norm": 0.18098995089530945,
+      "learning_rate": 6.27611543283696e-08,
+      "loss": 0.4376,
+      "step": 1084
+    },
+    {
+      "epoch": 6.656441717791411,
+      "grad_norm": 0.1942816972732544,
+      "learning_rate": 6.058274460166547e-08,
+      "loss": 0.4482,
+      "step": 1085
+    },
+    {
+      "epoch": 6.662576687116564,
+      "grad_norm": 0.1918404996395111,
+      "learning_rate": 5.8442580987259637e-08,
+      "loss": 0.41,
+      "step": 1086
+    },
+    {
+      "epoch": 6.668711656441718,
+      "grad_norm": 0.20013810694217682,
+      "learning_rate": 5.634068005652804e-08,
+      "loss": 0.383,
+      "step": 1087
+    },
+    {
+      "epoch": 6.674846625766871,
+      "grad_norm": 0.1878250539302826,
+      "learning_rate": 5.4277058084576394e-08,
+      "loss": 0.4637,
+      "step": 1088
+    },
+    {
+      "epoch": 6.680981595092025,
+      "grad_norm": 0.20763535797595978,
+      "learning_rate": 5.225173105011583e-08,
+      "loss": 0.4177,
+      "step": 1089
+    },
+    {
+      "epoch": 6.6871165644171775,
+      "grad_norm": 0.19868603348731995,
+      "learning_rate": 5.026471463533578e-08,
+      "loss": 0.405,
+      "step": 1090
+    },
+    {
+      "epoch": 6.693251533742331,
+      "grad_norm": 0.19065415859222412,
+      "learning_rate": 4.831602422578852e-08,
+      "loss": 0.3965,
+      "step": 1091
+    },
+    {
+      "epoch": 6.699386503067485,
+      "grad_norm": 0.2003355175256729,
+      "learning_rate": 4.640567491026316e-08,
+      "loss": 0.4255,
+      "step": 1092
+    },
+    {
+      "epoch": 6.705521472392638,
+      "grad_norm": 0.22053086757659912,
+      "learning_rate": 4.453368148067405e-08,
+      "loss": 0.4219,
+      "step": 1093
+    },
+    {
+      "epoch": 6.711656441717792,
+      "grad_norm": 0.18970713019371033,
+      "learning_rate": 4.2700058431943694e-08,
+      "loss": 0.445,
+      "step": 1094
+    },
+    {
+      "epoch": 6.717791411042945,
+      "grad_norm": 0.19756844639778137,
+      "learning_rate": 4.090481996189166e-08,
+      "loss": 0.4183,
+      "step": 1095
+    },
+    {
+      "epoch": 6.723926380368098,
+      "grad_norm": 0.20232686400413513,
+      "learning_rate": 3.914797997112307e-08,
+      "loss": 0.42,
+      "step": 1096
+    },
+    {
+      "epoch": 6.730061349693251,
+      "grad_norm": 0.1890253722667694,
+      "learning_rate": 3.7429552062923644e-08,
+      "loss": 0.4202,
+      "step": 1097
+    },
+    {
+      "epoch": 6.736196319018405,
+      "grad_norm": 0.18478119373321533,
+      "learning_rate": 3.574954954315091e-08,
+      "loss": 0.3997,
+      "step": 1098
+    },
+    {
+      "epoch": 6.742331288343558,
+      "grad_norm": 0.18402665853500366,
+      "learning_rate": 3.410798542013483e-08,
+      "loss": 0.4306,
+      "step": 1099
+    },
+    {
+      "epoch": 6.748466257668712,
+      "grad_norm": 0.2112373560667038,
+      "learning_rate": 3.25048724045729e-08,
+      "loss": 0.4423,
+      "step": 1100
+    },
+    {
+      "epoch": 6.754601226993865,
+      "grad_norm": 0.18628022074699402,
+      "learning_rate": 3.0940222909437434e-08,
+      "loss": 0.4172,
+      "step": 1101
+    },
+    {
+      "epoch": 6.7607361963190185,
+      "grad_norm": 0.18921811878681183,
+      "learning_rate": 2.9414049049872883e-08,
+      "loss": 0.4129,
+      "step": 1102
+    },
+    {
+      "epoch": 6.766871165644172,
+      "grad_norm": 0.20240849256515503,
+      "learning_rate": 2.7926362643106997e-08,
+      "loss": 0.4621,
+      "step": 1103
+    },
+    {
+      "epoch": 6.773006134969325,
+      "grad_norm": 0.18727080523967743,
+      "learning_rate": 2.647717520835702e-08,
+      "loss": 0.4166,
+      "step": 1104
+    },
+    {
+      "epoch": 6.779141104294479,
+      "grad_norm": 0.19762536883354187,
+      "learning_rate": 2.5066497966741987e-08,
+      "loss": 0.4223,
+      "step": 1105
+    },
+    {
+      "epoch": 6.785276073619632,
+      "grad_norm": 0.2059212028980255,
+      "learning_rate": 2.3694341841193902e-08,
+      "loss": 0.4323,
+      "step": 1106
+    },
+    {
+      "epoch": 6.791411042944786,
+      "grad_norm": 0.21039064228534698,
+      "learning_rate": 2.236071745637336e-08,
+      "loss": 0.4219,
+      "step": 1107
+    },
+    {
+      "epoch": 6.7975460122699385,
+      "grad_norm": 0.1902393400669098,
+      "learning_rate": 2.1065635138590724e-08,
+      "loss": 0.4157,
+      "step": 1108
+    },
+    {
+      "epoch": 6.803680981595092,
+      "grad_norm": 0.20540878176689148,
+      "learning_rate": 1.980910491572119e-08,
+      "loss": 0.4339,
+      "step": 1109
+    },
+    {
+      "epoch": 6.809815950920245,
+      "grad_norm": 0.22091785073280334,
+      "learning_rate": 1.8591136517130404e-08,
+      "loss": 0.4306,
+      "step": 1110
+    },
+    {
+      "epoch": 6.815950920245399,
+      "grad_norm": 0.20457226037979126,
+      "learning_rate": 1.7411739373598413e-08,
+      "loss": 0.4332,
+      "step": 1111
+    },
+    {
+      "epoch": 6.822085889570552,
+      "grad_norm": 0.23524275422096252,
+      "learning_rate": 1.627092261724583e-08,
+      "loss": 0.4358,
+      "step": 1112
+    },
+    {
+      "epoch": 6.828220858895706,
+      "grad_norm": 0.18031160533428192,
+      "learning_rate": 1.5168695081463904e-08,
+      "loss": 0.4402,
+      "step": 1113
+    },
+    {
+      "epoch": 6.8343558282208585,
+      "grad_norm": 0.21345287561416626,
+      "learning_rate": 1.410506530084621e-08,
+      "loss": 0.4317,
+      "step": 1114
+    },
+    {
+      "epoch": 6.840490797546012,
+      "grad_norm": 0.2095663696527481,
+      "learning_rate": 1.3080041511122077e-08,
+      "loss": 0.4239,
+      "step": 1115
+    },
+    {
+      "epoch": 6.846625766871165,
+      "grad_norm": 0.2007758915424347,
+      "learning_rate": 1.2093631649093828e-08,
+      "loss": 0.3961,
+      "step": 1116
+    },
+    {
+      "epoch": 6.852760736196319,
+      "grad_norm": 0.20119042694568634,
+      "learning_rate": 1.1145843352572406e-08,
+      "loss": 0.441,
+      "step": 1117
+    },
+    {
+      "epoch": 6.858895705521473,
+      "grad_norm": 0.18511272966861725,
+      "learning_rate": 1.0236683960323512e-08,
+      "loss": 0.4392,
+      "step": 1118
+    },
+    {
+      "epoch": 6.865030674846626,
+      "grad_norm": 0.198281928896904,
+      "learning_rate": 9.36616051200434e-09,
+      "loss": 0.4244,
+      "step": 1119
+    },
+    {
+      "epoch": 6.871165644171779,
+      "grad_norm": 0.18932987749576569,
+      "learning_rate": 8.53427974811527e-09,
+      "loss": 0.4387,
+      "step": 1120
+    },
+    {
+      "epoch": 6.877300613496932,
+      "grad_norm": 0.19044747948646545,
+      "learning_rate": 7.74104810994325e-09,
+      "loss": 0.4252,
+      "step": 1121
+    },
+    {
+      "epoch": 6.883435582822086,
+      "grad_norm": 0.1831924319267273,
+      "learning_rate": 6.986471739513501e-09,
+      "loss": 0.4281,
+      "step": 1122
+    },
+    {
+      "epoch": 6.889570552147239,
+      "grad_norm": 0.18994773924350739,
+      "learning_rate": 6.270556479541778e-09,
+      "loss": 0.4423,
+      "step": 1123
+    },
+    {
+      "epoch": 6.895705521472393,
+      "grad_norm": 0.20347394049167633,
+      "learning_rate": 5.593307873389963e-09,
+      "loss": 0.4119,
+      "step": 1124
+    },
+    {
+      "epoch": 6.901840490797546,
+      "grad_norm": 0.19256287813186646,
+      "learning_rate": 4.954731165022209e-09,
+      "loss": 0.4218,
+      "step": 1125
+    },
+    {
+      "epoch": 6.9079754601226995,
+      "grad_norm": 0.2047315537929535,
+      "learning_rate": 4.354831298963858e-09,
+      "loss": 0.428,
+      "step": 1126
+    },
+    {
+      "epoch": 6.914110429447852,
+      "grad_norm": 0.2375705987215042,
+      "learning_rate": 3.7936129202648106e-09,
+      "loss": 0.4402,
+      "step": 1127
+    },
+    {
+      "epoch": 6.920245398773006,
+      "grad_norm": 0.19552327692508698,
+      "learning_rate": 3.271080374462332e-09,
+      "loss": 0.4127,
+      "step": 1128
+    },
+    {
+      "epoch": 6.92638036809816,
+      "grad_norm": 0.22805239260196686,
+      "learning_rate": 2.7872377075471856e-09,
+      "loss": 0.4411,
+      "step": 1129
+    },
+    {
+      "epoch": 6.932515337423313,
+      "grad_norm": 0.2000674456357956,
+      "learning_rate": 2.3420886659331067e-09,
+      "loss": 0.4226,
+      "step": 1130
+    },
+    {
+      "epoch": 6.938650306748467,
+      "grad_norm": 0.22863265872001648,
+      "learning_rate": 1.9356366964279338e-09,
+      "loss": 0.4253,
+      "step": 1131
+    },
+    {
+      "epoch": 6.9447852760736195,
+      "grad_norm": 0.20869435369968414,
+      "learning_rate": 1.5678849462058554e-09,
+      "loss": 0.4374,
+      "step": 1132
+    },
+    {
+      "epoch": 6.950920245398773,
+      "grad_norm": 0.20107249915599823,
+      "learning_rate": 1.2388362627840934e-09,
+      "loss": 0.4414,
+      "step": 1133
+    },
+    {
+      "epoch": 6.957055214723926,
+      "grad_norm": 0.1913885623216629,
+      "learning_rate": 9.484931940001442e-10,
+      "loss": 0.4179,
+      "step": 1134
+    },
+    {
+      "epoch": 6.96319018404908,
+      "grad_norm": 0.19177880883216858,
+      "learning_rate": 6.968579879923498e-10,
+      "loss": 0.438,
+      "step": 1135
+    },
+    {
+      "epoch": 6.969325153374233,
+      "grad_norm": 0.21289180219173431,
+      "learning_rate": 4.8393259318269e-10,
+      "loss": 0.4331,
+      "step": 1136
+    },
+    {
+      "epoch": 6.975460122699387,
+      "grad_norm": 0.20402106642723083,
+      "learning_rate": 3.097186582606826e-10,
+      "loss": 0.4162,
+      "step": 1137
+    },
+    {
+      "epoch": 6.9815950920245395,
+      "grad_norm": 0.19780924916267395,
+      "learning_rate": 1.7421753217283788e-10,
+      "loss": 0.4331,
+      "step": 1138
+    },
+    {
+      "epoch": 6.987730061349693,
+      "grad_norm": 0.19273407757282257,
+      "learning_rate": 7.743026410989007e-11,
+      "loss": 0.3837,
+      "step": 1139
+    },
+    {
+      "epoch": 6.993865030674847,
+      "grad_norm": 0.19189181923866272,
+      "learning_rate": 1.9357603499026157e-11,
+      "loss": 0.4353,
+      "step": 1140
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.21704550087451935,
+      "learning_rate": 0.0,
+      "loss": 0.4444,
+      "step": 1141
+    },
+    {
+      "epoch": 7.0,
+      "step": 1141,
+      "total_flos": 2.996752297507999e+19,
+      "train_loss": 0.44955052823272323,
+      "train_runtime": 45593.8747,
+      "train_samples_per_second": 3.202,
+      "train_steps_per_second": 0.025
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1141,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 7,
+  "save_steps": 1000000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.996752297507999e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}