diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15013 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2519106300508283,
+  "eval_steps": 500,
+  "global_step": 2140,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005850733170000365,
+      "grad_norm": 0.3784801935906415,
+      "learning_rate": 7.999998479354072e-06,
+      "loss": 1.1815,
+      "step": 1
+    },
+    {
+      "epoch": 0.001170146634000073,
+      "grad_norm": 0.10369689761982695,
+      "learning_rate": 7.999993917417574e-06,
+      "loss": 1.0085,
+      "step": 2
+    },
+    {
+      "epoch": 0.0017552199510001097,
+      "grad_norm": 0.08165678751844849,
+      "learning_rate": 7.99998631419436e-06,
+      "loss": 0.934,
+      "step": 3
+    },
+    {
+      "epoch": 0.002340293268000146,
+      "grad_norm": 0.06369675215202451,
+      "learning_rate": 7.999975669690855e-06,
+      "loss": 1.0323,
+      "step": 4
+    },
+    {
+      "epoch": 0.002925366585000183,
+      "grad_norm": 0.08076101376684344,
+      "learning_rate": 7.999961983916048e-06,
+      "loss": 1.0676,
+      "step": 5
+    },
+    {
+      "epoch": 0.0035104399020002195,
+      "grad_norm": 0.06288930819751326,
+      "learning_rate": 7.999945256881502e-06,
+      "loss": 1.0421,
+      "step": 6
+    },
+    {
+      "epoch": 0.004095513219000256,
+      "grad_norm": 0.06851051488443692,
+      "learning_rate": 7.999925488601348e-06,
+      "loss": 0.9544,
+      "step": 7
+    },
+    {
+      "epoch": 0.004680586536000292,
+      "grad_norm": 0.06007707301663692,
+      "learning_rate": 7.999902679092287e-06,
+      "loss": 1.0349,
+      "step": 8
+    },
+    {
+      "epoch": 0.005265659853000329,
+      "grad_norm": 0.07674940091671953,
+      "learning_rate": 7.99987682837359e-06,
+      "loss": 1.1809,
+      "step": 9
+    },
+    {
+      "epoch": 0.005850733170000366,
+      "grad_norm": 0.06356726822529891,
+      "learning_rate": 7.999847936467093e-06,
+      "loss": 0.9502,
+      "step": 10
+    },
+    {
+      "epoch": 0.006435806487000402,
+      "grad_norm": 0.060093989725189235,
+      "learning_rate": 7.999816003397203e-06,
+      "loss": 1.0007,
+      "step": 11
+    },
+    {
+      "epoch": 0.007020879804000439,
+      "grad_norm": 0.04797598057222403,
+      "learning_rate": 7.999781029190902e-06,
+      "loss": 1.0316,
+      "step": 12
+    },
+    {
+      "epoch": 0.007605953121000476,
+      "grad_norm": 0.08827709210231136,
+      "learning_rate": 7.999743013877734e-06,
+      "loss": 1.0688,
+      "step": 13
+    },
+    {
+      "epoch": 0.008191026438000511,
+      "grad_norm": 0.04993529689396162,
+      "learning_rate": 7.999701957489811e-06,
+      "loss": 0.9695,
+      "step": 14
+    },
+    {
+      "epoch": 0.008776099755000549,
+      "grad_norm": 0.053685150458545855,
+      "learning_rate": 7.999657860061823e-06,
+      "loss": 1.0826,
+      "step": 15
+    },
+    {
+      "epoch": 0.009361173072000585,
+      "grad_norm": 0.04357397086005245,
+      "learning_rate": 7.99961072163102e-06,
+      "loss": 0.9788,
+      "step": 16
+    },
+    {
+      "epoch": 0.009946246389000622,
+      "grad_norm": 0.051838589957396514,
+      "learning_rate": 7.999560542237226e-06,
+      "loss": 1.0222,
+      "step": 17
+    },
+    {
+      "epoch": 0.010531319706000658,
+      "grad_norm": 0.04948384582310689,
+      "learning_rate": 7.999507321922833e-06,
+      "loss": 0.9882,
+      "step": 18
+    },
+    {
+      "epoch": 0.011116393023000696,
+      "grad_norm": 0.05488935540485584,
+      "learning_rate": 7.9994510607328e-06,
+      "loss": 1.0124,
+      "step": 19
+    },
+    {
+      "epoch": 0.011701466340000731,
+      "grad_norm": 0.04632979465330071,
+      "learning_rate": 7.999391758714659e-06,
+      "loss": 0.9646,
+      "step": 20
+    },
+    {
+      "epoch": 0.012286539657000767,
+      "grad_norm": 0.04102257708769348,
+      "learning_rate": 7.999329415918508e-06,
+      "loss": 0.8946,
+      "step": 21
+    },
+    {
+      "epoch": 0.012871612974000805,
+      "grad_norm": 0.045466366302330634,
+      "learning_rate": 7.999264032397013e-06,
+      "loss": 0.9973,
+      "step": 22
+    },
+    {
+      "epoch": 0.01345668629100084,
+      "grad_norm": 0.04440538659368297,
+      "learning_rate": 7.99919560820541e-06,
+      "loss": 1.0596,
+      "step": 23
+    },
+    {
+      "epoch": 0.014041759608000878,
+      "grad_norm": 0.042500986542824803,
+      "learning_rate": 7.999124143401507e-06,
+      "loss": 0.9303,
+      "step": 24
+    },
+    {
+      "epoch": 0.014626832925000914,
+      "grad_norm": 0.05180121412922999,
+      "learning_rate": 7.999049638045675e-06,
+      "loss": 0.9412,
+      "step": 25
+    },
+    {
+      "epoch": 0.015211906242000951,
+      "grad_norm": 0.06441302150178038,
+      "learning_rate": 7.998972092200859e-06,
+      "loss": 0.9986,
+      "step": 26
+    },
+    {
+      "epoch": 0.015796979559000987,
+      "grad_norm": 0.05297417453471626,
+      "learning_rate": 7.998891505932565e-06,
+      "loss": 1.0701,
+      "step": 27
+    },
+    {
+      "epoch": 0.016382052876001023,
+      "grad_norm": 0.04305692283062352,
+      "learning_rate": 7.998807879308878e-06,
+      "loss": 0.9823,
+      "step": 28
+    },
+    {
+      "epoch": 0.016967126193001062,
+      "grad_norm": 0.048331091733781785,
+      "learning_rate": 7.998721212400443e-06,
+      "loss": 0.9145,
+      "step": 29
+    },
+    {
+      "epoch": 0.017552199510001098,
+      "grad_norm": 0.04249327256576877,
+      "learning_rate": 7.998631505280477e-06,
+      "loss": 0.9469,
+      "step": 30
+    },
+    {
+      "epoch": 0.018137272827001134,
+      "grad_norm": 0.042672714299481426,
+      "learning_rate": 7.998538758024765e-06,
+      "loss": 0.9814,
+      "step": 31
+    },
+    {
+      "epoch": 0.01872234614400117,
+      "grad_norm": 0.04643044969291229,
+      "learning_rate": 7.998442970711661e-06,
+      "loss": 0.9342,
+      "step": 32
+    },
+    {
+      "epoch": 0.019307419461001205,
+      "grad_norm": 0.042411147754236,
+      "learning_rate": 7.998344143422087e-06,
+      "loss": 0.8851,
+      "step": 33
+    },
+    {
+      "epoch": 0.019892492778001244,
+      "grad_norm": 0.04407093546490425,
+      "learning_rate": 7.998242276239527e-06,
+      "loss": 1.0026,
+      "step": 34
+    },
+    {
+      "epoch": 0.02047756609500128,
+      "grad_norm": 0.044039695433035646,
+      "learning_rate": 7.998137369250046e-06,
+      "loss": 0.9637,
+      "step": 35
+    },
+    {
+      "epoch": 0.021062639412001316,
+      "grad_norm": 0.04258725681727216,
+      "learning_rate": 7.998029422542267e-06,
+      "loss": 0.9249,
+      "step": 36
+    },
+    {
+      "epoch": 0.021647712729001352,
+      "grad_norm": 0.045486820543060684,
+      "learning_rate": 7.997918436207383e-06,
+      "loss": 1.005,
+      "step": 37
+    },
+    {
+      "epoch": 0.02223278604600139,
+      "grad_norm": 0.03843585783476988,
+      "learning_rate": 7.997804410339156e-06,
+      "loss": 0.9815,
+      "step": 38
+    },
+    {
+      "epoch": 0.022817859363001427,
+      "grad_norm": 0.05544665751989288,
+      "learning_rate": 7.997687345033915e-06,
+      "loss": 0.94,
+      "step": 39
+    },
+    {
+      "epoch": 0.023402932680001463,
+      "grad_norm": 0.041440042889966715,
+      "learning_rate": 7.99756724039056e-06,
+      "loss": 0.9337,
+      "step": 40
+    },
+    {
+      "epoch": 0.0239880059970015,
+      "grad_norm": 0.040659761048442974,
+      "learning_rate": 7.997444096510552e-06,
+      "loss": 0.897,
+      "step": 41
+    },
+    {
+      "epoch": 0.024573079314001534,
+      "grad_norm": 0.06541743477477353,
+      "learning_rate": 7.997317913497925e-06,
+      "loss": 0.9325,
+      "step": 42
+    },
+    {
+      "epoch": 0.025158152631001574,
+      "grad_norm": 0.04101226832374059,
+      "learning_rate": 7.997188691459279e-06,
+      "loss": 0.9323,
+      "step": 43
+    },
+    {
+      "epoch": 0.02574322594800161,
+      "grad_norm": 0.03954326036578227,
+      "learning_rate": 7.997056430503782e-06,
+      "loss": 0.8793,
+      "step": 44
+    },
+    {
+      "epoch": 0.026328299265001645,
+      "grad_norm": 0.0396268421897809,
+      "learning_rate": 7.996921130743165e-06,
+      "loss": 0.9849,
+      "step": 45
+    },
+    {
+      "epoch": 0.02691337258200168,
+      "grad_norm": 0.03855187149117414,
+      "learning_rate": 7.996782792291732e-06,
+      "loss": 0.9328,
+      "step": 46
+    },
+    {
+      "epoch": 0.02749844589900172,
+      "grad_norm": 0.03791722568157901,
+      "learning_rate": 7.996641415266355e-06,
+      "loss": 0.9293,
+      "step": 47
+    },
+    {
+      "epoch": 0.028083519216001756,
+      "grad_norm": 0.04497262810695708,
+      "learning_rate": 7.996496999786465e-06,
+      "loss": 0.9884,
+      "step": 48
+    },
+    {
+      "epoch": 0.02866859253300179,
+      "grad_norm": 0.041089653577607424,
+      "learning_rate": 7.996349545974065e-06,
+      "loss": 0.9305,
+      "step": 49
+    },
+    {
+      "epoch": 0.029253665850001827,
+      "grad_norm": 0.04162057876282252,
+      "learning_rate": 7.996199053953729e-06,
+      "loss": 1.0771,
+      "step": 50
+    },
+    {
+      "epoch": 0.029838739167001863,
+      "grad_norm": 0.045060287442874644,
+      "learning_rate": 7.996045523852587e-06,
+      "loss": 0.9819,
+      "step": 51
+    },
+    {
+      "epoch": 0.030423812484001903,
+      "grad_norm": 0.03950486400361273,
+      "learning_rate": 7.995888955800346e-06,
+      "loss": 0.8911,
+      "step": 52
+    },
+    {
+      "epoch": 0.03100888580100194,
+      "grad_norm": 0.040638774400984866,
+      "learning_rate": 7.995729349929275e-06,
+      "loss": 0.9289,
+      "step": 53
+    },
+    {
+      "epoch": 0.031593959118001974,
+      "grad_norm": 0.04034434081592363,
+      "learning_rate": 7.995566706374206e-06,
+      "loss": 0.9529,
+      "step": 54
+    },
+    {
+      "epoch": 0.03217903243500201,
+      "grad_norm": 0.04315614004870732,
+      "learning_rate": 7.995401025272545e-06,
+      "loss": 0.9554,
+      "step": 55
+    },
+    {
+      "epoch": 0.032764105752002046,
+      "grad_norm": 0.03862522293897866,
+      "learning_rate": 7.995232306764258e-06,
+      "loss": 0.9604,
+      "step": 56
+    },
+    {
+      "epoch": 0.03334917906900208,
+      "grad_norm": 0.042952485422922036,
+      "learning_rate": 7.995060550991879e-06,
+      "loss": 0.9731,
+      "step": 57
+    },
+    {
+      "epoch": 0.033934252386002124,
+      "grad_norm": 0.04399639647859174,
+      "learning_rate": 7.994885758100508e-06,
+      "loss": 0.868,
+      "step": 58
+    },
+    {
+      "epoch": 0.03451932570300216,
+      "grad_norm": 0.03794998627596821,
+      "learning_rate": 7.99470792823781e-06,
+      "loss": 0.9176,
+      "step": 59
+    },
+    {
+      "epoch": 0.035104399020002196,
+      "grad_norm": 0.046410465067380204,
+      "learning_rate": 7.99452706155402e-06,
+      "loss": 0.9456,
+      "step": 60
+    },
+    {
+      "epoch": 0.03568947233700223,
+      "grad_norm": 0.04773216743183266,
+      "learning_rate": 7.994343158201927e-06,
+      "loss": 1.0049,
+      "step": 61
+    },
+    {
+      "epoch": 0.03627454565400227,
+      "grad_norm": 0.04395331205981261,
+      "learning_rate": 7.994156218336901e-06,
+      "loss": 0.9197,
+      "step": 62
+    },
+    {
+      "epoch": 0.0368596189710023,
+      "grad_norm": 0.04353152162367831,
+      "learning_rate": 7.993966242116865e-06,
+      "loss": 0.8978,
+      "step": 63
+    },
+    {
+      "epoch": 0.03744469228800234,
+      "grad_norm": 0.041195331266407585,
+      "learning_rate": 7.993773229702312e-06,
+      "loss": 0.965,
+      "step": 64
+    },
+    {
+      "epoch": 0.038029765605002375,
+      "grad_norm": 0.050163343033375746,
+      "learning_rate": 7.993577181256304e-06,
+      "loss": 0.8876,
+      "step": 65
+    },
+    {
+      "epoch": 0.03861483892200241,
+      "grad_norm": 0.06660490581550026,
+      "learning_rate": 7.993378096944456e-06,
+      "loss": 0.9082,
+      "step": 66
+    },
+    {
+      "epoch": 0.03919991223900245,
+      "grad_norm": 0.03984407119317513,
+      "learning_rate": 7.99317597693496e-06,
+      "loss": 0.9233,
+      "step": 67
+    },
+    {
+      "epoch": 0.03978498555600249,
+      "grad_norm": 0.04866369054757869,
+      "learning_rate": 7.992970821398567e-06,
+      "loss": 0.9349,
+      "step": 68
+    },
+    {
+      "epoch": 0.040370058873002525,
+      "grad_norm": 0.04990567711730111,
+      "learning_rate": 7.99276263050859e-06,
+      "loss": 0.985,
+      "step": 69
+    },
+    {
+      "epoch": 0.04095513219000256,
+      "grad_norm": 0.17411986574945643,
+      "learning_rate": 7.992551404440914e-06,
+      "loss": 1.0208,
+      "step": 70
+    },
+    {
+      "epoch": 0.041540205507002596,
+      "grad_norm": 0.04353986815925501,
+      "learning_rate": 7.992337143373981e-06,
+      "loss": 0.9633,
+      "step": 71
+    },
+    {
+      "epoch": 0.04212527882400263,
+      "grad_norm": 0.05318272939257367,
+      "learning_rate": 7.9921198474888e-06,
+      "loss": 1.0015,
+      "step": 72
+    },
+    {
+      "epoch": 0.04271035214100267,
+      "grad_norm": 0.059411931897509304,
+      "learning_rate": 7.991899516968942e-06,
+      "loss": 1.0224,
+      "step": 73
+    },
+    {
+      "epoch": 0.043295425458002704,
+      "grad_norm": 0.05746696657441689,
+      "learning_rate": 7.991676152000545e-06,
+      "loss": 0.9817,
+      "step": 74
+    },
+    {
+      "epoch": 0.04388049877500274,
+      "grad_norm": 0.03994063377968752,
+      "learning_rate": 7.991449752772307e-06,
+      "loss": 0.899,
+      "step": 75
+    },
+    {
+      "epoch": 0.04446557209200278,
+      "grad_norm": 0.05114807327866483,
+      "learning_rate": 7.991220319475492e-06,
+      "loss": 0.9627,
+      "step": 76
+    },
+    {
+      "epoch": 0.04505064540900282,
+      "grad_norm": 0.043413594203031196,
+      "learning_rate": 7.990987852303923e-06,
+      "loss": 0.9385,
+      "step": 77
+    },
+    {
+      "epoch": 0.045635718726002854,
+      "grad_norm": 0.044141893159488445,
+      "learning_rate": 7.990752351453994e-06,
+      "loss": 0.9214,
+      "step": 78
+    },
+    {
+      "epoch": 0.04622079204300289,
+      "grad_norm": 0.07098639926950194,
+      "learning_rate": 7.990513817124652e-06,
+      "loss": 0.9762,
+      "step": 79
+    },
+    {
+      "epoch": 0.046805865360002925,
+      "grad_norm": 0.04487096138718826,
+      "learning_rate": 7.990272249517416e-06,
+      "loss": 0.9379,
+      "step": 80
+    },
+    {
+      "epoch": 0.04739093867700296,
+      "grad_norm": 0.040488382771263605,
+      "learning_rate": 7.990027648836359e-06,
+      "loss": 0.9563,
+      "step": 81
+    },
+    {
+      "epoch": 0.047976011994003,
+      "grad_norm": 0.04352730030611419,
+      "learning_rate": 7.989780015288123e-06,
+      "loss": 0.9488,
+      "step": 82
+    },
+    {
+      "epoch": 0.04856108531100303,
+      "grad_norm": 0.04413441845817798,
+      "learning_rate": 7.98952934908191e-06,
+      "loss": 1.0336,
+      "step": 83
+    },
+    {
+      "epoch": 0.04914615862800307,
+      "grad_norm": 0.04193745680850997,
+      "learning_rate": 7.989275650429482e-06,
+      "loss": 0.8785,
+      "step": 84
+    },
+    {
+      "epoch": 0.04973123194500311,
+      "grad_norm": 0.04475381091812719,
+      "learning_rate": 7.989018919545165e-06,
+      "loss": 0.9443,
+      "step": 85
+    },
+    {
+      "epoch": 0.05031630526200315,
+      "grad_norm": 0.04234754821679888,
+      "learning_rate": 7.988759156645845e-06,
+      "loss": 0.9564,
+      "step": 86
+    },
+    {
+      "epoch": 0.05090137857900318,
+      "grad_norm": 0.054324472784088765,
+      "learning_rate": 7.988496361950972e-06,
+      "loss": 0.9824,
+      "step": 87
+    },
+    {
+      "epoch": 0.05148645189600322,
+      "grad_norm": 0.051967709927586946,
+      "learning_rate": 7.988230535682556e-06,
+      "loss": 0.914,
+      "step": 88
+    },
+    {
+      "epoch": 0.052071525213003254,
+      "grad_norm": 0.03996756018971234,
+      "learning_rate": 7.987961678065169e-06,
+      "loss": 0.9421,
+      "step": 89
+    },
+    {
+      "epoch": 0.05265659853000329,
+      "grad_norm": 0.07556612961446735,
+      "learning_rate": 7.987689789325939e-06,
+      "loss": 0.9791,
+      "step": 90
+    },
+    {
+      "epoch": 0.053241671847003326,
+      "grad_norm": 0.04557130060143385,
+      "learning_rate": 7.987414869694562e-06,
+      "loss": 0.9318,
+      "step": 91
+    },
+    {
+      "epoch": 0.05382674516400336,
+      "grad_norm": 0.0424700455078842,
+      "learning_rate": 7.98713691940329e-06,
+      "loss": 0.9745,
+      "step": 92
+    },
+    {
+      "epoch": 0.0544118184810034,
+      "grad_norm": 0.07879025661937863,
+      "learning_rate": 7.986855938686935e-06,
+      "loss": 0.9614,
+      "step": 93
+    },
+    {
+      "epoch": 0.05499689179800344,
+      "grad_norm": 0.04227267907542001,
+      "learning_rate": 7.986571927782871e-06,
+      "loss": 0.9317,
+      "step": 94
+    },
+    {
+      "epoch": 0.055581965115003476,
+      "grad_norm": 0.042275976377898476,
+      "learning_rate": 7.986284886931033e-06,
+      "loss": 0.8982,
+      "step": 95
+    },
+    {
+      "epoch": 0.05616703843200351,
+      "grad_norm": 0.0457764180229334,
+      "learning_rate": 7.985994816373913e-06,
+      "loss": 0.8803,
+      "step": 96
+    },
+    {
+      "epoch": 0.05675211174900355,
+      "grad_norm": 0.04305545472943831,
+      "learning_rate": 7.985701716356565e-06,
+      "loss": 0.9786,
+      "step": 97
+    },
+    {
+      "epoch": 0.05733718506600358,
+      "grad_norm": 0.045580685701444475,
+      "learning_rate": 7.985405587126597e-06,
+      "loss": 0.9036,
+      "step": 98
+    },
+    {
+      "epoch": 0.05792225838300362,
+      "grad_norm": 0.0442896526405325,
+      "learning_rate": 7.985106428934183e-06,
+      "loss": 0.9871,
+      "step": 99
+    },
+    {
+      "epoch": 0.058507331700003655,
+      "grad_norm": 0.04396583794537118,
+      "learning_rate": 7.984804242032051e-06,
+      "loss": 1.0145,
+      "step": 100
+    },
+    {
+      "epoch": 0.05909240501700369,
+      "grad_norm": 0.04488641679254391,
+      "learning_rate": 7.984499026675494e-06,
+      "loss": 0.9673,
+      "step": 101
+    },
+    {
+      "epoch": 0.05967747833400373,
+      "grad_norm": 0.0402258783966036,
+      "learning_rate": 7.984190783122351e-06,
+      "loss": 1.0458,
+      "step": 102
+    },
+    {
+      "epoch": 0.06026255165100377,
+      "grad_norm": 0.048502182555142354,
+      "learning_rate": 7.983879511633036e-06,
+      "loss": 0.8879,
+      "step": 103
+    },
+    {
+      "epoch": 0.060847624968003805,
+      "grad_norm": 0.04818474096979939,
+      "learning_rate": 7.983565212470504e-06,
+      "loss": 0.9467,
+      "step": 104
+    },
+    {
+      "epoch": 0.06143269828500384,
+      "grad_norm": 0.04176125713160911,
+      "learning_rate": 7.983247885900283e-06,
+      "loss": 0.9266,
+      "step": 105
+    },
+    {
+      "epoch": 0.06201777160200388,
+      "grad_norm": 0.04212530605033032,
+      "learning_rate": 7.982927532190447e-06,
+      "loss": 0.9179,
+      "step": 106
+    },
+    {
+      "epoch": 0.06260284491900392,
+      "grad_norm": 0.037198964746819226,
+      "learning_rate": 7.982604151611633e-06,
+      "loss": 1.0018,
+      "step": 107
+    },
+    {
+      "epoch": 0.06318791823600395,
+      "grad_norm": 0.04090799839826158,
+      "learning_rate": 7.982277744437035e-06,
+      "loss": 0.8756,
+      "step": 108
+    },
+    {
+      "epoch": 0.06377299155300399,
+      "grad_norm": 0.03857422273289164,
+      "learning_rate": 7.981948310942402e-06,
+      "loss": 0.8855,
+      "step": 109
+    },
+    {
+      "epoch": 0.06435806487000402,
+      "grad_norm": 0.05241752689478532,
+      "learning_rate": 7.981615851406039e-06,
+      "loss": 0.8862,
+      "step": 110
+    },
+    {
+      "epoch": 0.06494313818700406,
+      "grad_norm": 0.04086973369607326,
+      "learning_rate": 7.981280366108814e-06,
+      "loss": 0.9221,
+      "step": 111
+    },
+    {
+      "epoch": 0.06552821150400409,
+      "grad_norm": 0.03931043694281754,
+      "learning_rate": 7.98094185533414e-06,
+      "loss": 0.9417,
+      "step": 112
+    },
+    {
+      "epoch": 0.06611328482100413,
+      "grad_norm": 0.04169878931857935,
+      "learning_rate": 7.980600319367995e-06,
+      "loss": 0.958,
+      "step": 113
+    },
+    {
+      "epoch": 0.06669835813800416,
+      "grad_norm": 0.044230077430854955,
+      "learning_rate": 7.980255758498908e-06,
+      "loss": 0.9265,
+      "step": 114
+    },
+    {
+      "epoch": 0.0672834314550042,
+      "grad_norm": 0.04488148180330816,
+      "learning_rate": 7.979908173017968e-06,
+      "loss": 0.8908,
+      "step": 115
+    },
+    {
+      "epoch": 0.06786850477200425,
+      "grad_norm": 0.042038370275589085,
+      "learning_rate": 7.979557563218815e-06,
+      "loss": 0.8961,
+      "step": 116
+    },
+    {
+      "epoch": 0.06845357808900428,
+      "grad_norm": 0.04747064970378541,
+      "learning_rate": 7.979203929397646e-06,
+      "loss": 1.0609,
+      "step": 117
+    },
+    {
+      "epoch": 0.06903865140600432,
+      "grad_norm": 0.04392999493678844,
+      "learning_rate": 7.97884727185321e-06,
+      "loss": 0.9001,
+      "step": 118
+    },
+    {
+      "epoch": 0.06962372472300435,
+      "grad_norm": 0.040693633998808994,
+      "learning_rate": 7.978487590886814e-06,
+      "loss": 0.8562,
+      "step": 119
+    },
+    {
+      "epoch": 0.07020879804000439,
+      "grad_norm": 0.08337676841807191,
+      "learning_rate": 7.978124886802316e-06,
+      "loss": 0.9344,
+      "step": 120
+    },
+    {
+      "epoch": 0.07079387135700442,
+      "grad_norm": 0.03951441645023813,
+      "learning_rate": 7.977759159906134e-06,
+      "loss": 0.9182,
+      "step": 121
+    },
+    {
+      "epoch": 0.07137894467400446,
+      "grad_norm": 0.04427536962304041,
+      "learning_rate": 7.977390410507229e-06,
+      "loss": 0.9079,
+      "step": 122
+    },
+    {
+      "epoch": 0.07196401799100449,
+      "grad_norm": 0.047402666476443076,
+      "learning_rate": 7.977018638917126e-06,
+      "loss": 0.9442,
+      "step": 123
+    },
+    {
+      "epoch": 0.07254909130800453,
+      "grad_norm": 0.07806155365092546,
+      "learning_rate": 7.976643845449897e-06,
+      "loss": 0.9453,
+      "step": 124
+    },
+    {
+      "epoch": 0.07313416462500458,
+      "grad_norm": 0.04187989433422361,
+      "learning_rate": 7.97626603042217e-06,
+      "loss": 0.9762,
+      "step": 125
+    },
+    {
+      "epoch": 0.0737192379420046,
+      "grad_norm": 0.038153971027990764,
+      "learning_rate": 7.975885194153125e-06,
+      "loss": 0.9377,
+      "step": 126
+    },
+    {
+      "epoch": 0.07430431125900465,
+      "grad_norm": 0.04398811005912341,
+      "learning_rate": 7.975501336964492e-06,
+      "loss": 0.9655,
+      "step": 127
+    },
+    {
+      "epoch": 0.07488938457600468,
+      "grad_norm": 0.04310577216463342,
+      "learning_rate": 7.975114459180555e-06,
+      "loss": 0.9082,
+      "step": 128
+    },
+    {
+      "epoch": 0.07547445789300472,
+      "grad_norm": 0.04552741227747631,
+      "learning_rate": 7.97472456112815e-06,
+      "loss": 0.8667,
+      "step": 129
+    },
+    {
+      "epoch": 0.07605953121000475,
+      "grad_norm": 0.0406705187810207,
+      "learning_rate": 7.974331643136666e-06,
+      "loss": 0.9286,
+      "step": 130
+    },
+    {
+      "epoch": 0.07664460452700479,
+      "grad_norm": 0.16819457832404855,
+      "learning_rate": 7.973935705538039e-06,
+      "loss": 0.9724,
+      "step": 131
+    },
+    {
+      "epoch": 0.07722967784400482,
+      "grad_norm": 0.04524796652654962,
+      "learning_rate": 7.973536748666756e-06,
+      "loss": 0.961,
+      "step": 132
+    },
+    {
+      "epoch": 0.07781475116100486,
+      "grad_norm": 0.051171072191227115,
+      "learning_rate": 7.973134772859862e-06,
+      "loss": 0.9513,
+      "step": 133
+    },
+    {
+      "epoch": 0.0783998244780049,
+      "grad_norm": 0.044436153110071305,
+      "learning_rate": 7.972729778456946e-06,
+      "loss": 0.9363,
+      "step": 134
+    },
+    {
+      "epoch": 0.07898489779500494,
+      "grad_norm": 0.04485419817375143,
+      "learning_rate": 7.97232176580015e-06,
+      "loss": 0.8583,
+      "step": 135
+    },
+    {
+      "epoch": 0.07956997111200498,
+      "grad_norm": 0.044866876801222304,
+      "learning_rate": 7.971910735234161e-06,
+      "loss": 0.9859,
+      "step": 136
+    },
+    {
+      "epoch": 0.080155044429005,
+      "grad_norm": 0.03934716109284772,
+      "learning_rate": 7.971496687106219e-06,
+      "loss": 0.8592,
+      "step": 137
+    },
+    {
+      "epoch": 0.08074011774600505,
+      "grad_norm": 0.041852675369480244,
+      "learning_rate": 7.971079621766117e-06,
+      "loss": 0.9353,
+      "step": 138
+    },
+    {
+      "epoch": 0.08132519106300508,
+      "grad_norm": 0.3430745253576026,
+      "learning_rate": 7.97065953956619e-06,
+      "loss": 0.9602,
+      "step": 139
+    },
+    {
+      "epoch": 0.08191026438000512,
+      "grad_norm": 0.44783710464834237,
+      "learning_rate": 7.970236440861327e-06,
+      "loss": 0.9833,
+      "step": 140
+    },
+    {
+      "epoch": 0.08249533769700515,
+      "grad_norm": 0.29750847371388817,
+      "learning_rate": 7.96981032600896e-06,
+      "loss": 0.8244,
+      "step": 141
+    },
+    {
+      "epoch": 0.08308041101400519,
+      "grad_norm": 0.169443611740874,
+      "learning_rate": 7.969381195369076e-06,
+      "loss": 0.8983,
+      "step": 142
+    },
+    {
+      "epoch": 0.08366548433100524,
+      "grad_norm": 0.04111262427570532,
+      "learning_rate": 7.968949049304204e-06,
+      "loss": 0.9552,
+      "step": 143
+    },
+    {
+      "epoch": 0.08425055764800526,
+      "grad_norm": 0.04087231642049412,
+      "learning_rate": 7.968513888179421e-06,
+      "loss": 0.9051,
+      "step": 144
+    },
+    {
+      "epoch": 0.08483563096500531,
+      "grad_norm": 0.05663350951363164,
+      "learning_rate": 7.968075712362356e-06,
+      "loss": 0.8366,
+      "step": 145
+    },
+    {
+      "epoch": 0.08542070428200534,
+      "grad_norm": 0.04202895629977615,
+      "learning_rate": 7.967634522223179e-06,
+      "loss": 0.8115,
+      "step": 146
+    },
+    {
+      "epoch": 0.08600577759900538,
+      "grad_norm": 0.04238111014919017,
+      "learning_rate": 7.96719031813461e-06,
+      "loss": 0.9086,
+      "step": 147
+    },
+    {
+      "epoch": 0.08659085091600541,
+      "grad_norm": 0.04226501020383857,
+      "learning_rate": 7.966743100471913e-06,
+      "loss": 0.9286,
+      "step": 148
+    },
+    {
+      "epoch": 0.08717592423300545,
+      "grad_norm": 0.046325797442375215,
+      "learning_rate": 7.9662928696129e-06,
+      "loss": 0.9393,
+      "step": 149
+    },
+    {
+      "epoch": 0.08776099755000548,
+      "grad_norm": 0.05532068772615188,
+      "learning_rate": 7.965839625937926e-06,
+      "loss": 0.9202,
+      "step": 150
+    },
+    {
+      "epoch": 0.08834607086700552,
+      "grad_norm": 0.039360829289514745,
+      "learning_rate": 7.965383369829894e-06,
+      "loss": 0.908,
+      "step": 151
+    },
+    {
+      "epoch": 0.08893114418400556,
+      "grad_norm": 0.04965498407233415,
+      "learning_rate": 7.964924101674252e-06,
+      "loss": 0.9406,
+      "step": 152
+    },
+    {
+      "epoch": 0.0895162175010056,
+      "grad_norm": 0.04177674200968805,
+      "learning_rate": 7.964461821858987e-06,
+      "loss": 0.8933,
+      "step": 153
+    },
+    {
+      "epoch": 0.09010129081800564,
+      "grad_norm": 0.04711456314429998,
+      "learning_rate": 7.963996530774639e-06,
+      "loss": 1.0111,
+      "step": 154
+    },
+    {
+      "epoch": 0.09068636413500566,
+      "grad_norm": 0.045100723994096155,
+      "learning_rate": 7.963528228814285e-06,
+      "loss": 0.9806,
+      "step": 155
+    },
+    {
+      "epoch": 0.09127143745200571,
+      "grad_norm": 0.09448573408012474,
+      "learning_rate": 7.96305691637355e-06,
+      "loss": 0.9142,
+      "step": 156
+    },
+    {
+      "epoch": 0.09185651076900574,
+      "grad_norm": 0.04297907772876167,
+      "learning_rate": 7.962582593850596e-06,
+      "loss": 0.8852,
+      "step": 157
+    },
+    {
+      "epoch": 0.09244158408600578,
+      "grad_norm": 0.05151683390684187,
+      "learning_rate": 7.962105261646138e-06,
+      "loss": 0.9975,
+      "step": 158
+    },
+    {
+      "epoch": 0.09302665740300581,
+      "grad_norm": 0.07242957126413647,
+      "learning_rate": 7.961624920163423e-06,
+      "loss": 0.9196,
+      "step": 159
+    },
+    {
+      "epoch": 0.09361173072000585,
+      "grad_norm": 0.08840439035290122,
+      "learning_rate": 7.961141569808248e-06,
+      "loss": 0.92,
+      "step": 160
+    },
+    {
+      "epoch": 0.0941968040370059,
+      "grad_norm": 0.11047843778949552,
+      "learning_rate": 7.960655210988948e-06,
+      "loss": 0.9452,
+      "step": 161
+    },
+    {
+      "epoch": 0.09478187735400592,
+      "grad_norm": 0.04275273565605597,
+      "learning_rate": 7.960165844116399e-06,
+      "loss": 0.9641,
+      "step": 162
+    },
+    {
+      "epoch": 0.09536695067100597,
+      "grad_norm": 0.05662855927390397,
+      "learning_rate": 7.959673469604025e-06,
+      "loss": 0.9354,
+      "step": 163
+    },
+    {
+      "epoch": 0.095952023988006,
+      "grad_norm": 0.04769002643125012,
+      "learning_rate": 7.959178087867779e-06,
+      "loss": 0.9087,
+      "step": 164
+    },
+    {
+      "epoch": 0.09653709730500604,
+      "grad_norm": 0.050744023655463925,
+      "learning_rate": 7.958679699326164e-06,
+      "loss": 0.9561,
+      "step": 165
+    },
+    {
+      "epoch": 0.09712217062200607,
+      "grad_norm": 0.058451109924341715,
+      "learning_rate": 7.958178304400222e-06,
+      "loss": 0.9881,
+      "step": 166
+    },
+    {
+      "epoch": 0.09770724393900611,
+      "grad_norm": 0.04338399471073774,
+      "learning_rate": 7.95767390351353e-06,
+      "loss": 0.9705,
+      "step": 167
+    },
+    {
+      "epoch": 0.09829231725600614,
+      "grad_norm": 0.04901291931441639,
+      "learning_rate": 7.95716649709221e-06,
+      "loss": 0.9229,
+      "step": 168
+    },
+    {
+      "epoch": 0.09887739057300618,
+      "grad_norm": 0.044284928725944674,
+      "learning_rate": 7.95665608556492e-06,
+      "loss": 0.9493,
+      "step": 169
+    },
+    {
+      "epoch": 0.09946246389000622,
+      "grad_norm": 0.04914804115067185,
+      "learning_rate": 7.956142669362855e-06,
+      "loss": 0.9879,
+      "step": 170
+    },
+    {
+      "epoch": 0.10004753720700625,
+      "grad_norm": 0.04087870881565583,
+      "learning_rate": 7.955626248919752e-06,
+      "loss": 0.9435,
+      "step": 171
+    },
+    {
+      "epoch": 0.1006326105240063,
+      "grad_norm": 0.04938743203389444,
+      "learning_rate": 7.955106824671888e-06,
+      "loss": 0.8813,
+      "step": 172
+    },
+    {
+      "epoch": 0.10121768384100632,
+      "grad_norm": 0.04440441336302364,
+      "learning_rate": 7.95458439705807e-06,
+      "loss": 0.954,
+      "step": 173
+    },
+    {
+      "epoch": 0.10180275715800637,
+      "grad_norm": 0.04052619663535755,
+      "learning_rate": 7.954058966519649e-06,
+      "loss": 0.8759,
+      "step": 174
+    },
+    {
+      "epoch": 0.1023878304750064,
+      "grad_norm": 0.04315943430366373,
+      "learning_rate": 7.953530533500507e-06,
+      "loss": 0.8621,
+      "step": 175
+    },
+    {
+      "epoch": 0.10297290379200644,
+      "grad_norm": 0.07065437183104553,
+      "learning_rate": 7.952999098447072e-06,
+      "loss": 0.9796,
+      "step": 176
+    },
+    {
+      "epoch": 0.10355797710900647,
+      "grad_norm": 0.04286710979601013,
+      "learning_rate": 7.952464661808297e-06,
+      "loss": 0.9187,
+      "step": 177
+    },
+    {
+      "epoch": 0.10414305042600651,
+      "grad_norm": 0.044637305084684484,
+      "learning_rate": 7.951927224035678e-06,
+      "loss": 0.8772,
+      "step": 178
+    },
+    {
+      "epoch": 0.10472812374300655,
+      "grad_norm": 0.04369651860095562,
+      "learning_rate": 7.951386785583244e-06,
+      "loss": 0.8969,
+      "step": 179
+    },
+    {
+      "epoch": 0.10531319706000658,
+      "grad_norm": 0.05048322425196143,
+      "learning_rate": 7.950843346907559e-06,
+      "loss": 0.8907,
+      "step": 180
+    },
+    {
+      "epoch": 0.10589827037700662,
+      "grad_norm": 0.03884181381210202,
+      "learning_rate": 7.95029690846772e-06,
+      "loss": 0.931,
+      "step": 181
+    },
+    {
+      "epoch": 0.10648334369400665,
+      "grad_norm": 0.05060670376656713,
+      "learning_rate": 7.949747470725362e-06,
+      "loss": 0.8624,
+      "step": 182
+    },
+    {
+      "epoch": 0.1070684170110067,
+      "grad_norm": 0.04162121727818916,
+      "learning_rate": 7.949195034144653e-06,
+      "loss": 0.9141,
+      "step": 183
+    },
+    {
+      "epoch": 0.10765349032800672,
+      "grad_norm": 0.04729330740672752,
+      "learning_rate": 7.94863959919229e-06,
+      "loss": 0.921,
+      "step": 184
+    },
+    {
+      "epoch": 0.10823856364500677,
+      "grad_norm": 0.04459204791909808,
+      "learning_rate": 7.948081166337509e-06,
+      "loss": 0.8993,
+      "step": 185
+    },
+    {
+      "epoch": 0.1088236369620068,
+      "grad_norm": 0.04645986236352756,
+      "learning_rate": 7.947519736052075e-06,
+      "loss": 0.9158,
+      "step": 186
+    },
+    {
+      "epoch": 0.10940871027900684,
+      "grad_norm": 0.04275967646092167,
+      "learning_rate": 7.946955308810285e-06,
+      "loss": 0.9387,
+      "step": 187
+    },
+    {
+      "epoch": 0.10999378359600688,
+      "grad_norm": 0.07421648959437231,
+      "learning_rate": 7.94638788508897e-06,
+      "loss": 0.8497,
+      "step": 188
+    },
+    {
+      "epoch": 0.11057885691300691,
+      "grad_norm": 0.04955170539289863,
+      "learning_rate": 7.945817465367493e-06,
+      "loss": 0.8525,
+      "step": 189
+    },
+    {
+      "epoch": 0.11116393023000695,
+      "grad_norm": 0.12013696366466048,
+      "learning_rate": 7.945244050127744e-06,
+      "loss": 0.9616,
+      "step": 190
+    },
+    {
+      "epoch": 0.11174900354700698,
+      "grad_norm": 0.040829512145039985,
+      "learning_rate": 7.944667639854148e-06,
+      "loss": 0.8344,
+      "step": 191
+    },
+    {
+      "epoch": 0.11233407686400702,
+      "grad_norm": 0.05068364591326354,
+      "learning_rate": 7.944088235033657e-06,
+      "loss": 0.9403,
+      "step": 192
+    },
+    {
+      "epoch": 0.11291915018100705,
+      "grad_norm": 0.0499939002215986,
+      "learning_rate": 7.943505836155753e-06,
+      "loss": 0.9475,
+      "step": 193
+    },
+    {
+      "epoch": 0.1135042234980071,
+      "grad_norm": 0.05407026250866459,
+      "learning_rate": 7.94292044371245e-06,
+      "loss": 0.9101,
+      "step": 194
+    },
+    {
+      "epoch": 0.11408929681500712,
+      "grad_norm": 0.06417314751489454,
+      "learning_rate": 7.94233205819829e-06,
+      "loss": 0.8787,
+      "step": 195
+    },
+    {
+      "epoch": 0.11467437013200717,
+      "grad_norm": 0.03871241656337873,
+      "learning_rate": 7.941740680110343e-06,
+      "loss": 0.9059,
+      "step": 196
+    },
+    {
+      "epoch": 0.11525944344900721,
+      "grad_norm": 0.041501526897382096,
+      "learning_rate": 7.941146309948205e-06,
+      "loss": 0.8946,
+      "step": 197
+    },
+    {
+      "epoch": 0.11584451676600724,
+      "grad_norm": 0.037938175729775744,
+      "learning_rate": 7.940548948214005e-06,
+      "loss": 0.8534,
+      "step": 198
+    },
+    {
+      "epoch": 0.11642959008300728,
+      "grad_norm": 0.038697315108935856,
+      "learning_rate": 7.939948595412394e-06,
+      "loss": 0.9671,
+      "step": 199
+    },
+    {
+      "epoch": 0.11701466340000731,
+      "grad_norm": 0.040128164685532784,
+      "learning_rate": 7.939345252050552e-06,
+      "loss": 0.9452,
+      "step": 200
+    },
+    {
+      "epoch": 0.11759973671700735,
+      "grad_norm": 0.04159138776071103,
+      "learning_rate": 7.938738918638187e-06,
+      "loss": 0.9525,
+      "step": 201
+    },
+    {
+      "epoch": 0.11818481003400738,
+      "grad_norm": 0.044356876333774484,
+      "learning_rate": 7.93812959568753e-06,
+      "loss": 0.8863,
+      "step": 202
+    },
+    {
+      "epoch": 0.11876988335100742,
+      "grad_norm": 0.04110203273326361,
+      "learning_rate": 7.93751728371334e-06,
+      "loss": 0.8585,
+      "step": 203
+    },
+    {
+      "epoch": 0.11935495666800745,
+      "grad_norm": 0.03907623380015084,
+      "learning_rate": 7.9369019832329e-06,
+      "loss": 0.8901,
+      "step": 204
+    },
+    {
+      "epoch": 0.1199400299850075,
+      "grad_norm": 0.06841747716076531,
+      "learning_rate": 7.936283694766016e-06,
+      "loss": 0.9382,
+      "step": 205
+    },
+    {
+      "epoch": 0.12052510330200754,
+      "grad_norm": 0.03864267029597622,
+      "learning_rate": 7.935662418835023e-06,
+      "loss": 0.9445,
+      "step": 206
+    },
+    {
+      "epoch": 0.12111017661900757,
+      "grad_norm": 0.04187477888552647,
+      "learning_rate": 7.935038155964775e-06,
+      "loss": 0.949,
+      "step": 207
+    },
+    {
+      "epoch": 0.12169524993600761,
+      "grad_norm": 0.03844046778107278,
+      "learning_rate": 7.934410906682653e-06,
+      "loss": 0.8185,
+      "step": 208
+    },
+    {
+      "epoch": 0.12228032325300764,
+      "grad_norm": 0.03894778215959397,
+      "learning_rate": 7.933780671518558e-06,
+      "loss": 0.8226,
+      "step": 209
+    },
+    {
+      "epoch": 0.12286539657000768,
+      "grad_norm": 0.05316470088962357,
+      "learning_rate": 7.933147451004914e-06,
+      "loss": 0.9149,
+      "step": 210
+    },
+    {
+      "epoch": 0.12345046988700771,
+      "grad_norm": 0.04054196015487159,
+      "learning_rate": 7.932511245676669e-06,
+      "loss": 0.907,
+      "step": 211
+    },
+    {
+      "epoch": 0.12403554320400775,
+      "grad_norm": 0.04197248747401694,
+      "learning_rate": 7.931872056071292e-06,
+      "loss": 0.8974,
+      "step": 212
+    },
+    {
+      "epoch": 0.12462061652100778,
+      "grad_norm": 0.04107563523275673,
+      "learning_rate": 7.931229882728771e-06,
+      "loss": 0.8758,
+      "step": 213
+    },
+    {
+      "epoch": 0.12520568983800784,
+      "grad_norm": 0.04898006391812857,
+      "learning_rate": 7.930584726191616e-06,
+      "loss": 0.9015,
+      "step": 214
+    },
+    {
+      "epoch": 0.12579076315500787,
+      "grad_norm": 0.053828637223887026,
+      "learning_rate": 7.92993658700486e-06,
+      "loss": 0.9095,
+      "step": 215
+    },
+    {
+      "epoch": 0.1263758364720079,
+      "grad_norm": 0.041596264391364365,
+      "learning_rate": 7.929285465716051e-06,
+      "loss": 0.9324,
+      "step": 216
+    },
+    {
+      "epoch": 0.12696090978900793,
+      "grad_norm": 0.04270202089537964,
+      "learning_rate": 7.928631362875258e-06,
+      "loss": 0.9712,
+      "step": 217
+    },
+    {
+      "epoch": 0.12754598310600798,
+      "grad_norm": 0.04508600982672278,
+      "learning_rate": 7.927974279035069e-06,
+      "loss": 0.8526,
+      "step": 218
+    },
+    {
+      "epoch": 0.128131056423008,
+      "grad_norm": 0.04210505656160864,
+      "learning_rate": 7.927314214750592e-06,
+      "loss": 0.8483,
+      "step": 219
+    },
+    {
+      "epoch": 0.12871612974000804,
+      "grad_norm": 0.04228456764136032,
+      "learning_rate": 7.926651170579451e-06,
+      "loss": 1.0049,
+      "step": 220
+    },
+    {
+      "epoch": 0.12930120305700807,
+      "grad_norm": 0.039866451292078504,
+      "learning_rate": 7.92598514708179e-06,
+      "loss": 0.9421,
+      "step": 221
+    },
+    {
+      "epoch": 0.12988627637400812,
+      "grad_norm": 0.037613132725661406,
+      "learning_rate": 7.925316144820263e-06,
+      "loss": 0.9474,
+      "step": 222
+    },
+    {
+      "epoch": 0.13047134969100815,
+      "grad_norm": 0.04168086524600805,
+      "learning_rate": 7.92464416436005e-06,
+      "loss": 0.9058,
+      "step": 223
+    },
+    {
+      "epoch": 0.13105642300800818,
+      "grad_norm": 0.039621964711338775,
+      "learning_rate": 7.923969206268839e-06,
+      "loss": 0.9086,
+      "step": 224
+    },
+    {
+      "epoch": 0.13164149632500824,
+      "grad_norm": 0.05209282643387043,
+      "learning_rate": 7.923291271116838e-06,
+      "loss": 0.9298,
+      "step": 225
+    },
+    {
+      "epoch": 0.13222656964200827,
+      "grad_norm": 0.044258185244179175,
+      "learning_rate": 7.92261035947677e-06,
+      "loss": 0.8925,
+      "step": 226
+    },
+    {
+      "epoch": 0.1328116429590083,
+      "grad_norm": 0.039844108391859055,
+      "learning_rate": 7.92192647192387e-06,
+      "loss": 0.8392,
+      "step": 227
+    },
+    {
+      "epoch": 0.13339671627600833,
+      "grad_norm": 0.044744651789733476,
+      "learning_rate": 7.92123960903589e-06,
+      "loss": 0.9329,
+      "step": 228
+    },
+    {
+      "epoch": 0.13398178959300838,
+      "grad_norm": 0.04494192821446448,
+      "learning_rate": 7.92054977139309e-06,
+      "loss": 0.8606,
+      "step": 229
+    },
+    {
+      "epoch": 0.1345668629100084,
+      "grad_norm": 0.05863838322698434,
+      "learning_rate": 7.919856959578252e-06,
+      "loss": 0.9302,
+      "step": 230
+    },
+    {
+      "epoch": 0.13515193622700844,
+      "grad_norm": 0.04085984556832983,
+      "learning_rate": 7.919161174176663e-06,
+      "loss": 0.9993,
+      "step": 231
+    },
+    {
+      "epoch": 0.1357370095440085,
+      "grad_norm": 0.04500674877406008,
+      "learning_rate": 7.918462415776125e-06,
+      "loss": 0.9377,
+      "step": 232
+    },
+    {
+      "epoch": 0.13632208286100853,
+      "grad_norm": 0.039173528958006375,
+      "learning_rate": 7.917760684966955e-06,
+      "loss": 0.813,
+      "step": 233
+    },
+    {
+      "epoch": 0.13690715617800855,
+      "grad_norm": 0.04298295272921228,
+      "learning_rate": 7.91705598234197e-06,
+      "loss": 0.8952,
+      "step": 234
+    },
+    {
+      "epoch": 0.13749222949500858,
+      "grad_norm": 0.04028408438122686,
+      "learning_rate": 7.916348308496513e-06,
+      "loss": 0.9051,
+      "step": 235
+    },
+    {
+      "epoch": 0.13807730281200864,
+      "grad_norm": 0.06743695497866435,
+      "learning_rate": 7.915637664028423e-06,
+      "loss": 0.9475,
+      "step": 236
+    },
+    {
+      "epoch": 0.13866237612900867,
+      "grad_norm": 0.03949575625475006,
+      "learning_rate": 7.914924049538061e-06,
+      "loss": 0.9316,
+      "step": 237
+    },
+    {
+      "epoch": 0.1392474494460087,
+      "grad_norm": 0.04030623921903529,
+      "learning_rate": 7.914207465628284e-06,
+      "loss": 0.885,
+      "step": 238
+    },
+    {
+      "epoch": 0.13983252276300873,
+      "grad_norm": 0.047528726409039795,
+      "learning_rate": 7.91348791290447e-06,
+      "loss": 0.918,
+      "step": 239
+    },
+    {
+      "epoch": 0.14041759608000878,
+      "grad_norm": 0.04466991571391728,
+      "learning_rate": 7.912765391974496e-06,
+      "loss": 0.9305,
+      "step": 240
+    },
+    {
+      "epoch": 0.1410026693970088,
+      "grad_norm": 0.04186868219039162,
+      "learning_rate": 7.912039903448752e-06,
+      "loss": 0.9415,
+      "step": 241
+    },
+    {
+      "epoch": 0.14158774271400884,
+      "grad_norm": 0.05801571080351748,
+      "learning_rate": 7.91131144794013e-06,
+      "loss": 0.8787,
+      "step": 242
+    },
+    {
+      "epoch": 0.1421728160310089,
+      "grad_norm": 0.045088484806881386,
+      "learning_rate": 7.910580026064038e-06,
+      "loss": 1.0604,
+      "step": 243
+    },
+    {
+      "epoch": 0.14275788934800893,
+      "grad_norm": 0.06574863742707004,
+      "learning_rate": 7.909845638438377e-06,
+      "loss": 0.9216,
+      "step": 244
+    },
+    {
+      "epoch": 0.14334296266500896,
+      "grad_norm": 0.0372095564143721,
+      "learning_rate": 7.909108285683563e-06,
+      "loss": 0.9336,
+      "step": 245
+    },
+    {
+      "epoch": 0.14392803598200898,
+      "grad_norm": 0.04397627871472844,
+      "learning_rate": 7.908367968422515e-06,
+      "loss": 0.9261,
+      "step": 246
+    },
+    {
+      "epoch": 0.14451310929900904,
+      "grad_norm": 0.04276892267104892,
+      "learning_rate": 7.907624687280654e-06,
+      "loss": 0.8673,
+      "step": 247
+    },
+    {
+      "epoch": 0.14509818261600907,
+      "grad_norm": 0.045304865664725656,
+      "learning_rate": 7.906878442885907e-06,
+      "loss": 0.9475,
+      "step": 248
+    },
+    {
+      "epoch": 0.1456832559330091,
+      "grad_norm": 0.04370076931511423,
+      "learning_rate": 7.906129235868702e-06,
+      "loss": 0.8433,
+      "step": 249
+    },
+    {
+      "epoch": 0.14626832925000915,
+      "grad_norm": 0.0672588807453601,
+      "learning_rate": 7.905377066861973e-06,
+      "loss": 0.9124,
+      "step": 250
+    },
+    {
+      "epoch": 0.14685340256700918,
+      "grad_norm": 0.046912566764935076,
+      "learning_rate": 7.904621936501156e-06,
+      "loss": 0.9044,
+      "step": 251
+    },
+    {
+      "epoch": 0.1474384758840092,
+      "grad_norm": 0.07278415346178851,
+      "learning_rate": 7.903863845424185e-06,
+      "loss": 0.901,
+      "step": 252
+    },
+    {
+      "epoch": 0.14802354920100924,
+      "grad_norm": 0.04218918671820467,
+      "learning_rate": 7.9031027942715e-06,
+      "loss": 0.9019,
+      "step": 253
+    },
+    {
+      "epoch": 0.1486086225180093,
+      "grad_norm": 0.03858325311022567,
+      "learning_rate": 7.90233878368604e-06,
+      "loss": 0.8601,
+      "step": 254
+    },
+    {
+      "epoch": 0.14919369583500933,
+      "grad_norm": 0.04209886442727145,
+      "learning_rate": 7.90157181431324e-06,
+      "loss": 0.8681,
+      "step": 255
+    },
+    {
+      "epoch": 0.14977876915200936,
+      "grad_norm": 0.049400812285918,
+      "learning_rate": 7.90080188680104e-06,
+      "loss": 0.9494,
+      "step": 256
+    },
+    {
+      "epoch": 0.15036384246900938,
+      "grad_norm": 0.047941941474994906,
+      "learning_rate": 7.900029001799882e-06,
+      "loss": 0.8439,
+      "step": 257
+    },
+    {
+      "epoch": 0.15094891578600944,
+      "grad_norm": 0.042101153780900284,
+      "learning_rate": 7.899253159962694e-06,
+      "loss": 0.917,
+      "step": 258
+    },
+    {
+      "epoch": 0.15153398910300947,
+      "grad_norm": 0.040583990176968884,
+      "learning_rate": 7.898474361944915e-06,
+      "loss": 0.9328,
+      "step": 259
+    },
+    {
+      "epoch": 0.1521190624200095,
+      "grad_norm": 0.04000469028913075,
+      "learning_rate": 7.897692608404474e-06,
+      "loss": 0.9454,
+      "step": 260
+    },
+    {
+      "epoch": 0.15270413573700956,
+      "grad_norm": 0.04819787772789072,
+      "learning_rate": 7.8969079000018e-06,
+      "loss": 1.0061,
+      "step": 261
+    },
+    {
+      "epoch": 0.15328920905400958,
+      "grad_norm": 0.04670811295251141,
+      "learning_rate": 7.896120237399817e-06,
+      "loss": 0.8678,
+      "step": 262
+    },
+    {
+      "epoch": 0.1538742823710096,
+      "grad_norm": 0.03876277230266676,
+      "learning_rate": 7.895329621263945e-06,
+      "loss": 0.9465,
+      "step": 263
+    },
+    {
+      "epoch": 0.15445935568800964,
+      "grad_norm": 0.06917560168966953,
+      "learning_rate": 7.894536052262098e-06,
+      "loss": 0.9114,
+      "step": 264
+    },
+    {
+      "epoch": 0.1550444290050097,
+      "grad_norm": 0.052197549375990714,
+      "learning_rate": 7.893739531064688e-06,
+      "loss": 0.8966,
+      "step": 265
+    },
+    {
+      "epoch": 0.15562950232200973,
+      "grad_norm": 0.05089520469658166,
+      "learning_rate": 7.892940058344615e-06,
+      "loss": 0.8403,
+      "step": 266
+    },
+    {
+      "epoch": 0.15621457563900976,
+      "grad_norm": 0.04200303069403226,
+      "learning_rate": 7.89213763477728e-06,
+      "loss": 0.7954,
+      "step": 267
+    },
+    {
+      "epoch": 0.1567996489560098,
+      "grad_norm": 0.04045997821257577,
+      "learning_rate": 7.89133226104057e-06,
+      "loss": 0.9484,
+      "step": 268
+    },
+    {
+      "epoch": 0.15738472227300984,
+      "grad_norm": 0.04355833968928891,
+      "learning_rate": 7.890523937814872e-06,
+      "loss": 0.8871,
+      "step": 269
+    },
+    {
+      "epoch": 0.15796979559000987,
+      "grad_norm": 0.038980722774793516,
+      "learning_rate": 7.889712665783055e-06,
+      "loss": 0.8242,
+      "step": 270
+    },
+    {
+      "epoch": 0.1585548689070099,
+      "grad_norm": 0.0417431555190703,
+      "learning_rate": 7.888898445630486e-06,
+      "loss": 0.918,
+      "step": 271
+    },
+    {
+      "epoch": 0.15913994222400996,
+      "grad_norm": 0.04420422353440596,
+      "learning_rate": 7.888081278045022e-06,
+      "loss": 0.9345,
+      "step": 272
+    },
+    {
+      "epoch": 0.15972501554100998,
+      "grad_norm": 0.043453256463112454,
+      "learning_rate": 7.88726116371701e-06,
+      "loss": 0.8439,
+      "step": 273
+    },
+    {
+      "epoch": 0.16031008885801,
+      "grad_norm": 0.04018335492133053,
+      "learning_rate": 7.88643810333928e-06,
+      "loss": 0.9024,
+      "step": 274
+    },
+    {
+      "epoch": 0.16089516217501004,
+      "grad_norm": 0.03885288803364117,
+      "learning_rate": 7.885612097607161e-06,
+      "loss": 0.9005,
+      "step": 275
+    },
+    {
+      "epoch": 0.1614802354920101,
+      "grad_norm": 0.03940189643105726,
+      "learning_rate": 7.884783147218464e-06,
+      "loss": 0.8726,
+      "step": 276
+    },
+    {
+      "epoch": 0.16206530880901013,
+      "grad_norm": 0.04322545289358093,
+      "learning_rate": 7.88395125287349e-06,
+      "loss": 0.9309,
+      "step": 277
+    },
+    {
+      "epoch": 0.16265038212601016,
+      "grad_norm": 0.03991615308835092,
+      "learning_rate": 7.883116415275022e-06,
+      "loss": 0.9319,
+      "step": 278
+    },
+    {
+      "epoch": 0.1632354554430102,
+      "grad_norm": 0.04774591346437184,
+      "learning_rate": 7.882278635128339e-06,
+      "loss": 0.9976,
+      "step": 279
+    },
+    {
+      "epoch": 0.16382052876001024,
+      "grad_norm": 0.043726280435400645,
+      "learning_rate": 7.881437913141196e-06,
+      "loss": 0.9041,
+      "step": 280
+    },
+    {
+      "epoch": 0.16440560207701027,
+      "grad_norm": 0.05349577122745657,
+      "learning_rate": 7.880594250023842e-06,
+      "loss": 0.9109,
+      "step": 281
+    },
+    {
+      "epoch": 0.1649906753940103,
+      "grad_norm": 0.04664425873865413,
+      "learning_rate": 7.879747646489002e-06,
+      "loss": 0.8872,
+      "step": 282
+    },
+    {
+      "epoch": 0.16557574871101036,
+      "grad_norm": 0.04689991827127376,
+      "learning_rate": 7.878898103251891e-06,
+      "loss": 0.946,
+      "step": 283
+    },
+    {
+      "epoch": 0.16616082202801039,
+      "grad_norm": 0.04419788368438809,
+      "learning_rate": 7.87804562103021e-06,
+      "loss": 0.8699,
+      "step": 284
+    },
+    {
+      "epoch": 0.16674589534501041,
+      "grad_norm": 0.0498909584325992,
+      "learning_rate": 7.877190200544131e-06,
+      "loss": 0.8396,
+      "step": 285
+    },
+    {
+      "epoch": 0.16733096866201047,
+      "grad_norm": 0.04446094526551824,
+      "learning_rate": 7.876331842516323e-06,
+      "loss": 0.887,
+      "step": 286
+    },
+    {
+      "epoch": 0.1679160419790105,
+      "grad_norm": 0.06094680175561847,
+      "learning_rate": 7.875470547671926e-06,
+      "loss": 0.8834,
+      "step": 287
+    },
+    {
+      "epoch": 0.16850111529601053,
+      "grad_norm": 0.038876474999689326,
+      "learning_rate": 7.874606316738566e-06,
+      "loss": 0.8975,
+      "step": 288
+    },
+    {
+      "epoch": 0.16908618861301056,
+      "grad_norm": 0.04076135396799628,
+      "learning_rate": 7.873739150446349e-06,
+      "loss": 0.9094,
+      "step": 289
+    },
+    {
+      "epoch": 0.16967126193001061,
+      "grad_norm": 0.04242085203333459,
+      "learning_rate": 7.872869049527855e-06,
+      "loss": 0.9346,
+      "step": 290
+    },
+    {
+      "epoch": 0.17025633524701064,
+      "grad_norm": 0.04192270928126719,
+      "learning_rate": 7.871996014718154e-06,
+      "loss": 0.916,
+      "step": 291
+    },
+    {
+      "epoch": 0.17084140856401067,
+      "grad_norm": 0.06296131776401025,
+      "learning_rate": 7.871120046754787e-06,
+      "loss": 0.7869,
+      "step": 292
+    },
+    {
+      "epoch": 0.1714264818810107,
+      "grad_norm": 0.04208658542590707,
+      "learning_rate": 7.870241146377773e-06,
+      "loss": 0.863,
+      "step": 293
+    },
+    {
+      "epoch": 0.17201155519801076,
+      "grad_norm": 0.04221040890826715,
+      "learning_rate": 7.869359314329613e-06,
+      "loss": 0.8125,
+      "step": 294
+    },
+    {
+      "epoch": 0.17259662851501079,
+      "grad_norm": 0.044062682914531,
+      "learning_rate": 7.868474551355277e-06,
+      "loss": 0.8283,
+      "step": 295
+    },
+    {
+      "epoch": 0.17318170183201081,
+      "grad_norm": 0.053136044791769796,
+      "learning_rate": 7.867586858202221e-06,
+      "loss": 0.9321,
+      "step": 296
+    },
+    {
+      "epoch": 0.17376677514901087,
+      "grad_norm": 0.041998835388755755,
+      "learning_rate": 7.866696235620367e-06,
+      "loss": 0.9435,
+      "step": 297
+    },
+    {
+      "epoch": 0.1743518484660109,
+      "grad_norm": 0.04432061620173052,
+      "learning_rate": 7.865802684362119e-06,
+      "loss": 0.944,
+      "step": 298
+    },
+    {
+      "epoch": 0.17493692178301093,
+      "grad_norm": 0.03816132379303917,
+      "learning_rate": 7.864906205182347e-06,
+      "loss": 0.9222,
+      "step": 299
+    },
+    {
+      "epoch": 0.17552199510001096,
+      "grad_norm": 0.04061878988742196,
+      "learning_rate": 7.864006798838405e-06,
+      "loss": 0.9344,
+      "step": 300
+    },
+    {
+      "epoch": 0.17610706841701101,
+      "grad_norm": 0.038725999488975066,
+      "learning_rate": 7.863104466090113e-06,
+      "loss": 0.9477,
+      "step": 301
+    },
+    {
+      "epoch": 0.17669214173401104,
+      "grad_norm": 0.0675670307551766,
+      "learning_rate": 7.862199207699763e-06,
+      "loss": 0.8939,
+      "step": 302
+    },
+    {
+      "epoch": 0.17727721505101107,
+      "grad_norm": 0.04866290151842816,
+      "learning_rate": 7.861291024432122e-06,
+      "loss": 0.944,
+      "step": 303
+    },
+    {
+      "epoch": 0.17786228836801113,
+      "grad_norm": 0.04231401792052211,
+      "learning_rate": 7.860379917054426e-06,
+      "loss": 0.9108,
+      "step": 304
+    },
+    {
+      "epoch": 0.17844736168501116,
+      "grad_norm": 0.21645206187990054,
+      "learning_rate": 7.859465886336381e-06,
+      "loss": 0.9328,
+      "step": 305
+    },
+    {
+      "epoch": 0.1790324350020112,
+      "grad_norm": 0.04212212891416765,
+      "learning_rate": 7.858548933050162e-06,
+      "loss": 0.9755,
+      "step": 306
+    },
+    {
+      "epoch": 0.17961750831901122,
+      "grad_norm": 0.047331820432207364,
+      "learning_rate": 7.857629057970417e-06,
+      "loss": 0.8702,
+      "step": 307
+    },
+    {
+      "epoch": 0.18020258163601127,
+      "grad_norm": 0.040260553510288316,
+      "learning_rate": 7.856706261874258e-06,
+      "loss": 0.8934,
+      "step": 308
+    },
+    {
+      "epoch": 0.1807876549530113,
+      "grad_norm": 0.04326897093604986,
+      "learning_rate": 7.855780545541264e-06,
+      "loss": 0.877,
+      "step": 309
+    },
+    {
+      "epoch": 0.18137272827001133,
+      "grad_norm": 0.03936456606263684,
+      "learning_rate": 7.854851909753487e-06,
+      "loss": 0.9206,
+      "step": 310
+    },
+    {
+      "epoch": 0.18195780158701136,
+      "grad_norm": 0.040529920237504666,
+      "learning_rate": 7.853920355295438e-06,
+      "loss": 0.8469,
+      "step": 311
+    },
+    {
+      "epoch": 0.18254287490401142,
+      "grad_norm": 0.07984199172508148,
+      "learning_rate": 7.852985882954102e-06,
+      "loss": 0.856,
+      "step": 312
+    },
+    {
+      "epoch": 0.18312794822101144,
+      "grad_norm": 0.04341308255910565,
+      "learning_rate": 7.85204849351892e-06,
+      "loss": 0.8975,
+      "step": 313
+    },
+    {
+      "epoch": 0.18371302153801147,
+      "grad_norm": 0.05163482815007264,
+      "learning_rate": 7.851108187781802e-06,
+      "loss": 0.8516,
+      "step": 314
+    },
+    {
+      "epoch": 0.18429809485501153,
+      "grad_norm": 0.04363599571939765,
+      "learning_rate": 7.850164966537124e-06,
+      "loss": 0.9088,
+      "step": 315
+    },
+    {
+      "epoch": 0.18488316817201156,
+      "grad_norm": 0.043024652012164384,
+      "learning_rate": 7.84921883058172e-06,
+      "loss": 0.8291,
+      "step": 316
+    },
+    {
+      "epoch": 0.1854682414890116,
+      "grad_norm": 0.041659608271311335,
+      "learning_rate": 7.848269780714892e-06,
+      "loss": 0.9719,
+      "step": 317
+    },
+    {
+      "epoch": 0.18605331480601162,
+      "grad_norm": 0.04557805725835434,
+      "learning_rate": 7.847317817738394e-06,
+      "loss": 0.9638,
+      "step": 318
+    },
+    {
+      "epoch": 0.18663838812301167,
+      "grad_norm": 0.04388413396186285,
+      "learning_rate": 7.846362942456455e-06,
+      "loss": 0.93,
+      "step": 319
+    },
+    {
+      "epoch": 0.1872234614400117,
+      "grad_norm": 0.04052950822615268,
+      "learning_rate": 7.845405155675752e-06,
+      "loss": 0.8951,
+      "step": 320
+    },
+    {
+      "epoch": 0.18780853475701173,
+      "grad_norm": 0.03852369264523642,
+      "learning_rate": 7.844444458205428e-06,
+      "loss": 0.8521,
+      "step": 321
+    },
+    {
+      "epoch": 0.1883936080740118,
+      "grad_norm": 0.045799254500923765,
+      "learning_rate": 7.843480850857083e-06,
+      "loss": 0.8966,
+      "step": 322
+    },
+    {
+      "epoch": 0.18897868139101182,
+      "grad_norm": 0.04373024374645557,
+      "learning_rate": 7.842514334444776e-06,
+      "loss": 0.973,
+      "step": 323
+    },
+    {
+      "epoch": 0.18956375470801184,
+      "grad_norm": 0.037372310076302304,
+      "learning_rate": 7.841544909785022e-06,
+      "loss": 1.0232,
+      "step": 324
+    },
+    {
+      "epoch": 0.19014882802501187,
+      "grad_norm": 0.044911079932608254,
+      "learning_rate": 7.840572577696798e-06,
+      "loss": 0.8351,
+      "step": 325
+    },
+    {
+      "epoch": 0.19073390134201193,
+      "grad_norm": 0.04820487217400144,
+      "learning_rate": 7.839597339001529e-06,
+      "loss": 0.9381,
+      "step": 326
+    },
+    {
+      "epoch": 0.19131897465901196,
+      "grad_norm": 0.04460622903204633,
+      "learning_rate": 7.8386191945231e-06,
+      "loss": 0.9047,
+      "step": 327
+    },
+    {
+      "epoch": 0.191904047976012,
+      "grad_norm": 0.04344227987010105,
+      "learning_rate": 7.837638145087855e-06,
+      "loss": 0.8882,
+      "step": 328
+    },
+    {
+      "epoch": 0.19248912129301204,
+      "grad_norm": 0.041162516007716705,
+      "learning_rate": 7.836654191524583e-06,
+      "loss": 0.888,
+      "step": 329
+    },
+    {
+      "epoch": 0.19307419461001207,
+      "grad_norm": 0.050217278727892056,
+      "learning_rate": 7.835667334664533e-06,
+      "loss": 0.9425,
+      "step": 330
+    },
+    {
+      "epoch": 0.1936592679270121,
+      "grad_norm": 0.045239661384039016,
+      "learning_rate": 7.834677575341407e-06,
+      "loss": 0.845,
+      "step": 331
+    },
+    {
+      "epoch": 0.19424434124401213,
+      "grad_norm": 0.04069762643444013,
+      "learning_rate": 7.833684914391354e-06,
+      "loss": 0.9045,
+      "step": 332
+    },
+    {
+      "epoch": 0.1948294145610122,
+      "grad_norm": 0.03659391253836006,
+      "learning_rate": 7.832689352652978e-06,
+      "loss": 0.8415,
+      "step": 333
+    },
+    {
+      "epoch": 0.19541448787801222,
+      "grad_norm": 0.0478253523475305,
+      "learning_rate": 7.831690890967332e-06,
+      "loss": 0.9023,
+      "step": 334
+    },
+    {
+      "epoch": 0.19599956119501225,
+      "grad_norm": 0.03931532297283958,
+      "learning_rate": 7.830689530177923e-06,
+      "loss": 0.8757,
+      "step": 335
+    },
+    {
+      "epoch": 0.19658463451201227,
+      "grad_norm": 0.04480666786944768,
+      "learning_rate": 7.8296852711307e-06,
+      "loss": 0.8393,
+      "step": 336
+    },
+    {
+      "epoch": 0.19716970782901233,
+      "grad_norm": 0.0734058976406723,
+      "learning_rate": 7.828678114674066e-06,
+      "loss": 0.9038,
+      "step": 337
+    },
+    {
+      "epoch": 0.19775478114601236,
+      "grad_norm": 0.03967258990394233,
+      "learning_rate": 7.827668061658871e-06,
+      "loss": 0.8009,
+      "step": 338
+    },
+    {
+      "epoch": 0.1983398544630124,
+      "grad_norm": 0.0418553799297778,
+      "learning_rate": 7.82665511293841e-06,
+      "loss": 0.8865,
+      "step": 339
+    },
+    {
+      "epoch": 0.19892492778001244,
+      "grad_norm": 0.0384561013270465,
+      "learning_rate": 7.825639269368426e-06,
+      "loss": 0.872,
+      "step": 340
+    },
+    {
+      "epoch": 0.19951000109701247,
+      "grad_norm": 0.04182210062250734,
+      "learning_rate": 7.824620531807106e-06,
+      "loss": 0.8974,
+      "step": 341
+    },
+    {
+      "epoch": 0.2000950744140125,
+      "grad_norm": 0.0402445680829306,
+      "learning_rate": 7.823598901115085e-06,
+      "loss": 0.8017,
+      "step": 342
+    },
+    {
+      "epoch": 0.20068014773101253,
+      "grad_norm": 0.03976795416941979,
+      "learning_rate": 7.822574378155436e-06,
+      "loss": 0.9298,
+      "step": 343
+    },
+    {
+      "epoch": 0.2012652210480126,
+      "grad_norm": 0.043031163568613286,
+      "learning_rate": 7.821546963793683e-06,
+      "loss": 0.9508,
+      "step": 344
+    },
+    {
+      "epoch": 0.20185029436501262,
+      "grad_norm": 0.0485982687560341,
+      "learning_rate": 7.82051665889779e-06,
+      "loss": 0.8536,
+      "step": 345
+    },
+    {
+      "epoch": 0.20243536768201265,
+      "grad_norm": 0.04149899672912665,
+      "learning_rate": 7.819483464338156e-06,
+      "loss": 0.8767,
+      "step": 346
+    },
+    {
+      "epoch": 0.2030204409990127,
+      "grad_norm": 0.0971131140419449,
+      "learning_rate": 7.818447380987634e-06,
+      "loss": 0.9271,
+      "step": 347
+    },
+    {
+      "epoch": 0.20360551431601273,
+      "grad_norm": 0.04183099462603424,
+      "learning_rate": 7.817408409721506e-06,
+      "loss": 0.8362,
+      "step": 348
+    },
+    {
+      "epoch": 0.20419058763301276,
+      "grad_norm": 0.03719681766847752,
+      "learning_rate": 7.8163665514175e-06,
+      "loss": 0.8544,
+      "step": 349
+    },
+    {
+      "epoch": 0.2047756609500128,
+      "grad_norm": 0.08061763199476867,
+      "learning_rate": 7.815321806955782e-06,
+      "loss": 0.8335,
+      "step": 350
+    },
+    {
+      "epoch": 0.20536073426701285,
+      "grad_norm": 0.053441055551360056,
+      "learning_rate": 7.814274177218955e-06,
+      "loss": 0.8602,
+      "step": 351
+    },
+    {
+      "epoch": 0.20594580758401287,
+      "grad_norm": 0.1775803458323184,
+      "learning_rate": 7.81322366309206e-06,
+      "loss": 0.8998,
+      "step": 352
+    },
+    {
+      "epoch": 0.2065308809010129,
+      "grad_norm": 0.0637002971995377,
+      "learning_rate": 7.812170265462573e-06,
+      "loss": 0.9737,
+      "step": 353
+    },
+    {
+      "epoch": 0.20711595421801293,
+      "grad_norm": 0.04274639590829472,
+      "learning_rate": 7.81111398522041e-06,
+      "loss": 1.009,
+      "step": 354
+    },
+    {
+      "epoch": 0.207701027535013,
+      "grad_norm": 0.041282961278004064,
+      "learning_rate": 7.81005482325792e-06,
+      "loss": 0.9035,
+      "step": 355
+    },
+    {
+      "epoch": 0.20828610085201302,
+      "grad_norm": 0.04195399372218902,
+      "learning_rate": 7.808992780469889e-06,
+      "loss": 0.9128,
+      "step": 356
+    },
+    {
+      "epoch": 0.20887117416901305,
+      "grad_norm": 0.03698717701423449,
+      "learning_rate": 7.807927857753527e-06,
+      "loss": 0.8562,
+      "step": 357
+    },
+    {
+      "epoch": 0.2094562474860131,
+      "grad_norm": 0.1081591722204418,
+      "learning_rate": 7.80686005600849e-06,
+      "loss": 0.8931,
+      "step": 358
+    },
+    {
+      "epoch": 0.21004132080301313,
+      "grad_norm": 0.04152721891763337,
+      "learning_rate": 7.80578937613686e-06,
+      "loss": 0.8633,
+      "step": 359
+    },
+    {
+      "epoch": 0.21062639412001316,
+      "grad_norm": 0.04200660652374662,
+      "learning_rate": 7.804715819043148e-06,
+      "loss": 0.8773,
+      "step": 360
+    },
+    {
+      "epoch": 0.2112114674370132,
+      "grad_norm": 0.04405111681557889,
+      "learning_rate": 7.803639385634302e-06,
+      "loss": 0.8587,
+      "step": 361
+    },
+    {
+      "epoch": 0.21179654075401325,
+      "grad_norm": 0.04571477197647962,
+      "learning_rate": 7.802560076819694e-06,
+      "loss": 0.8334,
+      "step": 362
+    },
+    {
+      "epoch": 0.21238161407101327,
+      "grad_norm": 0.041607118224879065,
+      "learning_rate": 7.80147789351113e-06,
+      "loss": 0.8739,
+      "step": 363
+    },
+    {
+      "epoch": 0.2129666873880133,
+      "grad_norm": 0.06194034993642153,
+      "learning_rate": 7.800392836622838e-06,
+      "loss": 0.8956,
+      "step": 364
+    },
+    {
+      "epoch": 0.21355176070501336,
+      "grad_norm": 0.03798806008002444,
+      "learning_rate": 7.79930490707148e-06,
+      "loss": 0.8966,
+      "step": 365
+    },
+    {
+      "epoch": 0.2141368340220134,
+      "grad_norm": 0.04739784630899101,
+      "learning_rate": 7.798214105776146e-06,
+      "loss": 0.9552,
+      "step": 366
+    },
+    {
+      "epoch": 0.21472190733901342,
+      "grad_norm": 0.04497114482691456,
+      "learning_rate": 7.797120433658343e-06,
+      "loss": 0.8666,
+      "step": 367
+    },
+    {
+      "epoch": 0.21530698065601345,
+      "grad_norm": 0.04176901867071411,
+      "learning_rate": 7.796023891642011e-06,
+      "loss": 0.9051,
+      "step": 368
+    },
+    {
+      "epoch": 0.2158920539730135,
+      "grad_norm": 0.056698801623745465,
+      "learning_rate": 7.794924480653513e-06,
+      "loss": 0.8745,
+      "step": 369
+    },
+    {
+      "epoch": 0.21647712729001353,
+      "grad_norm": 0.04267067239376988,
+      "learning_rate": 7.793822201621633e-06,
+      "loss": 0.9129,
+      "step": 370
+    },
+    {
+      "epoch": 0.21706220060701356,
+      "grad_norm": 0.03999203881276761,
+      "learning_rate": 7.79271705547758e-06,
+      "loss": 0.8814,
+      "step": 371
+    },
+    {
+      "epoch": 0.2176472739240136,
+      "grad_norm": 0.04206496870855173,
+      "learning_rate": 7.79160904315499e-06,
+      "loss": 0.8936,
+      "step": 372
+    },
+    {
+      "epoch": 0.21823234724101365,
+      "grad_norm": 0.040567992793796616,
+      "learning_rate": 7.79049816558991e-06,
+      "loss": 0.8961,
+      "step": 373
+    },
+    {
+      "epoch": 0.21881742055801368,
+      "grad_norm": 0.04165915715538525,
+      "learning_rate": 7.789384423720815e-06,
+      "loss": 0.901,
+      "step": 374
+    },
+    {
+      "epoch": 0.2194024938750137,
+      "grad_norm": 0.04311161567240108,
+      "learning_rate": 7.788267818488597e-06,
+      "loss": 0.8571,
+      "step": 375
+    },
+    {
+      "epoch": 0.21998756719201376,
+      "grad_norm": 0.04090984120457054,
+      "learning_rate": 7.78714835083657e-06,
+      "loss": 0.879,
+      "step": 376
+    },
+    {
+      "epoch": 0.2205726405090138,
+      "grad_norm": 0.03723653818234615,
+      "learning_rate": 7.786026021710462e-06,
+      "loss": 0.8687,
+      "step": 377
+    },
+    {
+      "epoch": 0.22115771382601382,
+      "grad_norm": 0.040732056413017595,
+      "learning_rate": 7.78490083205842e-06,
+      "loss": 0.9033,
+      "step": 378
+    },
+    {
+      "epoch": 0.22174278714301385,
+      "grad_norm": 0.03755340315603773,
+      "learning_rate": 7.783772782831008e-06,
+      "loss": 0.8919,
+      "step": 379
+    },
+    {
+      "epoch": 0.2223278604600139,
+      "grad_norm": 0.04513035465018611,
+      "learning_rate": 7.782641874981207e-06,
+      "loss": 0.8766,
+      "step": 380
+    },
+    {
+      "epoch": 0.22291293377701393,
+      "grad_norm": 0.042927929340526826,
+      "learning_rate": 7.78150810946441e-06,
+      "loss": 0.8692,
+      "step": 381
+    },
+    {
+      "epoch": 0.22349800709401396,
+      "grad_norm": 0.03548281821425231,
+      "learning_rate": 7.780371487238428e-06,
+      "loss": 0.7295,
+      "step": 382
+    },
+    {
+      "epoch": 0.22408308041101402,
+      "grad_norm": 0.03978070402906236,
+      "learning_rate": 7.779232009263484e-06,
+      "loss": 0.8555,
+      "step": 383
+    },
+    {
+      "epoch": 0.22466815372801405,
+      "grad_norm": 0.036709565386030436,
+      "learning_rate": 7.778089676502209e-06,
+      "loss": 0.7492,
+      "step": 384
+    },
+    {
+      "epoch": 0.22525322704501408,
+      "grad_norm": 0.04547788495521431,
+      "learning_rate": 7.776944489919649e-06,
+      "loss": 0.8334,
+      "step": 385
+    },
+    {
+      "epoch": 0.2258383003620141,
+      "grad_norm": 0.04220442842369723,
+      "learning_rate": 7.775796450483267e-06,
+      "loss": 0.8244,
+      "step": 386
+    },
+    {
+      "epoch": 0.22642337367901416,
+      "grad_norm": 0.04778060193840243,
+      "learning_rate": 7.774645559162927e-06,
+      "loss": 0.8511,
+      "step": 387
+    },
+    {
+      "epoch": 0.2270084469960142,
+      "grad_norm": 0.05014682109980806,
+      "learning_rate": 7.773491816930904e-06,
+      "loss": 0.8334,
+      "step": 388
+    },
+    {
+      "epoch": 0.22759352031301422,
+      "grad_norm": 0.04001986651909544,
+      "learning_rate": 7.772335224761886e-06,
+      "loss": 0.8224,
+      "step": 389
+    },
+    {
+      "epoch": 0.22817859363001425,
+      "grad_norm": 0.040094874463681825,
+      "learning_rate": 7.771175783632966e-06,
+      "loss": 0.9069,
+      "step": 390
+    },
+    {
+      "epoch": 0.2287636669470143,
+      "grad_norm": 0.03678306586668835,
+      "learning_rate": 7.770013494523641e-06,
+      "loss": 0.8758,
+      "step": 391
+    },
+    {
+      "epoch": 0.22934874026401433,
+      "grad_norm": 0.04444193067206238,
+      "learning_rate": 7.768848358415819e-06,
+      "loss": 0.946,
+      "step": 392
+    },
+    {
+      "epoch": 0.22993381358101436,
+      "grad_norm": 0.042386218395158284,
+      "learning_rate": 7.767680376293811e-06,
+      "loss": 0.8395,
+      "step": 393
+    },
+    {
+      "epoch": 0.23051888689801442,
+      "grad_norm": 0.051874603825963005,
+      "learning_rate": 7.766509549144332e-06,
+      "loss": 0.8867,
+      "step": 394
+    },
+    {
+      "epoch": 0.23110396021501445,
+      "grad_norm": 0.04115935591406729,
+      "learning_rate": 7.765335877956498e-06,
+      "loss": 0.8509,
+      "step": 395
+    },
+    {
+      "epoch": 0.23168903353201448,
+      "grad_norm": 0.045377388296697053,
+      "learning_rate": 7.764159363721833e-06,
+      "loss": 0.8783,
+      "step": 396
+    },
+    {
+      "epoch": 0.2322741068490145,
+      "grad_norm": 0.03845331275197764,
+      "learning_rate": 7.762980007434261e-06,
+      "loss": 0.8721,
+      "step": 397
+    },
+    {
+      "epoch": 0.23285918016601456,
+      "grad_norm": 0.06130607399623932,
+      "learning_rate": 7.761797810090103e-06,
+      "loss": 0.896,
+      "step": 398
+    },
+    {
+      "epoch": 0.2334442534830146,
+      "grad_norm": 0.0653248890033106,
+      "learning_rate": 7.760612772688086e-06,
+      "loss": 0.9239,
+      "step": 399
+    },
+    {
+      "epoch": 0.23402932680001462,
+      "grad_norm": 0.04386903271065406,
+      "learning_rate": 7.759424896229329e-06,
+      "loss": 0.9055,
+      "step": 400
+    },
+    {
+      "epoch": 0.23461440011701468,
+      "grad_norm": 0.05112057938888221,
+      "learning_rate": 7.758234181717359e-06,
+      "loss": 0.8179,
+      "step": 401
+    },
+    {
+      "epoch": 0.2351994734340147,
+      "grad_norm": 0.03919834965990091,
+      "learning_rate": 7.757040630158094e-06,
+      "loss": 0.9131,
+      "step": 402
+    },
+    {
+      "epoch": 0.23578454675101473,
+      "grad_norm": 0.04080472273046829,
+      "learning_rate": 7.75584424255985e-06,
+      "loss": 0.8772,
+      "step": 403
+    },
+    {
+      "epoch": 0.23636962006801476,
+      "grad_norm": 0.036834620421617906,
+      "learning_rate": 7.754645019933338e-06,
+      "loss": 0.8155,
+      "step": 404
+    },
+    {
+      "epoch": 0.23695469338501482,
+      "grad_norm": 0.03654317750178369,
+      "learning_rate": 7.753442963291668e-06,
+      "loss": 0.8346,
+      "step": 405
+    },
+    {
+      "epoch": 0.23753976670201485,
+      "grad_norm": 0.04519798929105044,
+      "learning_rate": 7.752238073650339e-06,
+      "loss": 0.912,
+      "step": 406
+    },
+    {
+      "epoch": 0.23812484001901488,
+      "grad_norm": 0.03826695253269902,
+      "learning_rate": 7.751030352027246e-06,
+      "loss": 0.8772,
+      "step": 407
+    },
+    {
+      "epoch": 0.2387099133360149,
+      "grad_norm": 0.04486993387188878,
+      "learning_rate": 7.749819799442676e-06,
+      "loss": 0.8826,
+      "step": 408
+    },
+    {
+      "epoch": 0.23929498665301496,
+      "grad_norm": 0.0776681654995442,
+      "learning_rate": 7.74860641691931e-06,
+      "loss": 0.8987,
+      "step": 409
+    },
+    {
+      "epoch": 0.239880059970015,
+      "grad_norm": 0.05169098865704706,
+      "learning_rate": 7.747390205482216e-06,
+      "loss": 0.7904,
+      "step": 410
+    },
+    {
+      "epoch": 0.24046513328701502,
+      "grad_norm": 0.042053672534972886,
+      "learning_rate": 7.746171166158855e-06,
+      "loss": 0.9188,
+      "step": 411
+    },
+    {
+      "epoch": 0.24105020660401508,
+      "grad_norm": 0.044614916290635534,
+      "learning_rate": 7.744949299979071e-06,
+      "loss": 0.9118,
+      "step": 412
+    },
+    {
+      "epoch": 0.2416352799210151,
+      "grad_norm": 0.04594381480411999,
+      "learning_rate": 7.743724607975105e-06,
+      "loss": 0.8547,
+      "step": 413
+    },
+    {
+      "epoch": 0.24222035323801513,
+      "grad_norm": 0.037664202042716706,
+      "learning_rate": 7.742497091181578e-06,
+      "loss": 0.8446,
+      "step": 414
+    },
+    {
+      "epoch": 0.24280542655501516,
+      "grad_norm": 0.04794778502796113,
+      "learning_rate": 7.741266750635502e-06,
+      "loss": 0.897,
+      "step": 415
+    },
+    {
+      "epoch": 0.24339049987201522,
+      "grad_norm": 0.04188595026753968,
+      "learning_rate": 7.740033587376275e-06,
+      "loss": 0.9061,
+      "step": 416
+    },
+    {
+      "epoch": 0.24397557318901525,
+      "grad_norm": 0.04407359629305258,
+      "learning_rate": 7.738797602445671e-06,
+      "loss": 0.9146,
+      "step": 417
+    },
+    {
+      "epoch": 0.24456064650601528,
+      "grad_norm": 0.04077311126531583,
+      "learning_rate": 7.73755879688786e-06,
+      "loss": 0.8515,
+      "step": 418
+    },
+    {
+      "epoch": 0.24514571982301533,
+      "grad_norm": 0.055784116703384304,
+      "learning_rate": 7.736317171749385e-06,
+      "loss": 0.851,
+      "step": 419
+    },
+    {
+      "epoch": 0.24573079314001536,
+      "grad_norm": 0.049743142106565615,
+      "learning_rate": 7.735072728079179e-06,
+      "loss": 0.8718,
+      "step": 420
+    },
+    {
+      "epoch": 0.2463158664570154,
+      "grad_norm": 0.04470234941233142,
+      "learning_rate": 7.73382546692855e-06,
+      "loss": 0.9624,
+      "step": 421
+    },
+    {
+      "epoch": 0.24690093977401542,
+      "grad_norm": 0.040010409109375616,
+      "learning_rate": 7.732575389351187e-06,
+      "loss": 0.8925,
+      "step": 422
+    },
+    {
+      "epoch": 0.24748601309101548,
+      "grad_norm": 0.04513853855792064,
+      "learning_rate": 7.731322496403161e-06,
+      "loss": 0.8163,
+      "step": 423
+    },
+    {
+      "epoch": 0.2480710864080155,
+      "grad_norm": 0.04305001532204258,
+      "learning_rate": 7.730066789142922e-06,
+      "loss": 0.818,
+      "step": 424
+    },
+    {
+      "epoch": 0.24865615972501554,
+      "grad_norm": 0.0381328984157233,
+      "learning_rate": 7.728808268631291e-06,
+      "loss": 0.8655,
+      "step": 425
+    },
+    {
+      "epoch": 0.24924123304201556,
+      "grad_norm": 0.04141777979618345,
+      "learning_rate": 7.727546935931473e-06,
+      "loss": 0.8447,
+      "step": 426
+    },
+    {
+      "epoch": 0.24982630635901562,
+      "grad_norm": 0.0441532278076972,
+      "learning_rate": 7.726282792109049e-06,
+      "loss": 0.7839,
+      "step": 427
+    },
+    {
+      "epoch": 0.2504113796760157,
+      "grad_norm": 0.04369415131884431,
+      "learning_rate": 7.725015838231966e-06,
+      "loss": 0.8375,
+      "step": 428
+    },
+    {
+      "epoch": 0.2509964529930157,
+      "grad_norm": 0.04101978650087937,
+      "learning_rate": 7.723746075370553e-06,
+      "loss": 0.8001,
+      "step": 429
+    },
+    {
+      "epoch": 0.25158152631001574,
+      "grad_norm": 0.04974984131766339,
+      "learning_rate": 7.722473504597512e-06,
+      "loss": 0.8914,
+      "step": 430
+    },
+    {
+      "epoch": 0.25216659962701576,
+      "grad_norm": 0.042172544165836386,
+      "learning_rate": 7.721198126987914e-06,
+      "loss": 0.8266,
+      "step": 431
+    },
+    {
+      "epoch": 0.2527516729440158,
+      "grad_norm": 0.04137736709510405,
+      "learning_rate": 7.719919943619202e-06,
+      "loss": 0.9043,
+      "step": 432
+    },
+    {
+      "epoch": 0.2533367462610158,
+      "grad_norm": 0.04296090636863713,
+      "learning_rate": 7.718638955571187e-06,
+      "loss": 0.8638,
+      "step": 433
+    },
+    {
+      "epoch": 0.25392181957801585,
+      "grad_norm": 0.04617407215473608,
+      "learning_rate": 7.717355163926054e-06,
+      "loss": 0.8223,
+      "step": 434
+    },
+    {
+      "epoch": 0.2545068928950159,
+      "grad_norm": 0.05470709039494341,
+      "learning_rate": 7.716068569768357e-06,
+      "loss": 0.8702,
+      "step": 435
+    },
+    {
+      "epoch": 0.25509196621201596,
+      "grad_norm": 0.05481374522257052,
+      "learning_rate": 7.714779174185011e-06,
+      "loss": 0.8644,
+      "step": 436
+    },
+    {
+      "epoch": 0.255677039529016,
+      "grad_norm": 0.047154937968022254,
+      "learning_rate": 7.713486978265303e-06,
+      "loss": 0.8656,
+      "step": 437
+    },
+    {
+      "epoch": 0.256262112846016,
+      "grad_norm": 0.04351286538068476,
+      "learning_rate": 7.712191983100885e-06,
+      "loss": 0.9139,
+      "step": 438
+    },
+    {
+      "epoch": 0.25684718616301605,
+      "grad_norm": 0.036225679985398614,
+      "learning_rate": 7.710894189785773e-06,
+      "loss": 0.8589,
+      "step": 439
+    },
+    {
+      "epoch": 0.2574322594800161,
+      "grad_norm": 0.036619323150318614,
+      "learning_rate": 7.709593599416346e-06,
+      "loss": 0.8355,
+      "step": 440
+    },
+    {
+      "epoch": 0.2580173327970161,
+      "grad_norm": 0.052556029410115106,
+      "learning_rate": 7.708290213091348e-06,
+      "loss": 0.9407,
+      "step": 441
+    },
+    {
+      "epoch": 0.25860240611401614,
+      "grad_norm": 0.04045383405114572,
+      "learning_rate": 7.706984031911884e-06,
+      "loss": 0.8224,
+      "step": 442
+    },
+    {
+      "epoch": 0.2591874794310162,
+      "grad_norm": 0.03850958419485924,
+      "learning_rate": 7.705675056981419e-06,
+      "loss": 0.8289,
+      "step": 443
+    },
+    {
+      "epoch": 0.25977255274801625,
+      "grad_norm": 0.042025364903943374,
+      "learning_rate": 7.704363289405782e-06,
+      "loss": 0.8295,
+      "step": 444
+    },
+    {
+      "epoch": 0.2603576260650163,
+      "grad_norm": 0.18295957147270098,
+      "learning_rate": 7.703048730293156e-06,
+      "loss": 0.8591,
+      "step": 445
+    },
+    {
+      "epoch": 0.2609426993820163,
+      "grad_norm": 0.08559741878328085,
+      "learning_rate": 7.701731380754086e-06,
+      "loss": 0.8841,
+      "step": 446
+    },
+    {
+      "epoch": 0.26152777269901634,
+      "grad_norm": 0.04030996945096956,
+      "learning_rate": 7.700411241901473e-06,
+      "loss": 0.927,
+      "step": 447
+    },
+    {
+      "epoch": 0.26211284601601637,
+      "grad_norm": 0.04080969851872162,
+      "learning_rate": 7.699088314850574e-06,
+      "loss": 0.8448,
+      "step": 448
+    },
+    {
+      "epoch": 0.2626979193330164,
+      "grad_norm": 0.0378951860436084,
+      "learning_rate": 7.697762600719002e-06,
+      "loss": 0.8806,
+      "step": 449
+    },
+    {
+      "epoch": 0.2632829926500165,
+      "grad_norm": 0.03623506747860633,
+      "learning_rate": 7.696434100626727e-06,
+      "loss": 0.8899,
+      "step": 450
+    },
+    {
+      "epoch": 0.2638680659670165,
+      "grad_norm": 0.03949638769998007,
+      "learning_rate": 7.695102815696068e-06,
+      "loss": 0.896,
+      "step": 451
+    },
+    {
+      "epoch": 0.26445313928401654,
+      "grad_norm": 0.039703852605046996,
+      "learning_rate": 7.6937687470517e-06,
+      "loss": 0.9092,
+      "step": 452
+    },
+    {
+      "epoch": 0.26503821260101657,
+      "grad_norm": 0.04067815287205374,
+      "learning_rate": 7.692431895820648e-06,
+      "loss": 0.8948,
+      "step": 453
+    },
+    {
+      "epoch": 0.2656232859180166,
+      "grad_norm": 0.0498733499092084,
+      "learning_rate": 7.691092263132289e-06,
+      "loss": 0.8407,
+      "step": 454
+    },
+    {
+      "epoch": 0.2662083592350166,
+      "grad_norm": 0.06147495248603913,
+      "learning_rate": 7.689749850118347e-06,
+      "loss": 0.8343,
+      "step": 455
+    },
+    {
+      "epoch": 0.26679343255201665,
+      "grad_norm": 0.2184373267208327,
+      "learning_rate": 7.6884046579129e-06,
+      "loss": 0.909,
+      "step": 456
+    },
+    {
+      "epoch": 0.26737850586901674,
+      "grad_norm": 0.038431478494613806,
+      "learning_rate": 7.68705668765237e-06,
+      "loss": 0.8685,
+      "step": 457
+    },
+    {
+      "epoch": 0.26796357918601676,
+      "grad_norm": 0.15936476995911328,
+      "learning_rate": 7.685705940475523e-06,
+      "loss": 0.8176,
+      "step": 458
+    },
+    {
+      "epoch": 0.2685486525030168,
+      "grad_norm": 0.04683404608346142,
+      "learning_rate": 7.68435241752348e-06,
+      "loss": 0.8523,
+      "step": 459
+    },
+    {
+      "epoch": 0.2691337258200168,
+      "grad_norm": 0.04603467291785015,
+      "learning_rate": 7.6829961199397e-06,
+      "loss": 0.8134,
+      "step": 460
+    },
+    {
+      "epoch": 0.26971879913701685,
+      "grad_norm": 0.0502568206105835,
+      "learning_rate": 7.681637048869985e-06,
+      "loss": 0.9043,
+      "step": 461
+    },
+    {
+      "epoch": 0.2703038724540169,
+      "grad_norm": 0.03828249024856938,
+      "learning_rate": 7.680275205462485e-06,
+      "loss": 0.843,
+      "step": 462
+    },
+    {
+      "epoch": 0.2708889457710169,
+      "grad_norm": 0.043272538310479496,
+      "learning_rate": 7.67891059086769e-06,
+      "loss": 0.9456,
+      "step": 463
+    },
+    {
+      "epoch": 0.271474019088017,
+      "grad_norm": 0.04015211847803301,
+      "learning_rate": 7.67754320623843e-06,
+      "loss": 0.9414,
+      "step": 464
+    },
+    {
+      "epoch": 0.272059092405017,
+      "grad_norm": 0.04064423849996686,
+      "learning_rate": 7.676173052729877e-06,
+      "loss": 0.8705,
+      "step": 465
+    },
+    {
+      "epoch": 0.27264416572201705,
+      "grad_norm": 0.04969071039366313,
+      "learning_rate": 7.67480013149954e-06,
+      "loss": 0.8922,
+      "step": 466
+    },
+    {
+      "epoch": 0.2732292390390171,
+      "grad_norm": 0.05260125052428257,
+      "learning_rate": 7.67342444370727e-06,
+      "loss": 0.9733,
+      "step": 467
+    },
+    {
+      "epoch": 0.2738143123560171,
+      "grad_norm": 0.040268950739110694,
+      "learning_rate": 7.672045990515248e-06,
+      "loss": 0.8806,
+      "step": 468
+    },
+    {
+      "epoch": 0.27439938567301714,
+      "grad_norm": 0.042577022398831814,
+      "learning_rate": 7.670664773088e-06,
+      "loss": 0.9561,
+      "step": 469
+    },
+    {
+      "epoch": 0.27498445899001717,
+      "grad_norm": 0.04618878912895806,
+      "learning_rate": 7.669280792592383e-06,
+      "loss": 0.8403,
+      "step": 470
+    },
+    {
+      "epoch": 0.2755695323070172,
+      "grad_norm": 0.04343593914167416,
+      "learning_rate": 7.667894050197583e-06,
+      "loss": 0.8355,
+      "step": 471
+    },
+    {
+      "epoch": 0.2761546056240173,
+      "grad_norm": 0.040556959908718666,
+      "learning_rate": 7.66650454707513e-06,
+      "loss": 0.8608,
+      "step": 472
+    },
+    {
+      "epoch": 0.2767396789410173,
+      "grad_norm": 0.054697588596535354,
+      "learning_rate": 7.665112284398881e-06,
+      "loss": 0.839,
+      "step": 473
+    },
+    {
+      "epoch": 0.27732475225801734,
+      "grad_norm": 0.04393307328991595,
+      "learning_rate": 7.66371726334502e-06,
+      "loss": 0.8494,
+      "step": 474
+    },
+    {
+      "epoch": 0.27790982557501737,
+      "grad_norm": 0.04275309783082017,
+      "learning_rate": 7.662319485092067e-06,
+      "loss": 0.8259,
+      "step": 475
+    },
+    {
+      "epoch": 0.2784948988920174,
+      "grad_norm": 0.03953402161699839,
+      "learning_rate": 7.66091895082087e-06,
+      "loss": 0.7773,
+      "step": 476
+    },
+    {
+      "epoch": 0.2790799722090174,
+      "grad_norm": 0.03860846601474072,
+      "learning_rate": 7.659515661714608e-06,
+      "loss": 0.7962,
+      "step": 477
+    },
+    {
+      "epoch": 0.27966504552601745,
+      "grad_norm": 0.05366294387360709,
+      "learning_rate": 7.658109618958779e-06,
+      "loss": 0.9233,
+      "step": 478
+    },
+    {
+      "epoch": 0.28025011884301754,
+      "grad_norm": 0.04420018919513911,
+      "learning_rate": 7.656700823741216e-06,
+      "loss": 0.8156,
+      "step": 479
+    },
+    {
+      "epoch": 0.28083519216001757,
+      "grad_norm": 0.039049327120388326,
+      "learning_rate": 7.655289277252074e-06,
+      "loss": 0.8856,
+      "step": 480
+    },
+    {
+      "epoch": 0.2814202654770176,
+      "grad_norm": 0.26425632822736106,
+      "learning_rate": 7.653874980683828e-06,
+      "loss": 0.9503,
+      "step": 481
+    },
+    {
+      "epoch": 0.2820053387940176,
+      "grad_norm": 0.04289481203658443,
+      "learning_rate": 7.652457935231285e-06,
+      "loss": 0.8937,
+      "step": 482
+    },
+    {
+      "epoch": 0.28259041211101765,
+      "grad_norm": 0.04293700432214028,
+      "learning_rate": 7.651038142091568e-06,
+      "loss": 0.8571,
+      "step": 483
+    },
+    {
+      "epoch": 0.2831754854280177,
+      "grad_norm": 0.038244633724887706,
+      "learning_rate": 7.649615602464123e-06,
+      "loss": 0.7906,
+      "step": 484
+    },
+    {
+      "epoch": 0.2837605587450177,
+      "grad_norm": 0.03944291933937196,
+      "learning_rate": 7.648190317550717e-06,
+      "loss": 0.926,
+      "step": 485
+    },
+    {
+      "epoch": 0.2843456320620178,
+      "grad_norm": 0.10913228375108804,
+      "learning_rate": 7.646762288555433e-06,
+      "loss": 0.8592,
+      "step": 486
+    },
+    {
+      "epoch": 0.2849307053790178,
+      "grad_norm": 0.038286199243376444,
+      "learning_rate": 7.645331516684676e-06,
+      "loss": 0.8418,
+      "step": 487
+    },
+    {
+      "epoch": 0.28551577869601785,
+      "grad_norm": 0.043890441209433355,
+      "learning_rate": 7.643898003147167e-06,
+      "loss": 0.9614,
+      "step": 488
+    },
+    {
+      "epoch": 0.2861008520130179,
+      "grad_norm": 0.0382147078144481,
+      "learning_rate": 7.642461749153943e-06,
+      "loss": 0.83,
+      "step": 489
+    },
+    {
+      "epoch": 0.2866859253300179,
+      "grad_norm": 0.10202631306826099,
+      "learning_rate": 7.641022755918357e-06,
+      "loss": 0.9484,
+      "step": 490
+    },
+    {
+      "epoch": 0.28727099864701794,
+      "grad_norm": 0.04017584960145629,
+      "learning_rate": 7.639581024656072e-06,
+      "loss": 0.8278,
+      "step": 491
+    },
+    {
+      "epoch": 0.28785607196401797,
+      "grad_norm": 0.05571278346236126,
+      "learning_rate": 7.638136556585071e-06,
+      "loss": 0.9093,
+      "step": 492
+    },
+    {
+      "epoch": 0.28844114528101805,
+      "grad_norm": 0.04145030654143878,
+      "learning_rate": 7.636689352925643e-06,
+      "loss": 0.913,
+      "step": 493
+    },
+    {
+      "epoch": 0.2890262185980181,
+      "grad_norm": 0.05058778086388885,
+      "learning_rate": 7.635239414900393e-06,
+      "loss": 0.9366,
+      "step": 494
+    },
+    {
+      "epoch": 0.2896112919150181,
+      "grad_norm": 0.047476857954303966,
+      "learning_rate": 7.63378674373423e-06,
+      "loss": 0.8528,
+      "step": 495
+    },
+    {
+      "epoch": 0.29019636523201814,
+      "grad_norm": 0.043805250682478876,
+      "learning_rate": 7.632331340654377e-06,
+      "loss": 0.9953,
+      "step": 496
+    },
+    {
+      "epoch": 0.29078143854901817,
+      "grad_norm": 0.04691649904039108,
+      "learning_rate": 7.630873206890365e-06,
+      "loss": 0.8893,
+      "step": 497
+    },
+    {
+      "epoch": 0.2913665118660182,
+      "grad_norm": 0.061366723803107565,
+      "learning_rate": 7.629412343674026e-06,
+      "loss": 0.8895,
+      "step": 498
+    },
+    {
+      "epoch": 0.2919515851830182,
+      "grad_norm": 0.052047839285183604,
+      "learning_rate": 7.627948752239508e-06,
+      "loss": 0.8322,
+      "step": 499
+    },
+    {
+      "epoch": 0.2925366585000183,
+      "grad_norm": 0.04858401908225131,
+      "learning_rate": 7.6264824338232515e-06,
+      "loss": 0.7962,
+      "step": 500
+    },
+    {
+      "epoch": 0.29312173181701834,
+      "grad_norm": 0.05167192214353387,
+      "learning_rate": 7.625013389664012e-06,
+      "loss": 0.8178,
+      "step": 501
+    },
+    {
+      "epoch": 0.29370680513401837,
+      "grad_norm": 0.04571347435775933,
+      "learning_rate": 7.623541621002841e-06,
+      "loss": 0.9531,
+      "step": 502
+    },
+    {
+      "epoch": 0.2942918784510184,
+      "grad_norm": 0.05170428092011692,
+      "learning_rate": 7.622067129083092e-06,
+      "loss": 0.863,
+      "step": 503
+    },
+    {
+      "epoch": 0.2948769517680184,
+      "grad_norm": 0.04769020808960418,
+      "learning_rate": 7.620589915150423e-06,
+      "loss": 0.8693,
+      "step": 504
+    },
+    {
+      "epoch": 0.29546202508501845,
+      "grad_norm": 0.036447168096671895,
+      "learning_rate": 7.619109980452789e-06,
+      "loss": 0.8263,
+      "step": 505
+    },
+    {
+      "epoch": 0.2960470984020185,
+      "grad_norm": 0.04697422924818202,
+      "learning_rate": 7.617627326240441e-06,
+      "loss": 0.9127,
+      "step": 506
+    },
+    {
+      "epoch": 0.29663217171901857,
+      "grad_norm": 0.05026264821034648,
+      "learning_rate": 7.6161419537659345e-06,
+      "loss": 0.9163,
+      "step": 507
+    },
+    {
+      "epoch": 0.2972172450360186,
+      "grad_norm": 0.044605972655997896,
+      "learning_rate": 7.614653864284114e-06,
+      "loss": 0.8998,
+      "step": 508
+    },
+    {
+      "epoch": 0.2978023183530186,
+      "grad_norm": 0.06437245062396214,
+      "learning_rate": 7.613163059052123e-06,
+      "loss": 0.8039,
+      "step": 509
+    },
+    {
+      "epoch": 0.29838739167001865,
+      "grad_norm": 0.04554747767372419,
+      "learning_rate": 7.611669539329398e-06,
+      "loss": 0.8662,
+      "step": 510
+    },
+    {
+      "epoch": 0.2989724649870187,
+      "grad_norm": 0.0802330744693163,
+      "learning_rate": 7.610173306377671e-06,
+      "loss": 0.8343,
+      "step": 511
+    },
+    {
+      "epoch": 0.2995575383040187,
+      "grad_norm": 0.04192153945182111,
+      "learning_rate": 7.608674361460963e-06,
+      "loss": 0.8983,
+      "step": 512
+    },
+    {
+      "epoch": 0.30014261162101874,
+      "grad_norm": 0.04902593536059904,
+      "learning_rate": 7.607172705845589e-06,
+      "loss": 0.9242,
+      "step": 513
+    },
+    {
+      "epoch": 0.30072768493801877,
+      "grad_norm": 0.05436876049704265,
+      "learning_rate": 7.605668340800153e-06,
+      "loss": 0.834,
+      "step": 514
+    },
+    {
+      "epoch": 0.30131275825501885,
+      "grad_norm": 0.04047347680328035,
+      "learning_rate": 7.604161267595545e-06,
+      "loss": 0.8359,
+      "step": 515
+    },
+    {
+      "epoch": 0.3018978315720189,
+      "grad_norm": 0.03848909199643286,
+      "learning_rate": 7.602651487504946e-06,
+      "loss": 0.8126,
+      "step": 516
+    },
+    {
+      "epoch": 0.3024829048890189,
+      "grad_norm": 0.046207806514092946,
+      "learning_rate": 7.601139001803825e-06,
+      "loss": 0.869,
+      "step": 517
+    },
+    {
+      "epoch": 0.30306797820601894,
+      "grad_norm": 0.03852808749767389,
+      "learning_rate": 7.5996238117699344e-06,
+      "loss": 0.9808,
+      "step": 518
+    },
+    {
+      "epoch": 0.30365305152301897,
+      "grad_norm": 0.04005257123878126,
+      "learning_rate": 7.5981059186833114e-06,
+      "loss": 0.7767,
+      "step": 519
+    },
+    {
+      "epoch": 0.304238124840019,
+      "grad_norm": 0.05159848088030845,
+      "learning_rate": 7.596585323826277e-06,
+      "loss": 0.8932,
+      "step": 520
+    },
+    {
+      "epoch": 0.304823198157019,
+      "grad_norm": 0.040391024202221286,
+      "learning_rate": 7.595062028483434e-06,
+      "loss": 0.9093,
+      "step": 521
+    },
+    {
+      "epoch": 0.3054082714740191,
+      "grad_norm": 0.040428764785829324,
+      "learning_rate": 7.593536033941669e-06,
+      "loss": 0.9087,
+      "step": 522
+    },
+    {
+      "epoch": 0.30599334479101914,
+      "grad_norm": 0.09158364294178328,
+      "learning_rate": 7.592007341490145e-06,
+      "loss": 0.8601,
+      "step": 523
+    },
+    {
+      "epoch": 0.30657841810801917,
+      "grad_norm": 0.04136317840574456,
+      "learning_rate": 7.590475952420309e-06,
+      "loss": 0.8953,
+      "step": 524
+    },
+    {
+      "epoch": 0.3071634914250192,
+      "grad_norm": 0.06396409974499655,
+      "learning_rate": 7.588941868025881e-06,
+      "loss": 0.8297,
+      "step": 525
+    },
+    {
+      "epoch": 0.3077485647420192,
+      "grad_norm": 0.040333070246341814,
+      "learning_rate": 7.587405089602862e-06,
+      "loss": 0.7719,
+      "step": 526
+    },
+    {
+      "epoch": 0.30833363805901925,
+      "grad_norm": 0.04148918223122552,
+      "learning_rate": 7.585865618449528e-06,
+      "loss": 0.8007,
+      "step": 527
+    },
+    {
+      "epoch": 0.3089187113760193,
+      "grad_norm": 0.048132396423435676,
+      "learning_rate": 7.584323455866427e-06,
+      "loss": 0.8579,
+      "step": 528
+    },
+    {
+      "epoch": 0.30950378469301937,
+      "grad_norm": 0.04328208405834297,
+      "learning_rate": 7.582778603156387e-06,
+      "loss": 0.8071,
+      "step": 529
+    },
+    {
+      "epoch": 0.3100888580100194,
+      "grad_norm": 0.04318172969759895,
+      "learning_rate": 7.5812310616245e-06,
+      "loss": 0.83,
+      "step": 530
+    },
+    {
+      "epoch": 0.3106739313270194,
+      "grad_norm": 0.03768227605900526,
+      "learning_rate": 7.579680832578137e-06,
+      "loss": 0.8344,
+      "step": 531
+    },
+    {
+      "epoch": 0.31125900464401945,
+      "grad_norm": 0.04790982152353994,
+      "learning_rate": 7.578127917326936e-06,
+      "loss": 0.8974,
+      "step": 532
+    },
+    {
+      "epoch": 0.3118440779610195,
+      "grad_norm": 0.038002618502432514,
+      "learning_rate": 7.576572317182805e-06,
+      "loss": 0.792,
+      "step": 533
+    },
+    {
+      "epoch": 0.3124291512780195,
+      "grad_norm": 0.036571897121817694,
+      "learning_rate": 7.575014033459921e-06,
+      "loss": 0.8418,
+      "step": 534
+    },
+    {
+      "epoch": 0.31301422459501954,
+      "grad_norm": 0.03670017559323694,
+      "learning_rate": 7.573453067474724e-06,
+      "loss": 0.8834,
+      "step": 535
+    },
+    {
+      "epoch": 0.3135992979120196,
+      "grad_norm": 0.03621350431630049,
+      "learning_rate": 7.5718894205459284e-06,
+      "loss": 0.8549,
+      "step": 536
+    },
+    {
+      "epoch": 0.31418437122901965,
+      "grad_norm": 0.04296675197248163,
+      "learning_rate": 7.570323093994503e-06,
+      "loss": 0.7894,
+      "step": 537
+    },
+    {
+      "epoch": 0.3147694445460197,
+      "grad_norm": 0.044503765611357125,
+      "learning_rate": 7.568754089143688e-06,
+      "loss": 0.8655,
+      "step": 538
+    },
+    {
+      "epoch": 0.3153545178630197,
+      "grad_norm": 0.03715684392321602,
+      "learning_rate": 7.5671824073189845e-06,
+      "loss": 0.8697,
+      "step": 539
+    },
+    {
+      "epoch": 0.31593959118001974,
+      "grad_norm": 0.04056871817206291,
+      "learning_rate": 7.5656080498481535e-06,
+      "loss": 0.8803,
+      "step": 540
+    },
+    {
+      "epoch": 0.31652466449701977,
+      "grad_norm": 0.08089424036325778,
+      "learning_rate": 7.564031018061219e-06,
+      "loss": 0.8098,
+      "step": 541
+    },
+    {
+      "epoch": 0.3171097378140198,
+      "grad_norm": 0.04027917264460558,
+      "learning_rate": 7.562451313290459e-06,
+      "loss": 0.8939,
+      "step": 542
+    },
+    {
+      "epoch": 0.3176948111310199,
+      "grad_norm": 0.042852152415068484,
+      "learning_rate": 7.560868936870418e-06,
+      "loss": 0.8901,
+      "step": 543
+    },
+    {
+      "epoch": 0.3182798844480199,
+      "grad_norm": 0.05886634493495861,
+      "learning_rate": 7.559283890137889e-06,
+      "loss": 0.9286,
+      "step": 544
+    },
+    {
+      "epoch": 0.31886495776501994,
+      "grad_norm": 0.04216130498342499,
+      "learning_rate": 7.557696174431927e-06,
+      "loss": 0.8311,
+      "step": 545
+    },
+    {
+      "epoch": 0.31945003108201997,
+      "grad_norm": 0.053197981320438834,
+      "learning_rate": 7.556105791093838e-06,
+      "loss": 0.8952,
+      "step": 546
+    },
+    {
+      "epoch": 0.32003510439902,
+      "grad_norm": 0.038738558056749535,
+      "learning_rate": 7.554512741467183e-06,
+      "loss": 0.87,
+      "step": 547
+    },
+    {
+      "epoch": 0.32062017771602,
+      "grad_norm": 0.04255317774837396,
+      "learning_rate": 7.552917026897778e-06,
+      "loss": 0.8386,
+      "step": 548
+    },
+    {
+      "epoch": 0.32120525103302006,
+      "grad_norm": 0.06530091195329979,
+      "learning_rate": 7.551318648733684e-06,
+      "loss": 0.8862,
+      "step": 549
+    },
+    {
+      "epoch": 0.3217903243500201,
+      "grad_norm": 0.04640020878919504,
+      "learning_rate": 7.549717608325219e-06,
+      "loss": 0.8615,
+      "step": 550
+    },
+    {
+      "epoch": 0.32237539766702017,
+      "grad_norm": 0.038491606071055146,
+      "learning_rate": 7.548113907024948e-06,
+      "loss": 0.8581,
+      "step": 551
+    },
+    {
+      "epoch": 0.3229604709840202,
+      "grad_norm": 0.05470482648340031,
+      "learning_rate": 7.54650754618768e-06,
+      "loss": 0.8104,
+      "step": 552
+    },
+    {
+      "epoch": 0.3235455443010202,
+      "grad_norm": 0.04194329994008872,
+      "learning_rate": 7.544898527170475e-06,
+      "loss": 0.7725,
+      "step": 553
+    },
+    {
+      "epoch": 0.32413061761802026,
+      "grad_norm": 0.059518370147140835,
+      "learning_rate": 7.543286851332641e-06,
+      "loss": 0.8814,
+      "step": 554
+    },
+    {
+      "epoch": 0.3247156909350203,
+      "grad_norm": 0.041939649514720574,
+      "learning_rate": 7.5416725200357215e-06,
+      "loss": 0.8785,
+      "step": 555
+    },
+    {
+      "epoch": 0.3253007642520203,
+      "grad_norm": 0.038119584353072236,
+      "learning_rate": 7.540055534643512e-06,
+      "loss": 0.8787,
+      "step": 556
+    },
+    {
+      "epoch": 0.32588583756902034,
+      "grad_norm": 0.35043742170562575,
+      "learning_rate": 7.538435896522048e-06,
+      "loss": 0.9425,
+      "step": 557
+    },
+    {
+      "epoch": 0.3264709108860204,
+      "grad_norm": 0.037804212753556964,
+      "learning_rate": 7.536813607039603e-06,
+      "loss": 0.9275,
+      "step": 558
+    },
+    {
+      "epoch": 0.32705598420302046,
+      "grad_norm": 0.04797991190424691,
+      "learning_rate": 7.535188667566693e-06,
+      "loss": 0.8349,
+      "step": 559
+    },
+    {
+      "epoch": 0.3276410575200205,
+      "grad_norm": 0.07131490446157496,
+      "learning_rate": 7.533561079476073e-06,
+      "loss": 0.7922,
+      "step": 560
+    },
+    {
+      "epoch": 0.3282261308370205,
+      "grad_norm": 0.04273451205950798,
+      "learning_rate": 7.531930844142734e-06,
+      "loss": 0.9063,
+      "step": 561
+    },
+    {
+      "epoch": 0.32881120415402054,
+      "grad_norm": 0.04761650491169362,
+      "learning_rate": 7.5302979629439044e-06,
+      "loss": 0.874,
+      "step": 562
+    },
+    {
+      "epoch": 0.32939627747102057,
+      "grad_norm": 0.053621412028226095,
+      "learning_rate": 7.528662437259048e-06,
+      "loss": 0.8908,
+      "step": 563
+    },
+    {
+      "epoch": 0.3299813507880206,
+      "grad_norm": 0.08172735146274197,
+      "learning_rate": 7.527024268469862e-06,
+      "loss": 0.7873,
+      "step": 564
+    },
+    {
+      "epoch": 0.3305664241050207,
+      "grad_norm": 0.045936897060612714,
+      "learning_rate": 7.525383457960277e-06,
+      "loss": 0.8582,
+      "step": 565
+    },
+    {
+      "epoch": 0.3311514974220207,
+      "grad_norm": 0.043038357564032,
+      "learning_rate": 7.523740007116453e-06,
+      "loss": 0.8864,
+      "step": 566
+    },
+    {
+      "epoch": 0.33173657073902074,
+      "grad_norm": 0.0595338473222242,
+      "learning_rate": 7.5220939173267855e-06,
+      "loss": 0.8398,
+      "step": 567
+    },
+    {
+      "epoch": 0.33232164405602077,
+      "grad_norm": 0.03876298862367819,
+      "learning_rate": 7.520445189981897e-06,
+      "loss": 0.7608,
+      "step": 568
+    },
+    {
+      "epoch": 0.3329067173730208,
+      "grad_norm": 0.04137701061941132,
+      "learning_rate": 7.518793826474636e-06,
+      "loss": 0.9109,
+      "step": 569
+    },
+    {
+      "epoch": 0.33349179069002083,
+      "grad_norm": 0.04054950689452163,
+      "learning_rate": 7.517139828200079e-06,
+      "loss": 0.9109,
+      "step": 570
+    },
+    {
+      "epoch": 0.33407686400702086,
+      "grad_norm": 0.045942175432546556,
+      "learning_rate": 7.5154831965555315e-06,
+      "loss": 0.8569,
+      "step": 571
+    },
+    {
+      "epoch": 0.33466193732402094,
+      "grad_norm": 0.0412369088341243,
+      "learning_rate": 7.51382393294052e-06,
+      "loss": 0.837,
+      "step": 572
+    },
+    {
+      "epoch": 0.33524701064102097,
+      "grad_norm": 0.03984087759584128,
+      "learning_rate": 7.5121620387567955e-06,
+      "loss": 0.8173,
+      "step": 573
+    },
+    {
+      "epoch": 0.335832083958021,
+      "grad_norm": 0.03855595716614889,
+      "learning_rate": 7.510497515408333e-06,
+      "loss": 0.7942,
+      "step": 574
+    },
+    {
+      "epoch": 0.33641715727502103,
+      "grad_norm": 0.04123539453768172,
+      "learning_rate": 7.508830364301327e-06,
+      "loss": 0.9384,
+      "step": 575
+    },
+    {
+      "epoch": 0.33700223059202106,
+      "grad_norm": 0.047137253838708155,
+      "learning_rate": 7.507160586844191e-06,
+      "loss": 0.8449,
+      "step": 576
+    },
+    {
+      "epoch": 0.3375873039090211,
+      "grad_norm": 0.03987110874806033,
+      "learning_rate": 7.50548818444756e-06,
+      "loss": 0.8678,
+      "step": 577
+    },
+    {
+      "epoch": 0.3381723772260211,
+      "grad_norm": 0.05156972953667277,
+      "learning_rate": 7.503813158524284e-06,
+      "loss": 0.8369,
+      "step": 578
+    },
+    {
+      "epoch": 0.3387574505430212,
+      "grad_norm": 0.05429963825180294,
+      "learning_rate": 7.502135510489432e-06,
+      "loss": 0.9386,
+      "step": 579
+    },
+    {
+      "epoch": 0.33934252386002123,
+      "grad_norm": 0.04563390307441152,
+      "learning_rate": 7.500455241760284e-06,
+      "loss": 0.7892,
+      "step": 580
+    },
+    {
+      "epoch": 0.33992759717702126,
+      "grad_norm": 0.03772537720365855,
+      "learning_rate": 7.4987723537563395e-06,
+      "loss": 0.8171,
+      "step": 581
+    },
+    {
+      "epoch": 0.3405126704940213,
+      "grad_norm": 0.14591090698165343,
+      "learning_rate": 7.497086847899305e-06,
+      "loss": 0.9374,
+      "step": 582
+    },
+    {
+      "epoch": 0.3410977438110213,
+      "grad_norm": 0.04274276349409338,
+      "learning_rate": 7.495398725613103e-06,
+      "loss": 0.8886,
+      "step": 583
+    },
+    {
+      "epoch": 0.34168281712802134,
+      "grad_norm": 0.045419293797600416,
+      "learning_rate": 7.4937079883238644e-06,
+      "loss": 0.8874,
+      "step": 584
+    },
+    {
+      "epoch": 0.34226789044502137,
+      "grad_norm": 0.039369245868335276,
+      "learning_rate": 7.4920146374599305e-06,
+      "loss": 0.8515,
+      "step": 585
+    },
+    {
+      "epoch": 0.3428529637620214,
+      "grad_norm": 0.041038361870929824,
+      "learning_rate": 7.490318674451848e-06,
+      "loss": 0.8352,
+      "step": 586
+    },
+    {
+      "epoch": 0.3434380370790215,
+      "grad_norm": 0.03546313415290626,
+      "learning_rate": 7.488620100732373e-06,
+      "loss": 0.8126,
+      "step": 587
+    },
+    {
+      "epoch": 0.3440231103960215,
+      "grad_norm": 0.047686593783821035,
+      "learning_rate": 7.486918917736467e-06,
+      "loss": 0.8438,
+      "step": 588
+    },
+    {
+      "epoch": 0.34460818371302154,
+      "grad_norm": 0.03860874121220213,
+      "learning_rate": 7.485215126901294e-06,
+      "loss": 0.796,
+      "step": 589
+    },
+    {
+      "epoch": 0.34519325703002157,
+      "grad_norm": 0.04131878299287184,
+      "learning_rate": 7.483508729666222e-06,
+      "loss": 0.8787,
+      "step": 590
+    },
+    {
+      "epoch": 0.3457783303470216,
+      "grad_norm": 0.0626727371222094,
+      "learning_rate": 7.481799727472821e-06,
+      "loss": 0.8556,
+      "step": 591
+    },
+    {
+      "epoch": 0.34636340366402163,
+      "grad_norm": 0.046182038395419095,
+      "learning_rate": 7.480088121764862e-06,
+      "loss": 0.8362,
+      "step": 592
+    },
+    {
+      "epoch": 0.34694847698102166,
+      "grad_norm": 0.04057968397212824,
+      "learning_rate": 7.478373913988314e-06,
+      "loss": 0.8382,
+      "step": 593
+    },
+    {
+      "epoch": 0.34753355029802174,
+      "grad_norm": 0.04060993045831877,
+      "learning_rate": 7.476657105591347e-06,
+      "loss": 0.844,
+      "step": 594
+    },
+    {
+      "epoch": 0.34811862361502177,
+      "grad_norm": 0.042757904869809,
+      "learning_rate": 7.474937698024326e-06,
+      "loss": 0.8323,
+      "step": 595
+    },
+    {
+      "epoch": 0.3487036969320218,
+      "grad_norm": 0.042358138791525404,
+      "learning_rate": 7.4732156927398134e-06,
+      "loss": 0.8055,
+      "step": 596
+    },
+    {
+      "epoch": 0.34928877024902183,
+      "grad_norm": 0.04569798776629126,
+      "learning_rate": 7.4714910911925614e-06,
+      "loss": 0.7941,
+      "step": 597
+    },
+    {
+      "epoch": 0.34987384356602186,
+      "grad_norm": 0.0438874811170573,
+      "learning_rate": 7.469763894839523e-06,
+      "loss": 0.919,
+      "step": 598
+    },
+    {
+      "epoch": 0.3504589168830219,
+      "grad_norm": 0.06653725512945165,
+      "learning_rate": 7.468034105139836e-06,
+      "loss": 0.8574,
+      "step": 599
+    },
+    {
+      "epoch": 0.3510439902000219,
+      "grad_norm": 0.04995025638554342,
+      "learning_rate": 7.466301723554835e-06,
+      "loss": 0.8496,
+      "step": 600
+    },
+    {
+      "epoch": 0.351629063517022,
+      "grad_norm": 0.04283576163163911,
+      "learning_rate": 7.46456675154804e-06,
+      "loss": 0.8757,
+      "step": 601
+    },
+    {
+      "epoch": 0.35221413683402203,
+      "grad_norm": 0.039849657265063225,
+      "learning_rate": 7.462829190585162e-06,
+      "loss": 0.8945,
+      "step": 602
+    },
+    {
+      "epoch": 0.35279921015102206,
+      "grad_norm": 0.0440391834022003,
+      "learning_rate": 7.461089042134098e-06,
+      "loss": 0.8571,
+      "step": 603
+    },
+    {
+      "epoch": 0.3533842834680221,
+      "grad_norm": 0.043132549255425784,
+      "learning_rate": 7.45934630766493e-06,
+      "loss": 0.8668,
+      "step": 604
+    },
+    {
+      "epoch": 0.3539693567850221,
+      "grad_norm": 0.04249197374690922,
+      "learning_rate": 7.4576009886499285e-06,
+      "loss": 0.8797,
+      "step": 605
+    },
+    {
+      "epoch": 0.35455443010202214,
+      "grad_norm": 0.05324832438497396,
+      "learning_rate": 7.455853086563542e-06,
+      "loss": 0.9684,
+      "step": 606
+    },
+    {
+      "epoch": 0.3551395034190222,
+      "grad_norm": 0.04904764924574238,
+      "learning_rate": 7.454102602882405e-06,
+      "loss": 0.7753,
+      "step": 607
+    },
+    {
+      "epoch": 0.35572457673602226,
+      "grad_norm": 0.039014384352298336,
+      "learning_rate": 7.452349539085334e-06,
+      "loss": 0.8561,
+      "step": 608
+    },
+    {
+      "epoch": 0.3563096500530223,
+      "grad_norm": 0.03934058564150346,
+      "learning_rate": 7.4505938966533175e-06,
+      "loss": 0.8438,
+      "step": 609
+    },
+    {
+      "epoch": 0.3568947233700223,
+      "grad_norm": 0.05279799201688822,
+      "learning_rate": 7.448835677069536e-06,
+      "loss": 0.8912,
+      "step": 610
+    },
+    {
+      "epoch": 0.35747979668702234,
+      "grad_norm": 0.04001456755669222,
+      "learning_rate": 7.447074881819332e-06,
+      "loss": 0.8553,
+      "step": 611
+    },
+    {
+      "epoch": 0.3580648700040224,
+      "grad_norm": 0.042425970420024564,
+      "learning_rate": 7.445311512390233e-06,
+      "loss": 0.8327,
+      "step": 612
+    },
+    {
+      "epoch": 0.3586499433210224,
+      "grad_norm": 0.061113675404079935,
+      "learning_rate": 7.443545570271942e-06,
+      "loss": 0.8842,
+      "step": 613
+    },
+    {
+      "epoch": 0.35923501663802243,
+      "grad_norm": 0.046390573358408176,
+      "learning_rate": 7.44177705695633e-06,
+      "loss": 0.869,
+      "step": 614
+    },
+    {
+      "epoch": 0.3598200899550225,
+      "grad_norm": 0.04323809663682179,
+      "learning_rate": 7.440005973937445e-06,
+      "loss": 0.9184,
+      "step": 615
+    },
+    {
+      "epoch": 0.36040516327202254,
+      "grad_norm": 0.04915044460354856,
+      "learning_rate": 7.4382323227115e-06,
+      "loss": 0.8376,
+      "step": 616
+    },
+    {
+      "epoch": 0.3609902365890226,
+      "grad_norm": 0.0385546720018955,
+      "learning_rate": 7.436456104776885e-06,
+      "loss": 0.8515,
+      "step": 617
+    },
+    {
+      "epoch": 0.3615753099060226,
+      "grad_norm": 0.042848988320383044,
+      "learning_rate": 7.4346773216341545e-06,
+      "loss": 0.867,
+      "step": 618
+    },
+    {
+      "epoch": 0.36216038322302263,
+      "grad_norm": 0.04584623182025803,
+      "learning_rate": 7.432895974786029e-06,
+      "loss": 0.8974,
+      "step": 619
+    },
+    {
+      "epoch": 0.36274545654002266,
+      "grad_norm": 0.04282342241688822,
+      "learning_rate": 7.431112065737397e-06,
+      "loss": 0.8623,
+      "step": 620
+    },
+    {
+      "epoch": 0.3633305298570227,
+      "grad_norm": 0.04076113650735267,
+      "learning_rate": 7.429325595995311e-06,
+      "loss": 0.8682,
+      "step": 621
+    },
+    {
+      "epoch": 0.3639156031740227,
+      "grad_norm": 0.05126027130856797,
+      "learning_rate": 7.427536567068985e-06,
+      "loss": 0.906,
+      "step": 622
+    },
+    {
+      "epoch": 0.3645006764910228,
+      "grad_norm": 0.049318751888490414,
+      "learning_rate": 7.4257449804697975e-06,
+      "loss": 0.8397,
+      "step": 623
+    },
+    {
+      "epoch": 0.36508574980802283,
+      "grad_norm": 0.043254726523200684,
+      "learning_rate": 7.423950837711287e-06,
+      "loss": 0.8622,
+      "step": 624
+    },
+    {
+      "epoch": 0.36567082312502286,
+      "grad_norm": 0.05038943712513383,
+      "learning_rate": 7.422154140309151e-06,
+      "loss": 0.8775,
+      "step": 625
+    },
+    {
+      "epoch": 0.3662558964420229,
+      "grad_norm": 0.04134166981874639,
+      "learning_rate": 7.420354889781245e-06,
+      "loss": 0.8226,
+      "step": 626
+    },
+    {
+      "epoch": 0.3668409697590229,
+      "grad_norm": 0.03866147658091302,
+      "learning_rate": 7.418553087647582e-06,
+      "loss": 0.8664,
+      "step": 627
+    },
+    {
+      "epoch": 0.36742604307602295,
+      "grad_norm": 0.038894973852294755,
+      "learning_rate": 7.416748735430332e-06,
+      "loss": 0.8796,
+      "step": 628
+    },
+    {
+      "epoch": 0.368011116393023,
+      "grad_norm": 0.05664602547094577,
+      "learning_rate": 7.4149418346538144e-06,
+      "loss": 0.9451,
+      "step": 629
+    },
+    {
+      "epoch": 0.36859618971002306,
+      "grad_norm": 0.03682659768264277,
+      "learning_rate": 7.413132386844507e-06,
+      "loss": 0.7877,
+      "step": 630
+    },
+    {
+      "epoch": 0.3691812630270231,
+      "grad_norm": 0.04070102583713315,
+      "learning_rate": 7.411320393531038e-06,
+      "loss": 0.8647,
+      "step": 631
+    },
+    {
+      "epoch": 0.3697663363440231,
+      "grad_norm": 0.037583772238652974,
+      "learning_rate": 7.4095058562441835e-06,
+      "loss": 0.8984,
+      "step": 632
+    },
+    {
+      "epoch": 0.37035140966102315,
+      "grad_norm": 0.03745618837783022,
+      "learning_rate": 7.407688776516873e-06,
+      "loss": 0.8077,
+      "step": 633
+    },
+    {
+      "epoch": 0.3709364829780232,
+      "grad_norm": 0.04719274188930425,
+      "learning_rate": 7.405869155884178e-06,
+      "loss": 0.7846,
+      "step": 634
+    },
+    {
+      "epoch": 0.3715215562950232,
+      "grad_norm": 0.04726483533088948,
+      "learning_rate": 7.404046995883322e-06,
+      "loss": 0.9625,
+      "step": 635
+    },
+    {
+      "epoch": 0.37210662961202323,
+      "grad_norm": 0.03718098121534363,
+      "learning_rate": 7.402222298053672e-06,
+      "loss": 0.7673,
+      "step": 636
+    },
+    {
+      "epoch": 0.3726917029290233,
+      "grad_norm": 0.041699609984716224,
+      "learning_rate": 7.400395063936738e-06,
+      "loss": 0.8846,
+      "step": 637
+    },
+    {
+      "epoch": 0.37327677624602335,
+      "grad_norm": 0.03781340458428561,
+      "learning_rate": 7.3985652950761734e-06,
+      "loss": 0.8589,
+      "step": 638
+    },
+    {
+      "epoch": 0.3738618495630234,
+      "grad_norm": 0.05169204738245716,
+      "learning_rate": 7.396732993017774e-06,
+      "loss": 0.8737,
+      "step": 639
+    },
+    {
+      "epoch": 0.3744469228800234,
+      "grad_norm": 0.040199437093993116,
+      "learning_rate": 7.394898159309474e-06,
+      "loss": 0.8432,
+      "step": 640
+    },
+    {
+      "epoch": 0.37503199619702343,
+      "grad_norm": 0.04129820957000115,
+      "learning_rate": 7.393060795501346e-06,
+      "loss": 0.8363,
+      "step": 641
+    },
+    {
+      "epoch": 0.37561706951402346,
+      "grad_norm": 0.04346950641443829,
+      "learning_rate": 7.391220903145602e-06,
+      "loss": 0.8008,
+      "step": 642
+    },
+    {
+      "epoch": 0.3762021428310235,
+      "grad_norm": 0.043967535702394124,
+      "learning_rate": 7.389378483796589e-06,
+      "loss": 0.8695,
+      "step": 643
+    },
+    {
+      "epoch": 0.3767872161480236,
+      "grad_norm": 0.037870985542036356,
+      "learning_rate": 7.387533539010789e-06,
+      "loss": 0.7874,
+      "step": 644
+    },
+    {
+      "epoch": 0.3773722894650236,
+      "grad_norm": 0.0394925511025262,
+      "learning_rate": 7.385686070346818e-06,
+      "loss": 0.8427,
+      "step": 645
+    },
+    {
+      "epoch": 0.37795736278202363,
+      "grad_norm": 0.03932506659821688,
+      "learning_rate": 7.383836079365423e-06,
+      "loss": 0.8773,
+      "step": 646
+    },
+    {
+      "epoch": 0.37854243609902366,
+      "grad_norm": 0.04174103140864924,
+      "learning_rate": 7.381983567629482e-06,
+      "loss": 0.8532,
+      "step": 647
+    },
+    {
+      "epoch": 0.3791275094160237,
+      "grad_norm": 0.03777546260763163,
+      "learning_rate": 7.380128536704003e-06,
+      "loss": 0.842,
+      "step": 648
+    },
+    {
+      "epoch": 0.3797125827330237,
+      "grad_norm": 0.038903962254009807,
+      "learning_rate": 7.378270988156122e-06,
+      "loss": 0.9141,
+      "step": 649
+    },
+    {
+      "epoch": 0.38029765605002375,
+      "grad_norm": 0.03931172416456551,
+      "learning_rate": 7.376410923555104e-06,
+      "loss": 0.8382,
+      "step": 650
+    },
+    {
+      "epoch": 0.38088272936702383,
+      "grad_norm": 0.03830193944287262,
+      "learning_rate": 7.374548344472336e-06,
+      "loss": 0.913,
+      "step": 651
+    },
+    {
+      "epoch": 0.38146780268402386,
+      "grad_norm": 0.05689973145188307,
+      "learning_rate": 7.372683252481333e-06,
+      "loss": 0.9233,
+      "step": 652
+    },
+    {
+      "epoch": 0.3820528760010239,
+      "grad_norm": 0.11927191175958904,
+      "learning_rate": 7.370815649157728e-06,
+      "loss": 0.8497,
+      "step": 653
+    },
+    {
+      "epoch": 0.3826379493180239,
+      "grad_norm": 0.04879685015775887,
+      "learning_rate": 7.36894553607928e-06,
+      "loss": 0.8902,
+      "step": 654
+    },
+    {
+      "epoch": 0.38322302263502395,
+      "grad_norm": 0.038545163591213454,
+      "learning_rate": 7.3670729148258655e-06,
+      "loss": 0.8101,
+      "step": 655
+    },
+    {
+      "epoch": 0.383808095952024,
+      "grad_norm": 0.07018595700302209,
+      "learning_rate": 7.365197786979483e-06,
+      "loss": 0.8732,
+      "step": 656
+    },
+    {
+      "epoch": 0.384393169269024,
+      "grad_norm": 0.03883816516578689,
+      "learning_rate": 7.3633201541242465e-06,
+      "loss": 0.9438,
+      "step": 657
+    },
+    {
+      "epoch": 0.3849782425860241,
+      "grad_norm": 0.045681207579253334,
+      "learning_rate": 7.3614400178463834e-06,
+      "loss": 0.8083,
+      "step": 658
+    },
+    {
+      "epoch": 0.3855633159030241,
+      "grad_norm": 0.03536545477300677,
+      "learning_rate": 7.359557379734242e-06,
+      "loss": 0.7559,
+      "step": 659
+    },
+    {
+      "epoch": 0.38614838922002415,
+      "grad_norm": 0.03989482911665265,
+      "learning_rate": 7.357672241378282e-06,
+      "loss": 0.8969,
+      "step": 660
+    },
+    {
+      "epoch": 0.3867334625370242,
+      "grad_norm": 0.03858611847647411,
+      "learning_rate": 7.355784604371071e-06,
+      "loss": 0.835,
+      "step": 661
+    },
+    {
+      "epoch": 0.3873185358540242,
+      "grad_norm": 0.04523699088789738,
+      "learning_rate": 7.353894470307294e-06,
+      "loss": 0.8641,
+      "step": 662
+    },
+    {
+      "epoch": 0.38790360917102423,
+      "grad_norm": 0.041661369856111206,
+      "learning_rate": 7.352001840783741e-06,
+      "loss": 0.834,
+      "step": 663
+    },
+    {
+      "epoch": 0.38848868248802426,
+      "grad_norm": 0.041538283343954825,
+      "learning_rate": 7.3501067173993115e-06,
+      "loss": 0.9114,
+      "step": 664
+    },
+    {
+      "epoch": 0.3890737558050243,
+      "grad_norm": 0.039602491788351755,
+      "learning_rate": 7.348209101755012e-06,
+      "loss": 0.8479,
+      "step": 665
+    },
+    {
+      "epoch": 0.3896588291220244,
+      "grad_norm": 0.04077290004316095,
+      "learning_rate": 7.346308995453956e-06,
+      "loss": 0.8511,
+      "step": 666
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 0.044595088605505394,
+      "learning_rate": 7.344406400101358e-06,
+      "loss": 0.8121,
+      "step": 667
+    },
+    {
+      "epoch": 0.39082897575602443,
+      "grad_norm": 0.04520053665259006,
+      "learning_rate": 7.342501317304538e-06,
+      "loss": 0.8916,
+      "step": 668
+    },
+    {
+      "epoch": 0.39141404907302446,
+      "grad_norm": 0.04048302305644283,
+      "learning_rate": 7.340593748672915e-06,
+      "loss": 0.9092,
+      "step": 669
+    },
+    {
+      "epoch": 0.3919991223900245,
+      "grad_norm": 0.04037233961111267,
+      "learning_rate": 7.33868369581801e-06,
+      "loss": 0.881,
+      "step": 670
+    },
+    {
+      "epoch": 0.3925841957070245,
+      "grad_norm": 0.05051075945293654,
+      "learning_rate": 7.336771160353441e-06,
+      "loss": 0.8109,
+      "step": 671
+    },
+    {
+      "epoch": 0.39316926902402455,
+      "grad_norm": 0.049551953529107955,
+      "learning_rate": 7.334856143894927e-06,
+      "loss": 0.9017,
+      "step": 672
+    },
+    {
+      "epoch": 0.39375434234102463,
+      "grad_norm": 0.03993075169550292,
+      "learning_rate": 7.332938648060276e-06,
+      "loss": 0.8542,
+      "step": 673
+    },
+    {
+      "epoch": 0.39433941565802466,
+      "grad_norm": 0.038451189858571745,
+      "learning_rate": 7.331018674469396e-06,
+      "loss": 0.8112,
+      "step": 674
+    },
+    {
+      "epoch": 0.3949244889750247,
+      "grad_norm": 0.0505499896286193,
+      "learning_rate": 7.329096224744291e-06,
+      "loss": 0.8813,
+      "step": 675
+    },
+    {
+      "epoch": 0.3955095622920247,
+      "grad_norm": 0.071098260548373,
+      "learning_rate": 7.3271713005090494e-06,
+      "loss": 0.7823,
+      "step": 676
+    },
+    {
+      "epoch": 0.39609463560902475,
+      "grad_norm": 0.039554913835090445,
+      "learning_rate": 7.325243903389853e-06,
+      "loss": 0.8741,
+      "step": 677
+    },
+    {
+      "epoch": 0.3966797089260248,
+      "grad_norm": 0.03860100123135944,
+      "learning_rate": 7.323314035014974e-06,
+      "loss": 0.8042,
+      "step": 678
+    },
+    {
+      "epoch": 0.3972647822430248,
+      "grad_norm": 0.043534126452199624,
+      "learning_rate": 7.321381697014771e-06,
+      "loss": 0.8062,
+      "step": 679
+    },
+    {
+      "epoch": 0.3978498555600249,
+      "grad_norm": 0.04429511697906778,
+      "learning_rate": 7.319446891021693e-06,
+      "loss": 0.8726,
+      "step": 680
+    },
+    {
+      "epoch": 0.3984349288770249,
+      "grad_norm": 0.03812425577668118,
+      "learning_rate": 7.317509618670267e-06,
+      "loss": 0.8136,
+      "step": 681
+    },
+    {
+      "epoch": 0.39902000219402495,
+      "grad_norm": 0.07345134546444418,
+      "learning_rate": 7.315569881597106e-06,
+      "loss": 0.8483,
+      "step": 682
+    },
+    {
+      "epoch": 0.399605075511025,
+      "grad_norm": 0.038808207078641584,
+      "learning_rate": 7.313627681440909e-06,
+      "loss": 0.9122,
+      "step": 683
+    },
+    {
+      "epoch": 0.400190148828025,
+      "grad_norm": 0.03882381554371386,
+      "learning_rate": 7.311683019842453e-06,
+      "loss": 0.8767,
+      "step": 684
+    },
+    {
+      "epoch": 0.40077522214502503,
+      "grad_norm": 0.09532542796696841,
+      "learning_rate": 7.309735898444593e-06,
+      "loss": 0.817,
+      "step": 685
+    },
+    {
+      "epoch": 0.40136029546202506,
+      "grad_norm": 0.0401829342578114,
+      "learning_rate": 7.307786318892265e-06,
+      "loss": 0.9071,
+      "step": 686
+    },
+    {
+      "epoch": 0.40194536877902515,
+      "grad_norm": 0.03903463743661068,
+      "learning_rate": 7.305834282832478e-06,
+      "loss": 0.9161,
+      "step": 687
+    },
+    {
+      "epoch": 0.4025304420960252,
+      "grad_norm": 0.03813349508001005,
+      "learning_rate": 7.303879791914321e-06,
+      "loss": 0.7688,
+      "step": 688
+    },
+    {
+      "epoch": 0.4031155154130252,
+      "grad_norm": 0.03801357430065186,
+      "learning_rate": 7.301922847788953e-06,
+      "loss": 0.7473,
+      "step": 689
+    },
+    {
+      "epoch": 0.40370058873002523,
+      "grad_norm": 0.0391358430740653,
+      "learning_rate": 7.299963452109607e-06,
+      "loss": 0.84,
+      "step": 690
+    },
+    {
+      "epoch": 0.40428566204702526,
+      "grad_norm": 0.04237700699339236,
+      "learning_rate": 7.298001606531588e-06,
+      "loss": 0.7707,
+      "step": 691
+    },
+    {
+      "epoch": 0.4048707353640253,
+      "grad_norm": 0.04484671718594054,
+      "learning_rate": 7.296037312712267e-06,
+      "loss": 0.8456,
+      "step": 692
+    },
+    {
+      "epoch": 0.4054558086810253,
+      "grad_norm": 0.04025195813747376,
+      "learning_rate": 7.2940705723110895e-06,
+      "loss": 0.8882,
+      "step": 693
+    },
+    {
+      "epoch": 0.4060408819980254,
+      "grad_norm": 0.2399660293834239,
+      "learning_rate": 7.292101386989561e-06,
+      "loss": 0.8086,
+      "step": 694
+    },
+    {
+      "epoch": 0.40662595531502543,
+      "grad_norm": 0.03485270510298816,
+      "learning_rate": 7.290129758411258e-06,
+      "loss": 0.8997,
+      "step": 695
+    },
+    {
+      "epoch": 0.40721102863202546,
+      "grad_norm": 0.03819929313036679,
+      "learning_rate": 7.288155688241819e-06,
+      "loss": 0.9212,
+      "step": 696
+    },
+    {
+      "epoch": 0.4077961019490255,
+      "grad_norm": 0.03510738864060292,
+      "learning_rate": 7.286179178148942e-06,
+      "loss": 0.8754,
+      "step": 697
+    },
+    {
+      "epoch": 0.4083811752660255,
+      "grad_norm": 0.043632341643109356,
+      "learning_rate": 7.284200229802391e-06,
+      "loss": 0.7717,
+      "step": 698
+    },
+    {
+      "epoch": 0.40896624858302555,
+      "grad_norm": 0.04089651755929757,
+      "learning_rate": 7.28221884487399e-06,
+      "loss": 0.88,
+      "step": 699
+    },
+    {
+      "epoch": 0.4095513219000256,
+      "grad_norm": 0.037780685942199126,
+      "learning_rate": 7.280235025037616e-06,
+      "loss": 0.8197,
+      "step": 700
+    },
+    {
+      "epoch": 0.4101363952170256,
+      "grad_norm": 0.03655073966270369,
+      "learning_rate": 7.27824877196921e-06,
+      "loss": 0.8279,
+      "step": 701
+    },
+    {
+      "epoch": 0.4107214685340257,
+      "grad_norm": 0.05600910511927575,
+      "learning_rate": 7.2762600873467624e-06,
+      "loss": 0.8476,
+      "step": 702
+    },
+    {
+      "epoch": 0.4113065418510257,
+      "grad_norm": 0.034951743393564536,
+      "learning_rate": 7.274268972850321e-06,
+      "loss": 0.7609,
+      "step": 703
+    },
+    {
+      "epoch": 0.41189161516802575,
+      "grad_norm": 0.09106547558232417,
+      "learning_rate": 7.272275430161988e-06,
+      "loss": 0.8996,
+      "step": 704
+    },
+    {
+      "epoch": 0.4124766884850258,
+      "grad_norm": 0.051359026093647085,
+      "learning_rate": 7.270279460965912e-06,
+      "loss": 0.9052,
+      "step": 705
+    },
+    {
+      "epoch": 0.4130617618020258,
+      "grad_norm": 0.03905089647983441,
+      "learning_rate": 7.268281066948296e-06,
+      "loss": 0.918,
+      "step": 706
+    },
+    {
+      "epoch": 0.41364683511902584,
+      "grad_norm": 0.04206925775516373,
+      "learning_rate": 7.2662802497973875e-06,
+      "loss": 0.8053,
+      "step": 707
+    },
+    {
+      "epoch": 0.41423190843602586,
+      "grad_norm": 0.040474218310822804,
+      "learning_rate": 7.264277011203488e-06,
+      "loss": 0.891,
+      "step": 708
+    },
+    {
+      "epoch": 0.41481698175302595,
+      "grad_norm": 0.0442335071782456,
+      "learning_rate": 7.262271352858936e-06,
+      "loss": 0.8593,
+      "step": 709
+    },
+    {
+      "epoch": 0.415402055070026,
+      "grad_norm": 0.047122148951545366,
+      "learning_rate": 7.26026327645812e-06,
+      "loss": 0.844,
+      "step": 710
+    },
+    {
+      "epoch": 0.415987128387026,
+      "grad_norm": 0.03850054120717149,
+      "learning_rate": 7.258252783697469e-06,
+      "loss": 0.7795,
+      "step": 711
+    },
+    {
+      "epoch": 0.41657220170402603,
+      "grad_norm": 0.050566843036294336,
+      "learning_rate": 7.2562398762754554e-06,
+      "loss": 0.7794,
+      "step": 712
+    },
+    {
+      "epoch": 0.41715727502102606,
+      "grad_norm": 0.058106581796564256,
+      "learning_rate": 7.254224555892587e-06,
+      "loss": 0.9735,
+      "step": 713
+    },
+    {
+      "epoch": 0.4177423483380261,
+      "grad_norm": 0.03831983478639148,
+      "learning_rate": 7.252206824251416e-06,
+      "loss": 0.7832,
+      "step": 714
+    },
+    {
+      "epoch": 0.4183274216550261,
+      "grad_norm": 0.0431158884889512,
+      "learning_rate": 7.250186683056527e-06,
+      "loss": 0.7865,
+      "step": 715
+    },
+    {
+      "epoch": 0.4189124949720262,
+      "grad_norm": 0.046081855255802225,
+      "learning_rate": 7.248164134014544e-06,
+      "loss": 0.8512,
+      "step": 716
+    },
+    {
+      "epoch": 0.41949756828902623,
+      "grad_norm": 0.04502443182272067,
+      "learning_rate": 7.246139178834119e-06,
+      "loss": 0.853,
+      "step": 717
+    },
+    {
+      "epoch": 0.42008264160602626,
+      "grad_norm": 0.045461255479903595,
+      "learning_rate": 7.244111819225946e-06,
+      "loss": 0.8262,
+      "step": 718
+    },
+    {
+      "epoch": 0.4206677149230263,
+      "grad_norm": 0.038558457568846585,
+      "learning_rate": 7.24208205690274e-06,
+      "loss": 0.7975,
+      "step": 719
+    },
+    {
+      "epoch": 0.4212527882400263,
+      "grad_norm": 0.04654917549136812,
+      "learning_rate": 7.240049893579256e-06,
+      "loss": 0.8849,
+      "step": 720
+    },
+    {
+      "epoch": 0.42183786155702635,
+      "grad_norm": 0.04307894480353439,
+      "learning_rate": 7.238015330972268e-06,
+      "loss": 0.805,
+      "step": 721
+    },
+    {
+      "epoch": 0.4224229348740264,
+      "grad_norm": 0.04281636756195477,
+      "learning_rate": 7.235978370800583e-06,
+      "loss": 0.8471,
+      "step": 722
+    },
+    {
+      "epoch": 0.42300800819102646,
+      "grad_norm": 0.04507551309506065,
+      "learning_rate": 7.233939014785032e-06,
+      "loss": 0.8468,
+      "step": 723
+    },
+    {
+      "epoch": 0.4235930815080265,
+      "grad_norm": 0.06918610285412721,
+      "learning_rate": 7.2318972646484685e-06,
+      "loss": 0.8655,
+      "step": 724
+    },
+    {
+      "epoch": 0.4241781548250265,
+      "grad_norm": 0.04721791843997281,
+      "learning_rate": 7.229853122115772e-06,
+      "loss": 0.7927,
+      "step": 725
+    },
+    {
+      "epoch": 0.42476322814202655,
+      "grad_norm": 0.041834283551617495,
+      "learning_rate": 7.227806588913838e-06,
+      "loss": 0.8712,
+      "step": 726
+    },
+    {
+      "epoch": 0.4253483014590266,
+      "grad_norm": 0.05505118490346748,
+      "learning_rate": 7.225757666771585e-06,
+      "loss": 0.8584,
+      "step": 727
+    },
+    {
+      "epoch": 0.4259333747760266,
+      "grad_norm": 0.047773194756069906,
+      "learning_rate": 7.223706357419951e-06,
+      "loss": 0.7893,
+      "step": 728
+    },
+    {
+      "epoch": 0.42651844809302664,
+      "grad_norm": 0.04822106540139186,
+      "learning_rate": 7.221652662591887e-06,
+      "loss": 0.8277,
+      "step": 729
+    },
+    {
+      "epoch": 0.4271035214100267,
+      "grad_norm": 0.03969595772236796,
+      "learning_rate": 7.219596584022363e-06,
+      "loss": 0.8394,
+      "step": 730
+    },
+    {
+      "epoch": 0.42768859472702675,
+      "grad_norm": 0.0422480401855194,
+      "learning_rate": 7.217538123448359e-06,
+      "loss": 0.8094,
+      "step": 731
+    },
+    {
+      "epoch": 0.4282736680440268,
+      "grad_norm": 0.0394391284081235,
+      "learning_rate": 7.215477282608871e-06,
+      "loss": 0.8847,
+      "step": 732
+    },
+    {
+      "epoch": 0.4288587413610268,
+      "grad_norm": 0.05956332380443304,
+      "learning_rate": 7.213414063244903e-06,
+      "loss": 0.8427,
+      "step": 733
+    },
+    {
+      "epoch": 0.42944381467802684,
+      "grad_norm": 0.03974032496622027,
+      "learning_rate": 7.21134846709947e-06,
+      "loss": 0.7986,
+      "step": 734
+    },
+    {
+      "epoch": 0.43002888799502687,
+      "grad_norm": 0.061979287274211155,
+      "learning_rate": 7.209280495917594e-06,
+      "loss": 0.8178,
+      "step": 735
+    },
+    {
+      "epoch": 0.4306139613120269,
+      "grad_norm": 0.04099124105139921,
+      "learning_rate": 7.2072101514463045e-06,
+      "loss": 0.7936,
+      "step": 736
+    },
+    {
+      "epoch": 0.4311990346290269,
+      "grad_norm": 0.04101203671080088,
+      "learning_rate": 7.205137435434634e-06,
+      "loss": 0.8607,
+      "step": 737
+    },
+    {
+      "epoch": 0.431784107946027,
+      "grad_norm": 0.04063679329152861,
+      "learning_rate": 7.203062349633622e-06,
+      "loss": 0.8066,
+      "step": 738
+    },
+    {
+      "epoch": 0.43236918126302704,
+      "grad_norm": 0.045567608816787585,
+      "learning_rate": 7.200984895796305e-06,
+      "loss": 0.8558,
+      "step": 739
+    },
+    {
+      "epoch": 0.43295425458002706,
+      "grad_norm": 0.0514210080543997,
+      "learning_rate": 7.198905075677726e-06,
+      "loss": 0.7855,
+      "step": 740
+    },
+    {
+      "epoch": 0.4335393278970271,
+      "grad_norm": 0.0404180379498117,
+      "learning_rate": 7.196822891034922e-06,
+      "loss": 0.9028,
+      "step": 741
+    },
+    {
+      "epoch": 0.4341244012140271,
+      "grad_norm": 0.03964126441343688,
+      "learning_rate": 7.1947383436269295e-06,
+      "loss": 0.872,
+      "step": 742
+    },
+    {
+      "epoch": 0.43470947453102715,
+      "grad_norm": 0.04863222203139086,
+      "learning_rate": 7.192651435214781e-06,
+      "loss": 0.9288,
+      "step": 743
+    },
+    {
+      "epoch": 0.4352945478480272,
+      "grad_norm": 0.04412474170590896,
+      "learning_rate": 7.190562167561505e-06,
+      "loss": 0.836,
+      "step": 744
+    },
+    {
+      "epoch": 0.43587962116502726,
+      "grad_norm": 0.16048273545352948,
+      "learning_rate": 7.188470542432119e-06,
+      "loss": 0.7639,
+      "step": 745
+    },
+    {
+      "epoch": 0.4364646944820273,
+      "grad_norm": 0.04767484164607536,
+      "learning_rate": 7.1863765615936375e-06,
+      "loss": 0.8481,
+      "step": 746
+    },
+    {
+      "epoch": 0.4370497677990273,
+      "grad_norm": 0.06989126022408862,
+      "learning_rate": 7.184280226815061e-06,
+      "loss": 0.8569,
+      "step": 747
+    },
+    {
+      "epoch": 0.43763484111602735,
+      "grad_norm": 0.041900678395983416,
+      "learning_rate": 7.18218153986738e-06,
+      "loss": 0.931,
+      "step": 748
+    },
+    {
+      "epoch": 0.4382199144330274,
+      "grad_norm": 0.04216846873986442,
+      "learning_rate": 7.180080502523572e-06,
+      "loss": 0.8646,
+      "step": 749
+    },
+    {
+      "epoch": 0.4388049877500274,
+      "grad_norm": 0.050264349955050885,
+      "learning_rate": 7.177977116558601e-06,
+      "loss": 0.8199,
+      "step": 750
+    },
+    {
+      "epoch": 0.43939006106702744,
+      "grad_norm": 0.03843848162959958,
+      "learning_rate": 7.175871383749415e-06,
+      "loss": 0.8097,
+      "step": 751
+    },
+    {
+      "epoch": 0.4399751343840275,
+      "grad_norm": 0.05545552486681674,
+      "learning_rate": 7.173763305874942e-06,
+      "loss": 0.9036,
+      "step": 752
+    },
+    {
+      "epoch": 0.44056020770102755,
+      "grad_norm": 0.040372444436672136,
+      "learning_rate": 7.1716528847160944e-06,
+      "loss": 0.7861,
+      "step": 753
+    },
+    {
+      "epoch": 0.4411452810180276,
+      "grad_norm": 0.03823936143241982,
+      "learning_rate": 7.169540122055764e-06,
+      "loss": 0.7976,
+      "step": 754
+    },
+    {
+      "epoch": 0.4417303543350276,
+      "grad_norm": 0.040595268808395715,
+      "learning_rate": 7.167425019678817e-06,
+      "loss": 0.8007,
+      "step": 755
+    },
+    {
+      "epoch": 0.44231542765202764,
+      "grad_norm": 0.04580612084615541,
+      "learning_rate": 7.1653075793721e-06,
+      "loss": 0.741,
+      "step": 756
+    },
+    {
+      "epoch": 0.44290050096902767,
+      "grad_norm": 0.044516769340694984,
+      "learning_rate": 7.163187802924435e-06,
+      "loss": 0.911,
+      "step": 757
+    },
+    {
+      "epoch": 0.4434855742860277,
+      "grad_norm": 0.04502988747564978,
+      "learning_rate": 7.161065692126614e-06,
+      "loss": 0.8775,
+      "step": 758
+    },
+    {
+      "epoch": 0.4440706476030278,
+      "grad_norm": 0.041288645539491076,
+      "learning_rate": 7.1589412487714055e-06,
+      "loss": 0.8393,
+      "step": 759
+    },
+    {
+      "epoch": 0.4446557209200278,
+      "grad_norm": 0.04219916867709284,
+      "learning_rate": 7.156814474653542e-06,
+      "loss": 0.8203,
+      "step": 760
+    },
+    {
+      "epoch": 0.44524079423702784,
+      "grad_norm": 0.04510176879579754,
+      "learning_rate": 7.154685371569736e-06,
+      "loss": 0.7882,
+      "step": 761
+    },
+    {
+      "epoch": 0.44582586755402787,
+      "grad_norm": 0.03956297558604167,
+      "learning_rate": 7.152553941318655e-06,
+      "loss": 0.8313,
+      "step": 762
+    },
+    {
+      "epoch": 0.4464109408710279,
+      "grad_norm": 0.03781666670962381,
+      "learning_rate": 7.15042018570094e-06,
+      "loss": 0.9057,
+      "step": 763
+    },
+    {
+      "epoch": 0.4469960141880279,
+      "grad_norm": 0.04278185971588543,
+      "learning_rate": 7.148284106519195e-06,
+      "loss": 0.842,
+      "step": 764
+    },
+    {
+      "epoch": 0.44758108750502795,
+      "grad_norm": 0.03726277158845805,
+      "learning_rate": 7.1461457055779875e-06,
+      "loss": 0.8003,
+      "step": 765
+    },
+    {
+      "epoch": 0.44816616082202804,
+      "grad_norm": 0.044396475499465786,
+      "learning_rate": 7.144004984683844e-06,
+      "loss": 0.8393,
+      "step": 766
+    },
+    {
+      "epoch": 0.44875123413902807,
+      "grad_norm": 0.03484609580213948,
+      "learning_rate": 7.141861945645254e-06,
+      "loss": 0.8255,
+      "step": 767
+    },
+    {
+      "epoch": 0.4493363074560281,
+      "grad_norm": 0.04524092857108677,
+      "learning_rate": 7.139716590272663e-06,
+      "loss": 0.7811,
+      "step": 768
+    },
+    {
+      "epoch": 0.4499213807730281,
+      "grad_norm": 0.03734702819768394,
+      "learning_rate": 7.1375689203784755e-06,
+      "loss": 0.8644,
+      "step": 769
+    },
+    {
+      "epoch": 0.45050645409002815,
+      "grad_norm": 0.05019144085295421,
+      "learning_rate": 7.135418937777049e-06,
+      "loss": 0.9044,
+      "step": 770
+    },
+    {
+      "epoch": 0.4510915274070282,
+      "grad_norm": 0.04029746203545264,
+      "learning_rate": 7.133266644284696e-06,
+      "loss": 0.7769,
+      "step": 771
+    },
+    {
+      "epoch": 0.4516766007240282,
+      "grad_norm": 0.04262838461849833,
+      "learning_rate": 7.131112041719681e-06,
+      "loss": 0.9143,
+      "step": 772
+    },
+    {
+      "epoch": 0.45226167404102824,
+      "grad_norm": 0.043876149904780545,
+      "learning_rate": 7.1289551319022195e-06,
+      "loss": 0.8828,
+      "step": 773
+    },
+    {
+      "epoch": 0.4528467473580283,
+      "grad_norm": 0.04711071987260168,
+      "learning_rate": 7.126795916654477e-06,
+      "loss": 0.7762,
+      "step": 774
+    },
+    {
+      "epoch": 0.45343182067502835,
+      "grad_norm": 0.04116549252293238,
+      "learning_rate": 7.124634397800565e-06,
+      "loss": 0.7778,
+      "step": 775
+    },
+    {
+      "epoch": 0.4540168939920284,
+      "grad_norm": 0.0393006386390472,
+      "learning_rate": 7.1224705771665405e-06,
+      "loss": 0.8465,
+      "step": 776
+    },
+    {
+      "epoch": 0.4546019673090284,
+      "grad_norm": 0.03826784736833335,
+      "learning_rate": 7.120304456580408e-06,
+      "loss": 0.8359,
+      "step": 777
+    },
+    {
+      "epoch": 0.45518704062602844,
+      "grad_norm": 0.04985454473324124,
+      "learning_rate": 7.118136037872112e-06,
+      "loss": 0.8552,
+      "step": 778
+    },
+    {
+      "epoch": 0.45577211394302847,
+      "grad_norm": 0.04012812213905606,
+      "learning_rate": 7.115965322873541e-06,
+      "loss": 0.8249,
+      "step": 779
+    },
+    {
+      "epoch": 0.4563571872600285,
+      "grad_norm": 0.03736935441616661,
+      "learning_rate": 7.113792313418522e-06,
+      "loss": 0.8399,
+      "step": 780
+    },
+    {
+      "epoch": 0.4569422605770286,
+      "grad_norm": 0.036669437710784104,
+      "learning_rate": 7.1116170113428194e-06,
+      "loss": 0.79,
+      "step": 781
+    },
+    {
+      "epoch": 0.4575273338940286,
+      "grad_norm": 0.03882415136861196,
+      "learning_rate": 7.109439418484137e-06,
+      "loss": 0.8016,
+      "step": 782
+    },
+    {
+      "epoch": 0.45811240721102864,
+      "grad_norm": 0.0452764493144211,
+      "learning_rate": 7.107259536682111e-06,
+      "loss": 0.8138,
+      "step": 783
+    },
+    {
+      "epoch": 0.45869748052802867,
+      "grad_norm": 0.04729394522582173,
+      "learning_rate": 7.105077367778313e-06,
+      "loss": 0.7596,
+      "step": 784
+    },
+    {
+      "epoch": 0.4592825538450287,
+      "grad_norm": 0.05019146434388651,
+      "learning_rate": 7.102892913616248e-06,
+      "loss": 0.8015,
+      "step": 785
+    },
+    {
+      "epoch": 0.4598676271620287,
+      "grad_norm": 0.039099244411788574,
+      "learning_rate": 7.100706176041348e-06,
+      "loss": 0.8098,
+      "step": 786
+    },
+    {
+      "epoch": 0.46045270047902875,
+      "grad_norm": 0.03895433813179543,
+      "learning_rate": 7.098517156900978e-06,
+      "loss": 0.8851,
+      "step": 787
+    },
+    {
+      "epoch": 0.46103777379602884,
+      "grad_norm": 0.03731969942911145,
+      "learning_rate": 7.096325858044427e-06,
+      "loss": 0.7721,
+      "step": 788
+    },
+    {
+      "epoch": 0.46162284711302887,
+      "grad_norm": 0.04097182225913861,
+      "learning_rate": 7.094132281322912e-06,
+      "loss": 0.8223,
+      "step": 789
+    },
+    {
+      "epoch": 0.4622079204300289,
+      "grad_norm": 0.03696891399238777,
+      "learning_rate": 7.091936428589576e-06,
+      "loss": 0.8938,
+      "step": 790
+    },
+    {
+      "epoch": 0.4627929937470289,
+      "grad_norm": 0.03792585935917287,
+      "learning_rate": 7.089738301699479e-06,
+      "loss": 0.8393,
+      "step": 791
+    },
+    {
+      "epoch": 0.46337806706402895,
+      "grad_norm": 0.10218168425909542,
+      "learning_rate": 7.087537902509607e-06,
+      "loss": 0.8016,
+      "step": 792
+    },
+    {
+      "epoch": 0.463963140381029,
+      "grad_norm": 0.03989943757222258,
+      "learning_rate": 7.085335232878865e-06,
+      "loss": 0.7431,
+      "step": 793
+    },
+    {
+      "epoch": 0.464548213698029,
+      "grad_norm": 0.03900627784304605,
+      "learning_rate": 7.083130294668076e-06,
+      "loss": 0.8431,
+      "step": 794
+    },
+    {
+      "epoch": 0.4651332870150291,
+      "grad_norm": 0.05897148270799823,
+      "learning_rate": 7.080923089739978e-06,
+      "loss": 0.83,
+      "step": 795
+    },
+    {
+      "epoch": 0.4657183603320291,
+      "grad_norm": 0.03888063452127548,
+      "learning_rate": 7.078713619959228e-06,
+      "loss": 0.7915,
+      "step": 796
+    },
+    {
+      "epoch": 0.46630343364902915,
+      "grad_norm": 0.04604225990759911,
+      "learning_rate": 7.076501887192387e-06,
+      "loss": 0.8419,
+      "step": 797
+    },
+    {
+      "epoch": 0.4668885069660292,
+      "grad_norm": 0.040295564919243125,
+      "learning_rate": 7.074287893307941e-06,
+      "loss": 0.8764,
+      "step": 798
+    },
+    {
+      "epoch": 0.4674735802830292,
+      "grad_norm": 0.03769433658091701,
+      "learning_rate": 7.072071640176274e-06,
+      "loss": 0.8566,
+      "step": 799
+    },
+    {
+      "epoch": 0.46805865360002924,
+      "grad_norm": 0.03728814454795478,
+      "learning_rate": 7.069853129669688e-06,
+      "loss": 0.771,
+      "step": 800
+    },
+    {
+      "epoch": 0.46864372691702927,
+      "grad_norm": 0.03547834831253728,
+      "learning_rate": 7.067632363662386e-06,
+      "loss": 0.7874,
+      "step": 801
+    },
+    {
+      "epoch": 0.46922880023402935,
+      "grad_norm": 0.04116629646593246,
+      "learning_rate": 7.065409344030479e-06,
+      "loss": 0.8579,
+      "step": 802
+    },
+    {
+      "epoch": 0.4698138735510294,
+      "grad_norm": 0.04284181743182457,
+      "learning_rate": 7.063184072651981e-06,
+      "loss": 0.874,
+      "step": 803
+    },
+    {
+      "epoch": 0.4703989468680294,
+      "grad_norm": 0.5453076047222728,
+      "learning_rate": 7.060956551406807e-06,
+      "loss": 0.8275,
+      "step": 804
+    },
+    {
+      "epoch": 0.47098402018502944,
+      "grad_norm": 0.04285257240888046,
+      "learning_rate": 7.058726782176778e-06,
+      "loss": 0.8748,
+      "step": 805
+    },
+    {
+      "epoch": 0.47156909350202947,
+      "grad_norm": 0.0426593418584649,
+      "learning_rate": 7.056494766845606e-06,
+      "loss": 0.805,
+      "step": 806
+    },
+    {
+      "epoch": 0.4721541668190295,
+      "grad_norm": 0.03749564999202563,
+      "learning_rate": 7.05426050729891e-06,
+      "loss": 0.8289,
+      "step": 807
+    },
+    {
+      "epoch": 0.4727392401360295,
+      "grad_norm": 0.044556182570763894,
+      "learning_rate": 7.052024005424194e-06,
+      "loss": 0.8757,
+      "step": 808
+    },
+    {
+      "epoch": 0.4733243134530296,
+      "grad_norm": 0.04281051498943111,
+      "learning_rate": 7.049785263110867e-06,
+      "loss": 0.8838,
+      "step": 809
+    },
+    {
+      "epoch": 0.47390938677002964,
+      "grad_norm": 0.04188599336409731,
+      "learning_rate": 7.047544282250223e-06,
+      "loss": 0.8371,
+      "step": 810
+    },
+    {
+      "epoch": 0.47449446008702967,
+      "grad_norm": 0.04079367612711138,
+      "learning_rate": 7.045301064735451e-06,
+      "loss": 0.8383,
+      "step": 811
+    },
+    {
+      "epoch": 0.4750795334040297,
+      "grad_norm": 0.037411483434431944,
+      "learning_rate": 7.0430556124616294e-06,
+      "loss": 0.7866,
+      "step": 812
+    },
+    {
+      "epoch": 0.4756646067210297,
+      "grad_norm": 0.0497395887833391,
+      "learning_rate": 7.040807927325723e-06,
+      "loss": 0.836,
+      "step": 813
+    },
+    {
+      "epoch": 0.47624968003802975,
+      "grad_norm": 0.039340018189295584,
+      "learning_rate": 7.038558011226583e-06,
+      "loss": 0.7925,
+      "step": 814
+    },
+    {
+      "epoch": 0.4768347533550298,
+      "grad_norm": 0.04067266509967937,
+      "learning_rate": 7.036305866064947e-06,
+      "loss": 0.8246,
+      "step": 815
+    },
+    {
+      "epoch": 0.4774198266720298,
+      "grad_norm": 0.03756080114515082,
+      "learning_rate": 7.0340514937434334e-06,
+      "loss": 0.8091,
+      "step": 816
+    },
+    {
+      "epoch": 0.4780048999890299,
+      "grad_norm": 0.04009793142877897,
+      "learning_rate": 7.031794896166544e-06,
+      "loss": 0.8367,
+      "step": 817
+    },
+    {
+      "epoch": 0.4785899733060299,
+      "grad_norm": 0.045719600380512315,
+      "learning_rate": 7.029536075240659e-06,
+      "loss": 0.8698,
+      "step": 818
+    },
+    {
+      "epoch": 0.47917504662302995,
+      "grad_norm": 0.06146708297478866,
+      "learning_rate": 7.0272750328740394e-06,
+      "loss": 0.7769,
+      "step": 819
+    },
+    {
+      "epoch": 0.47976011994003,
+      "grad_norm": 0.0379549936030894,
+      "learning_rate": 7.025011770976821e-06,
+      "loss": 0.8307,
+      "step": 820
+    },
+    {
+      "epoch": 0.48034519325703,
+      "grad_norm": 0.046629898254568806,
+      "learning_rate": 7.022746291461013e-06,
+      "loss": 0.8296,
+      "step": 821
+    },
+    {
+      "epoch": 0.48093026657403004,
+      "grad_norm": 0.04903726722961758,
+      "learning_rate": 7.020478596240503e-06,
+      "loss": 0.8578,
+      "step": 822
+    },
+    {
+      "epoch": 0.48151533989103007,
+      "grad_norm": 0.044482172504589565,
+      "learning_rate": 7.018208687231045e-06,
+      "loss": 0.8339,
+      "step": 823
+    },
+    {
+      "epoch": 0.48210041320803015,
+      "grad_norm": 0.04652606756149258,
+      "learning_rate": 7.015936566350267e-06,
+      "loss": 0.8629,
+      "step": 824
+    },
+    {
+      "epoch": 0.4826854865250302,
+      "grad_norm": 0.03952788329765694,
+      "learning_rate": 7.013662235517661e-06,
+      "loss": 0.8851,
+      "step": 825
+    },
+    {
+      "epoch": 0.4832705598420302,
+      "grad_norm": 0.04737073432575187,
+      "learning_rate": 7.011385696654594e-06,
+      "loss": 0.8662,
+      "step": 826
+    },
+    {
+      "epoch": 0.48385563315903024,
+      "grad_norm": 0.03751881453871839,
+      "learning_rate": 7.0091069516842915e-06,
+      "loss": 0.8559,
+      "step": 827
+    },
+    {
+      "epoch": 0.48444070647603027,
+      "grad_norm": 0.03716116535695275,
+      "learning_rate": 7.006826002531843e-06,
+      "loss": 0.7718,
+      "step": 828
+    },
+    {
+      "epoch": 0.4850257797930303,
+      "grad_norm": 0.056266031939443406,
+      "learning_rate": 7.004542851124203e-06,
+      "loss": 0.7714,
+      "step": 829
+    },
+    {
+      "epoch": 0.4856108531100303,
+      "grad_norm": 0.05535556673447491,
+      "learning_rate": 7.0022574993901865e-06,
+      "loss": 0.8676,
+      "step": 830
+    },
+    {
+      "epoch": 0.4861959264270304,
+      "grad_norm": 0.03845486724104416,
+      "learning_rate": 6.999969949260464e-06,
+      "loss": 0.7885,
+      "step": 831
+    },
+    {
+      "epoch": 0.48678099974403044,
+      "grad_norm": 0.04062056962499875,
+      "learning_rate": 6.99768020266757e-06,
+      "loss": 0.8322,
+      "step": 832
+    },
+    {
+      "epoch": 0.48736607306103047,
+      "grad_norm": 0.06424079693979534,
+      "learning_rate": 6.995388261545884e-06,
+      "loss": 0.905,
+      "step": 833
+    },
+    {
+      "epoch": 0.4879511463780305,
+      "grad_norm": 0.04236693639090112,
+      "learning_rate": 6.993094127831649e-06,
+      "loss": 0.8726,
+      "step": 834
+    },
+    {
+      "epoch": 0.4885362196950305,
+      "grad_norm": 0.038818701632739046,
+      "learning_rate": 6.990797803462955e-06,
+      "loss": 0.7753,
+      "step": 835
+    },
+    {
+      "epoch": 0.48912129301203056,
+      "grad_norm": 0.05513690717963214,
+      "learning_rate": 6.988499290379746e-06,
+      "loss": 0.888,
+      "step": 836
+    },
+    {
+      "epoch": 0.4897063663290306,
+      "grad_norm": 0.040849030445044794,
+      "learning_rate": 6.986198590523812e-06,
+      "loss": 0.8466,
+      "step": 837
+    },
+    {
+      "epoch": 0.49029143964603067,
+      "grad_norm": 0.040919168417802985,
+      "learning_rate": 6.983895705838793e-06,
+      "loss": 0.8589,
+      "step": 838
+    },
+    {
+      "epoch": 0.4908765129630307,
+      "grad_norm": 0.038496795188997325,
+      "learning_rate": 6.9815906382701725e-06,
+      "loss": 0.862,
+      "step": 839
+    },
+    {
+      "epoch": 0.4914615862800307,
+      "grad_norm": 0.043613843698726576,
+      "learning_rate": 6.97928338976528e-06,
+      "loss": 0.8702,
+      "step": 840
+    },
+    {
+      "epoch": 0.49204665959703076,
+      "grad_norm": 0.03941143092937186,
+      "learning_rate": 6.9769739622732855e-06,
+      "loss": 0.7831,
+      "step": 841
+    },
+    {
+      "epoch": 0.4926317329140308,
+      "grad_norm": 0.06878143755782574,
+      "learning_rate": 6.974662357745203e-06,
+      "loss": 0.9225,
+      "step": 842
+    },
+    {
+      "epoch": 0.4932168062310308,
+      "grad_norm": 0.05850718491585764,
+      "learning_rate": 6.972348578133881e-06,
+      "loss": 0.8781,
+      "step": 843
+    },
+    {
+      "epoch": 0.49380187954803084,
+      "grad_norm": 0.039118395131995005,
+      "learning_rate": 6.9700326253940095e-06,
+      "loss": 0.7985,
+      "step": 844
+    },
+    {
+      "epoch": 0.4943869528650309,
+      "grad_norm": 0.057835827757031166,
+      "learning_rate": 6.967714501482114e-06,
+      "loss": 0.781,
+      "step": 845
+    },
+    {
+      "epoch": 0.49497202618203096,
+      "grad_norm": 0.05588852958063585,
+      "learning_rate": 6.965394208356551e-06,
+      "loss": 0.8423,
+      "step": 846
+    },
+    {
+      "epoch": 0.495557099499031,
+      "grad_norm": 0.039374530415438855,
+      "learning_rate": 6.9630717479775145e-06,
+      "loss": 0.8456,
+      "step": 847
+    },
+    {
+      "epoch": 0.496142172816031,
+      "grad_norm": 0.0358193083222334,
+      "learning_rate": 6.960747122307025e-06,
+      "loss": 0.7992,
+      "step": 848
+    },
+    {
+      "epoch": 0.49672724613303104,
+      "grad_norm": 0.03788661691723569,
+      "learning_rate": 6.9584203333089325e-06,
+      "loss": 0.8037,
+      "step": 849
+    },
+    {
+      "epoch": 0.49731231945003107,
+      "grad_norm": 0.04201649774003793,
+      "learning_rate": 6.956091382948918e-06,
+      "loss": 0.8882,
+      "step": 850
+    },
+    {
+      "epoch": 0.4978973927670311,
+      "grad_norm": 0.04434150316820075,
+      "learning_rate": 6.953760273194487e-06,
+      "loss": 0.8166,
+      "step": 851
+    },
+    {
+      "epoch": 0.49848246608403113,
+      "grad_norm": 0.07426388002691353,
+      "learning_rate": 6.951427006014967e-06,
+      "loss": 0.8424,
+      "step": 852
+    },
+    {
+      "epoch": 0.4990675394010312,
+      "grad_norm": 0.037544290713932796,
+      "learning_rate": 6.949091583381511e-06,
+      "loss": 0.7957,
+      "step": 853
+    },
+    {
+      "epoch": 0.49965261271803124,
+      "grad_norm": 0.05489041840699949,
+      "learning_rate": 6.946754007267091e-06,
+      "loss": 0.9257,
+      "step": 854
+    },
+    {
+      "epoch": 0.5002376860350313,
+      "grad_norm": 0.042595883415080985,
+      "learning_rate": 6.944414279646499e-06,
+      "loss": 0.8277,
+      "step": 855
+    },
+    {
+      "epoch": 0.5008227593520314,
+      "grad_norm": 0.08019312917038988,
+      "learning_rate": 6.942072402496345e-06,
+      "loss": 0.9115,
+      "step": 856
+    },
+    {
+      "epoch": 0.5014078326690313,
+      "grad_norm": 0.05086333614690917,
+      "learning_rate": 6.9397283777950545e-06,
+      "loss": 0.9041,
+      "step": 857
+    },
+    {
+      "epoch": 0.5019929059860314,
+      "grad_norm": 0.04214362484163065,
+      "learning_rate": 6.937382207522867e-06,
+      "loss": 0.8299,
+      "step": 858
+    },
+    {
+      "epoch": 0.5025779793030314,
+      "grad_norm": 0.04083088848742396,
+      "learning_rate": 6.935033893661835e-06,
+      "loss": 0.8356,
+      "step": 859
+    },
+    {
+      "epoch": 0.5031630526200315,
+      "grad_norm": 0.04687863044141472,
+      "learning_rate": 6.932683438195821e-06,
+      "loss": 0.865,
+      "step": 860
+    },
+    {
+      "epoch": 0.5037481259370314,
+      "grad_norm": 0.04280142790393131,
+      "learning_rate": 6.9303308431105e-06,
+      "loss": 0.8862,
+      "step": 861
+    },
+    {
+      "epoch": 0.5043331992540315,
+      "grad_norm": 0.04180226845557804,
+      "learning_rate": 6.92797611039335e-06,
+      "loss": 0.8046,
+      "step": 862
+    },
+    {
+      "epoch": 0.5049182725710316,
+      "grad_norm": 0.035262278854158474,
+      "learning_rate": 6.925619242033656e-06,
+      "loss": 0.8197,
+      "step": 863
+    },
+    {
+      "epoch": 0.5055033458880316,
+      "grad_norm": 0.03754079468139451,
+      "learning_rate": 6.92326024002251e-06,
+      "loss": 0.8101,
+      "step": 864
+    },
+    {
+      "epoch": 0.5060884192050317,
+      "grad_norm": 0.039581024469053656,
+      "learning_rate": 6.9208991063528045e-06,
+      "loss": 0.8607,
+      "step": 865
+    },
+    {
+      "epoch": 0.5066734925220316,
+      "grad_norm": 0.0371798589671245,
+      "learning_rate": 6.918535843019233e-06,
+      "loss": 0.8102,
+      "step": 866
+    },
+    {
+      "epoch": 0.5072585658390317,
+      "grad_norm": 0.037768636271352095,
+      "learning_rate": 6.916170452018288e-06,
+      "loss": 0.8418,
+      "step": 867
+    },
+    {
+      "epoch": 0.5078436391560317,
+      "grad_norm": 0.10093759593186688,
+      "learning_rate": 6.913802935348258e-06,
+      "loss": 0.8629,
+      "step": 868
+    },
+    {
+      "epoch": 0.5084287124730318,
+      "grad_norm": 0.03663052824269648,
+      "learning_rate": 6.911433295009232e-06,
+      "loss": 0.8162,
+      "step": 869
+    },
+    {
+      "epoch": 0.5090137857900318,
+      "grad_norm": 0.04855772810380649,
+      "learning_rate": 6.909061533003088e-06,
+      "loss": 0.8616,
+      "step": 870
+    },
+    {
+      "epoch": 0.5095988591070318,
+      "grad_norm": 0.043219673612599416,
+      "learning_rate": 6.906687651333498e-06,
+      "loss": 0.7216,
+      "step": 871
+    },
+    {
+      "epoch": 0.5101839324240319,
+      "grad_norm": 0.042711861232268646,
+      "learning_rate": 6.904311652005925e-06,
+      "loss": 0.7547,
+      "step": 872
+    },
+    {
+      "epoch": 0.5107690057410319,
+      "grad_norm": 0.05574403716352006,
+      "learning_rate": 6.9019335370276225e-06,
+      "loss": 0.795,
+      "step": 873
+    },
+    {
+      "epoch": 0.511354079058032,
+      "grad_norm": 0.06264056674635722,
+      "learning_rate": 6.899553308407629e-06,
+      "loss": 0.85,
+      "step": 874
+    },
+    {
+      "epoch": 0.511939152375032,
+      "grad_norm": 0.04875354726056701,
+      "learning_rate": 6.89717096815677e-06,
+      "loss": 0.8407,
+      "step": 875
+    },
+    {
+      "epoch": 0.512524225692032,
+      "grad_norm": 0.038870649597744084,
+      "learning_rate": 6.894786518287653e-06,
+      "loss": 0.8076,
+      "step": 876
+    },
+    {
+      "epoch": 0.513109299009032,
+      "grad_norm": 0.05118848362688543,
+      "learning_rate": 6.8923999608146705e-06,
+      "loss": 0.8363,
+      "step": 877
+    },
+    {
+      "epoch": 0.5136943723260321,
+      "grad_norm": 0.0376112846207315,
+      "learning_rate": 6.890011297753994e-06,
+      "loss": 0.7743,
+      "step": 878
+    },
+    {
+      "epoch": 0.5142794456430322,
+      "grad_norm": 0.06896727299920731,
+      "learning_rate": 6.887620531123574e-06,
+      "loss": 0.8359,
+      "step": 879
+    },
+    {
+      "epoch": 0.5148645189600322,
+      "grad_norm": 0.03963337869499397,
+      "learning_rate": 6.885227662943136e-06,
+      "loss": 0.8206,
+      "step": 880
+    },
+    {
+      "epoch": 0.5154495922770322,
+      "grad_norm": 0.04870453554962347,
+      "learning_rate": 6.882832695234186e-06,
+      "loss": 0.8189,
+      "step": 881
+    },
+    {
+      "epoch": 0.5160346655940322,
+      "grad_norm": 0.03835351064840454,
+      "learning_rate": 6.880435630019998e-06,
+      "loss": 0.8973,
+      "step": 882
+    },
+    {
+      "epoch": 0.5166197389110323,
+      "grad_norm": 0.058777196577975345,
+      "learning_rate": 6.8780364693256224e-06,
+      "loss": 0.8733,
+      "step": 883
+    },
+    {
+      "epoch": 0.5172048122280323,
+      "grad_norm": 0.04023477251259275,
+      "learning_rate": 6.875635215177878e-06,
+      "loss": 0.7674,
+      "step": 884
+    },
+    {
+      "epoch": 0.5177898855450324,
+      "grad_norm": 0.10180465827282513,
+      "learning_rate": 6.873231869605351e-06,
+      "loss": 0.8238,
+      "step": 885
+    },
+    {
+      "epoch": 0.5183749588620324,
+      "grad_norm": 0.03697008145978301,
+      "learning_rate": 6.870826434638396e-06,
+      "loss": 0.8113,
+      "step": 886
+    },
+    {
+      "epoch": 0.5189600321790324,
+      "grad_norm": 0.03765250989596081,
+      "learning_rate": 6.868418912309133e-06,
+      "loss": 0.8409,
+      "step": 887
+    },
+    {
+      "epoch": 0.5195451054960325,
+      "grad_norm": 0.03863967736287724,
+      "learning_rate": 6.866009304651444e-06,
+      "loss": 0.7622,
+      "step": 888
+    },
+    {
+      "epoch": 0.5201301788130325,
+      "grad_norm": 0.04809769182496589,
+      "learning_rate": 6.8635976137009735e-06,
+      "loss": 0.838,
+      "step": 889
+    },
+    {
+      "epoch": 0.5207152521300326,
+      "grad_norm": 0.04589536972341525,
+      "learning_rate": 6.861183841495127e-06,
+      "loss": 0.9291,
+      "step": 890
+    },
+    {
+      "epoch": 0.5213003254470325,
+      "grad_norm": 0.04495994452664241,
+      "learning_rate": 6.858767990073066e-06,
+      "loss": 0.9015,
+      "step": 891
+    },
+    {
+      "epoch": 0.5218853987640326,
+      "grad_norm": 0.03736417738835852,
+      "learning_rate": 6.856350061475712e-06,
+      "loss": 0.7575,
+      "step": 892
+    },
+    {
+      "epoch": 0.5224704720810327,
+      "grad_norm": 0.039650470820273634,
+      "learning_rate": 6.853930057745735e-06,
+      "loss": 0.7939,
+      "step": 893
+    },
+    {
+      "epoch": 0.5230555453980327,
+      "grad_norm": 0.04072420592259055,
+      "learning_rate": 6.8515079809275656e-06,
+      "loss": 0.8026,
+      "step": 894
+    },
+    {
+      "epoch": 0.5236406187150328,
+      "grad_norm": 0.04022954757166665,
+      "learning_rate": 6.849083833067381e-06,
+      "loss": 0.7891,
+      "step": 895
+    },
+    {
+      "epoch": 0.5242256920320327,
+      "grad_norm": 0.04135409431135059,
+      "learning_rate": 6.846657616213109e-06,
+      "loss": 0.758,
+      "step": 896
+    },
+    {
+      "epoch": 0.5248107653490328,
+      "grad_norm": 0.03960521359539724,
+      "learning_rate": 6.844229332414427e-06,
+      "loss": 0.8038,
+      "step": 897
+    },
+    {
+      "epoch": 0.5253958386660328,
+      "grad_norm": 0.03765896158372491,
+      "learning_rate": 6.841798983722755e-06,
+      "loss": 0.7877,
+      "step": 898
+    },
+    {
+      "epoch": 0.5259809119830329,
+      "grad_norm": 0.058839179066597226,
+      "learning_rate": 6.839366572191262e-06,
+      "loss": 0.7523,
+      "step": 899
+    },
+    {
+      "epoch": 0.526565985300033,
+      "grad_norm": 0.040054239033310936,
+      "learning_rate": 6.836932099874856e-06,
+      "loss": 0.8321,
+      "step": 900
+    },
+    {
+      "epoch": 0.5271510586170329,
+      "grad_norm": 0.0627577307413516,
+      "learning_rate": 6.834495568830187e-06,
+      "loss": 0.8389,
+      "step": 901
+    },
+    {
+      "epoch": 0.527736131934033,
+      "grad_norm": 0.036542541933322875,
+      "learning_rate": 6.832056981115644e-06,
+      "loss": 0.8108,
+      "step": 902
+    },
+    {
+      "epoch": 0.528321205251033,
+      "grad_norm": 0.039969851332150676,
+      "learning_rate": 6.8296163387913545e-06,
+      "loss": 0.8144,
+      "step": 903
+    },
+    {
+      "epoch": 0.5289062785680331,
+      "grad_norm": 0.04091122066442238,
+      "learning_rate": 6.827173643919181e-06,
+      "loss": 0.8865,
+      "step": 904
+    },
+    {
+      "epoch": 0.529491351885033,
+      "grad_norm": 0.04077179519179353,
+      "learning_rate": 6.824728898562721e-06,
+      "loss": 0.8353,
+      "step": 905
+    },
+    {
+      "epoch": 0.5300764252020331,
+      "grad_norm": 0.045472282387591885,
+      "learning_rate": 6.822282104787305e-06,
+      "loss": 0.8175,
+      "step": 906
+    },
+    {
+      "epoch": 0.5306614985190332,
+      "grad_norm": 0.03919660004429457,
+      "learning_rate": 6.819833264659988e-06,
+      "loss": 0.7968,
+      "step": 907
+    },
+    {
+      "epoch": 0.5312465718360332,
+      "grad_norm": 0.0550475630156648,
+      "learning_rate": 6.81738238024956e-06,
+      "loss": 0.8005,
+      "step": 908
+    },
+    {
+      "epoch": 0.5318316451530333,
+      "grad_norm": 0.04044516211789697,
+      "learning_rate": 6.814929453626538e-06,
+      "loss": 0.8056,
+      "step": 909
+    },
+    {
+      "epoch": 0.5324167184700332,
+      "grad_norm": 0.04512920702041598,
+      "learning_rate": 6.81247448686316e-06,
+      "loss": 0.8061,
+      "step": 910
+    },
+    {
+      "epoch": 0.5330017917870333,
+      "grad_norm": 0.06440782132528662,
+      "learning_rate": 6.810017482033392e-06,
+      "loss": 0.8471,
+      "step": 911
+    },
+    {
+      "epoch": 0.5335868651040333,
+      "grad_norm": 0.03695317579916004,
+      "learning_rate": 6.8075584412129205e-06,
+      "loss": 0.8222,
+      "step": 912
+    },
+    {
+      "epoch": 0.5341719384210334,
+      "grad_norm": 0.03666660940321042,
+      "learning_rate": 6.805097366479148e-06,
+      "loss": 0.7822,
+      "step": 913
+    },
+    {
+      "epoch": 0.5347570117380335,
+      "grad_norm": 0.055445109178676184,
+      "learning_rate": 6.802634259911201e-06,
+      "loss": 0.844,
+      "step": 914
+    },
+    {
+      "epoch": 0.5353420850550334,
+      "grad_norm": 0.05279543181613761,
+      "learning_rate": 6.800169123589919e-06,
+      "loss": 0.7463,
+      "step": 915
+    },
+    {
+      "epoch": 0.5359271583720335,
+      "grad_norm": 0.041859624286024466,
+      "learning_rate": 6.797701959597859e-06,
+      "loss": 0.8604,
+      "step": 916
+    },
+    {
+      "epoch": 0.5365122316890335,
+      "grad_norm": 0.0383422427552055,
+      "learning_rate": 6.795232770019286e-06,
+      "loss": 0.7703,
+      "step": 917
+    },
+    {
+      "epoch": 0.5370973050060336,
+      "grad_norm": 0.043028959290976825,
+      "learning_rate": 6.7927615569401815e-06,
+      "loss": 0.8212,
+      "step": 918
+    },
+    {
+      "epoch": 0.5376823783230336,
+      "grad_norm": 0.03804962859282499,
+      "learning_rate": 6.790288322448235e-06,
+      "loss": 0.8366,
+      "step": 919
+    },
+    {
+      "epoch": 0.5382674516400336,
+      "grad_norm": 0.039087833943377275,
+      "learning_rate": 6.787813068632843e-06,
+      "loss": 0.7831,
+      "step": 920
+    },
+    {
+      "epoch": 0.5388525249570337,
+      "grad_norm": 0.04141823524439823,
+      "learning_rate": 6.785335797585107e-06,
+      "loss": 0.8828,
+      "step": 921
+    },
+    {
+      "epoch": 0.5394375982740337,
+      "grad_norm": 0.05007681264530146,
+      "learning_rate": 6.782856511397835e-06,
+      "loss": 0.8005,
+      "step": 922
+    },
+    {
+      "epoch": 0.5400226715910338,
+      "grad_norm": 0.0358413269090128,
+      "learning_rate": 6.780375212165535e-06,
+      "loss": 0.8488,
+      "step": 923
+    },
+    {
+      "epoch": 0.5406077449080338,
+      "grad_norm": 0.039662947137159635,
+      "learning_rate": 6.777891901984417e-06,
+      "loss": 0.8269,
+      "step": 924
+    },
+    {
+      "epoch": 0.5411928182250338,
+      "grad_norm": 0.03708918191821332,
+      "learning_rate": 6.775406582952389e-06,
+      "loss": 0.8561,
+      "step": 925
+    },
+    {
+      "epoch": 0.5417778915420338,
+      "grad_norm": 0.0512337126087228,
+      "learning_rate": 6.772919257169059e-06,
+      "loss": 0.8225,
+      "step": 926
+    },
+    {
+      "epoch": 0.5423629648590339,
+      "grad_norm": 0.15259566476683822,
+      "learning_rate": 6.770429926735727e-06,
+      "loss": 0.7892,
+      "step": 927
+    },
+    {
+      "epoch": 0.542948038176034,
+      "grad_norm": 0.036454692983525445,
+      "learning_rate": 6.767938593755386e-06,
+      "loss": 0.87,
+      "step": 928
+    },
+    {
+      "epoch": 0.543533111493034,
+      "grad_norm": 0.08503045559657693,
+      "learning_rate": 6.765445260332723e-06,
+      "loss": 0.878,
+      "step": 929
+    },
+    {
+      "epoch": 0.544118184810034,
+      "grad_norm": 0.04291403088850566,
+      "learning_rate": 6.7629499285741155e-06,
+      "loss": 0.8633,
+      "step": 930
+    },
+    {
+      "epoch": 0.544703258127034,
+      "grad_norm": 0.04417233325326183,
+      "learning_rate": 6.7604526005876265e-06,
+      "loss": 0.7777,
+      "step": 931
+    },
+    {
+      "epoch": 0.5452883314440341,
+      "grad_norm": 0.04685514307801011,
+      "learning_rate": 6.7579532784830075e-06,
+      "loss": 0.8233,
+      "step": 932
+    },
+    {
+      "epoch": 0.5458734047610341,
+      "grad_norm": 0.04216821721923777,
+      "learning_rate": 6.755451964371696e-06,
+      "loss": 0.8055,
+      "step": 933
+    },
+    {
+      "epoch": 0.5464584780780342,
+      "grad_norm": 0.05899439548668003,
+      "learning_rate": 6.752948660366807e-06,
+      "loss": 0.7423,
+      "step": 934
+    },
+    {
+      "epoch": 0.5470435513950342,
+      "grad_norm": 0.039547439339394544,
+      "learning_rate": 6.750443368583141e-06,
+      "loss": 0.7959,
+      "step": 935
+    },
+    {
+      "epoch": 0.5476286247120342,
+      "grad_norm": 0.046082967838638036,
+      "learning_rate": 6.747936091137179e-06,
+      "loss": 0.7691,
+      "step": 936
+    },
+    {
+      "epoch": 0.5482136980290343,
+      "grad_norm": 0.04724571857592231,
+      "learning_rate": 6.745426830147074e-06,
+      "loss": 0.7716,
+      "step": 937
+    },
+    {
+      "epoch": 0.5487987713460343,
+      "grad_norm": 0.049126540248396626,
+      "learning_rate": 6.74291558773266e-06,
+      "loss": 0.8051,
+      "step": 938
+    },
+    {
+      "epoch": 0.5493838446630344,
+      "grad_norm": 0.038162530207834874,
+      "learning_rate": 6.740402366015442e-06,
+      "loss": 0.8182,
+      "step": 939
+    },
+    {
+      "epoch": 0.5499689179800343,
+      "grad_norm": 0.07362395009585133,
+      "learning_rate": 6.737887167118597e-06,
+      "loss": 0.8025,
+      "step": 940
+    },
+    {
+      "epoch": 0.5505539912970344,
+      "grad_norm": 0.03855880618683095,
+      "learning_rate": 6.735369993166977e-06,
+      "loss": 0.8257,
+      "step": 941
+    },
+    {
+      "epoch": 0.5511390646140344,
+      "grad_norm": 0.04403565508981498,
+      "learning_rate": 6.732850846287096e-06,
+      "loss": 0.7377,
+      "step": 942
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.039173991395047666,
+      "learning_rate": 6.730329728607137e-06,
+      "loss": 0.8063,
+      "step": 943
+    },
+    {
+      "epoch": 0.5523092112480346,
+      "grad_norm": 0.042531239336748856,
+      "learning_rate": 6.72780664225695e-06,
+      "loss": 0.8237,
+      "step": 944
+    },
+    {
+      "epoch": 0.5528942845650345,
+      "grad_norm": 0.039715480086954504,
+      "learning_rate": 6.725281589368046e-06,
+      "loss": 0.8229,
+      "step": 945
+    },
+    {
+      "epoch": 0.5534793578820346,
+      "grad_norm": 0.04910369460983757,
+      "learning_rate": 6.722754572073599e-06,
+      "loss": 0.8503,
+      "step": 946
+    },
+    {
+      "epoch": 0.5540644311990346,
+      "grad_norm": 0.03827200257320983,
+      "learning_rate": 6.720225592508439e-06,
+      "loss": 0.9069,
+      "step": 947
+    },
+    {
+      "epoch": 0.5546495045160347,
+      "grad_norm": 0.06511818499208381,
+      "learning_rate": 6.7176946528090585e-06,
+      "loss": 0.9065,
+      "step": 948
+    },
+    {
+      "epoch": 0.5552345778330346,
+      "grad_norm": 0.03694464014581456,
+      "learning_rate": 6.715161755113604e-06,
+      "loss": 0.8588,
+      "step": 949
+    },
+    {
+      "epoch": 0.5558196511500347,
+      "grad_norm": 0.03684150463907197,
+      "learning_rate": 6.712626901561876e-06,
+      "loss": 0.8272,
+      "step": 950
+    },
+    {
+      "epoch": 0.5564047244670348,
+      "grad_norm": 0.04096608730346328,
+      "learning_rate": 6.710090094295323e-06,
+      "loss": 0.9031,
+      "step": 951
+    },
+    {
+      "epoch": 0.5569897977840348,
+      "grad_norm": 0.04008860398814125,
+      "learning_rate": 6.707551335457054e-06,
+      "loss": 0.8452,
+      "step": 952
+    },
+    {
+      "epoch": 0.5575748711010349,
+      "grad_norm": 0.06233285275932101,
+      "learning_rate": 6.705010627191816e-06,
+      "loss": 0.8813,
+      "step": 953
+    },
+    {
+      "epoch": 0.5581599444180348,
+      "grad_norm": 0.04077134247799943,
+      "learning_rate": 6.7024679716460114e-06,
+      "loss": 0.8493,
+      "step": 954
+    },
+    {
+      "epoch": 0.5587450177350349,
+      "grad_norm": 0.046352641350363474,
+      "learning_rate": 6.699923370967682e-06,
+      "loss": 0.8309,
+      "step": 955
+    },
+    {
+      "epoch": 0.5593300910520349,
+      "grad_norm": 0.041407416461248854,
+      "learning_rate": 6.6973768273065145e-06,
+      "loss": 0.8149,
+      "step": 956
+    },
+    {
+      "epoch": 0.559915164369035,
+      "grad_norm": 0.03903915190219945,
+      "learning_rate": 6.694828342813839e-06,
+      "loss": 0.8669,
+      "step": 957
+    },
+    {
+      "epoch": 0.5605002376860351,
+      "grad_norm": 0.05659630453049191,
+      "learning_rate": 6.692277919642623e-06,
+      "loss": 0.8291,
+      "step": 958
+    },
+    {
+      "epoch": 0.561085311003035,
+      "grad_norm": 0.048419920107990906,
+      "learning_rate": 6.6897255599474705e-06,
+      "loss": 0.7891,
+      "step": 959
+    },
+    {
+      "epoch": 0.5616703843200351,
+      "grad_norm": 0.04526622141241147,
+      "learning_rate": 6.6871712658846255e-06,
+      "loss": 0.82,
+      "step": 960
+    },
+    {
+      "epoch": 0.5622554576370351,
+      "grad_norm": 0.05616992844910552,
+      "learning_rate": 6.684615039611963e-06,
+      "loss": 0.819,
+      "step": 961
+    },
+    {
+      "epoch": 0.5628405309540352,
+      "grad_norm": 0.05533671965109322,
+      "learning_rate": 6.682056883288993e-06,
+      "loss": 0.8278,
+      "step": 962
+    },
+    {
+      "epoch": 0.5634256042710352,
+      "grad_norm": 0.04723246972770138,
+      "learning_rate": 6.679496799076853e-06,
+      "loss": 0.8255,
+      "step": 963
+    },
+    {
+      "epoch": 0.5640106775880352,
+      "grad_norm": 0.04762545006180037,
+      "learning_rate": 6.67693478913831e-06,
+      "loss": 0.8867,
+      "step": 964
+    },
+    {
+      "epoch": 0.5645957509050353,
+      "grad_norm": 0.04368900157724721,
+      "learning_rate": 6.674370855637759e-06,
+      "loss": 0.8527,
+      "step": 965
+    },
+    {
+      "epoch": 0.5651808242220353,
+      "grad_norm": 0.04639910967539116,
+      "learning_rate": 6.671805000741221e-06,
+      "loss": 0.8147,
+      "step": 966
+    },
+    {
+      "epoch": 0.5657658975390354,
+      "grad_norm": 0.04122171566933528,
+      "learning_rate": 6.6692372266163365e-06,
+      "loss": 0.8176,
+      "step": 967
+    },
+    {
+      "epoch": 0.5663509708560354,
+      "grad_norm": 0.04160364438101482,
+      "learning_rate": 6.666667535432371e-06,
+      "loss": 0.8588,
+      "step": 968
+    },
+    {
+      "epoch": 0.5669360441730354,
+      "grad_norm": 0.04172707626334754,
+      "learning_rate": 6.664095929360207e-06,
+      "loss": 0.8315,
+      "step": 969
+    },
+    {
+      "epoch": 0.5675211174900354,
+      "grad_norm": 0.04646561250778188,
+      "learning_rate": 6.661522410572346e-06,
+      "loss": 0.7923,
+      "step": 970
+    },
+    {
+      "epoch": 0.5681061908070355,
+      "grad_norm": 0.04188298952966753,
+      "learning_rate": 6.658946981242906e-06,
+      "loss": 0.7711,
+      "step": 971
+    },
+    {
+      "epoch": 0.5686912641240356,
+      "grad_norm": 0.03908796787695422,
+      "learning_rate": 6.656369643547617e-06,
+      "loss": 0.8856,
+      "step": 972
+    },
+    {
+      "epoch": 0.5692763374410356,
+      "grad_norm": 0.042788844509774104,
+      "learning_rate": 6.653790399663823e-06,
+      "loss": 0.7808,
+      "step": 973
+    },
+    {
+      "epoch": 0.5698614107580356,
+      "grad_norm": 0.05201703823511508,
+      "learning_rate": 6.651209251770478e-06,
+      "loss": 0.8618,
+      "step": 974
+    },
+    {
+      "epoch": 0.5704464840750356,
+      "grad_norm": 0.04585801993937681,
+      "learning_rate": 6.648626202048144e-06,
+      "loss": 0.8373,
+      "step": 975
+    },
+    {
+      "epoch": 0.5710315573920357,
+      "grad_norm": 0.040661578495326846,
+      "learning_rate": 6.646041252678989e-06,
+      "loss": 0.7641,
+      "step": 976
+    },
+    {
+      "epoch": 0.5716166307090357,
+      "grad_norm": 0.035144726376477184,
+      "learning_rate": 6.643454405846788e-06,
+      "loss": 0.7272,
+      "step": 977
+    },
+    {
+      "epoch": 0.5722017040260358,
+      "grad_norm": 0.04251759434748344,
+      "learning_rate": 6.640865663736917e-06,
+      "loss": 0.8063,
+      "step": 978
+    },
+    {
+      "epoch": 0.5727867773430358,
+      "grad_norm": 0.037521500861554015,
+      "learning_rate": 6.638275028536356e-06,
+      "loss": 0.7797,
+      "step": 979
+    },
+    {
+      "epoch": 0.5733718506600358,
+      "grad_norm": 0.041622067012318506,
+      "learning_rate": 6.6356825024336784e-06,
+      "loss": 0.8679,
+      "step": 980
+    },
+    {
+      "epoch": 0.5739569239770359,
+      "grad_norm": 0.04290995619424604,
+      "learning_rate": 6.63308808761906e-06,
+      "loss": 0.8477,
+      "step": 981
+    },
+    {
+      "epoch": 0.5745419972940359,
+      "grad_norm": 0.0429619795472441,
+      "learning_rate": 6.630491786284273e-06,
+      "loss": 0.8145,
+      "step": 982
+    },
+    {
+      "epoch": 0.575127070611036,
+      "grad_norm": 0.03771407315818989,
+      "learning_rate": 6.6278936006226795e-06,
+      "loss": 0.796,
+      "step": 983
+    },
+    {
+      "epoch": 0.5757121439280359,
+      "grad_norm": 0.0720666677058357,
+      "learning_rate": 6.625293532829236e-06,
+      "loss": 0.7856,
+      "step": 984
+    },
+    {
+      "epoch": 0.576297217245036,
+      "grad_norm": 0.05035693118475441,
+      "learning_rate": 6.622691585100488e-06,
+      "loss": 0.8212,
+      "step": 985
+    },
+    {
+      "epoch": 0.5768822905620361,
+      "grad_norm": 0.0661002129965687,
+      "learning_rate": 6.620087759634569e-06,
+      "loss": 0.8248,
+      "step": 986
+    },
+    {
+      "epoch": 0.5774673638790361,
+      "grad_norm": 0.043048564438655845,
+      "learning_rate": 6.617482058631201e-06,
+      "loss": 0.7865,
+      "step": 987
+    },
+    {
+      "epoch": 0.5780524371960362,
+      "grad_norm": 0.04716626817210766,
+      "learning_rate": 6.614874484291688e-06,
+      "loss": 0.7806,
+      "step": 988
+    },
+    {
+      "epoch": 0.5786375105130361,
+      "grad_norm": 0.043537127489980836,
+      "learning_rate": 6.612265038818915e-06,
+      "loss": 0.8248,
+      "step": 989
+    },
+    {
+      "epoch": 0.5792225838300362,
+      "grad_norm": 0.036528767071679374,
+      "learning_rate": 6.609653724417354e-06,
+      "loss": 0.8464,
+      "step": 990
+    },
+    {
+      "epoch": 0.5798076571470362,
+      "grad_norm": 0.041985098940740775,
+      "learning_rate": 6.6070405432930495e-06,
+      "loss": 0.8371,
+      "step": 991
+    },
+    {
+      "epoch": 0.5803927304640363,
+      "grad_norm": 0.037661629039364965,
+      "learning_rate": 6.604425497653627e-06,
+      "loss": 0.8133,
+      "step": 992
+    },
+    {
+      "epoch": 0.5809778037810364,
+      "grad_norm": 0.04709785126617865,
+      "learning_rate": 6.6018085897082845e-06,
+      "loss": 0.7926,
+      "step": 993
+    },
+    {
+      "epoch": 0.5815628770980363,
+      "grad_norm": 0.04920382998315035,
+      "learning_rate": 6.5991898216677945e-06,
+      "loss": 0.8511,
+      "step": 994
+    },
+    {
+      "epoch": 0.5821479504150364,
+      "grad_norm": 0.0384231269239162,
+      "learning_rate": 6.596569195744502e-06,
+      "loss": 0.8767,
+      "step": 995
+    },
+    {
+      "epoch": 0.5827330237320364,
+      "grad_norm": 0.0469334210563224,
+      "learning_rate": 6.59394671415232e-06,
+      "loss": 0.8445,
+      "step": 996
+    },
+    {
+      "epoch": 0.5833180970490365,
+      "grad_norm": 0.07961405813390293,
+      "learning_rate": 6.591322379106728e-06,
+      "loss": 0.7951,
+      "step": 997
+    },
+    {
+      "epoch": 0.5839031703660365,
+      "grad_norm": 0.039593707752364676,
+      "learning_rate": 6.588696192824775e-06,
+      "loss": 0.8325,
+      "step": 998
+    },
+    {
+      "epoch": 0.5844882436830365,
+      "grad_norm": 0.04184294201712494,
+      "learning_rate": 6.5860681575250706e-06,
+      "loss": 0.8599,
+      "step": 999
+    },
+    {
+      "epoch": 0.5850733170000366,
+      "grad_norm": 0.10237775503964221,
+      "learning_rate": 6.5834382754277885e-06,
+      "loss": 0.8159,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5856583903170366,
+      "grad_norm": 0.03881481319706578,
+      "learning_rate": 6.580806548754661e-06,
+      "loss": 0.7969,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5862434636340367,
+      "grad_norm": 0.043457660890875695,
+      "learning_rate": 6.578172979728979e-06,
+      "loss": 0.8012,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5868285369510367,
+      "grad_norm": 0.037934756068100706,
+      "learning_rate": 6.5755375705755924e-06,
+      "loss": 0.8248,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5874136102680367,
+      "grad_norm": 0.12610177824785532,
+      "learning_rate": 6.572900323520901e-06,
+      "loss": 0.7342,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5879986835850367,
+      "grad_norm": 0.03661390969286209,
+      "learning_rate": 6.570261240792861e-06,
+      "loss": 0.7682,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5885837569020368,
+      "grad_norm": 0.06798106192365762,
+      "learning_rate": 6.5676203246209785e-06,
+      "loss": 0.793,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5891688302190369,
+      "grad_norm": 0.05106297642369706,
+      "learning_rate": 6.564977577236309e-06,
+      "loss": 0.881,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5897539035360368,
+      "grad_norm": 0.04384540012407079,
+      "learning_rate": 6.5623330008714505e-06,
+      "loss": 0.7453,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5903389768530369,
+      "grad_norm": 0.10758927483483961,
+      "learning_rate": 6.559686597760555e-06,
+      "loss": 0.8367,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5909240501700369,
+      "grad_norm": 0.04079443571028819,
+      "learning_rate": 6.557038370139307e-06,
+      "loss": 0.8972,
+      "step": 1010
+    },
+    {
+      "epoch": 0.591509123487037,
+      "grad_norm": 0.04314093124710063,
+      "learning_rate": 6.554388320244943e-06,
+      "loss": 0.8316,
+      "step": 1011
+    },
+    {
+      "epoch": 0.592094196804037,
+      "grad_norm": 0.037939172015033606,
+      "learning_rate": 6.5517364503162315e-06,
+      "loss": 0.8587,
+      "step": 1012
+    },
+    {
+      "epoch": 0.592679270121037,
+      "grad_norm": 0.08510073848850286,
+      "learning_rate": 6.549082762593481e-06,
+      "loss": 0.8269,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5932643434380371,
+      "grad_norm": 0.04708547345385738,
+      "learning_rate": 6.546427259318535e-06,
+      "loss": 0.8102,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5938494167550371,
+      "grad_norm": 0.053444431590032856,
+      "learning_rate": 6.543769942734772e-06,
+      "loss": 0.7774,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5944344900720372,
+      "grad_norm": 0.04005959184610443,
+      "learning_rate": 6.541110815087104e-06,
+      "loss": 0.7808,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5950195633890372,
+      "grad_norm": 0.05077795167298267,
+      "learning_rate": 6.538449878621966e-06,
+      "loss": 0.8758,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5956046367060372,
+      "grad_norm": 0.03833355241397989,
+      "learning_rate": 6.535787135587331e-06,
+      "loss": 0.8331,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5961897100230372,
+      "grad_norm": 0.053233926527690635,
+      "learning_rate": 6.533122588232689e-06,
+      "loss": 0.8339,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5967747833400373,
+      "grad_norm": 0.044734695814882965,
+      "learning_rate": 6.530456238809062e-06,
+      "loss": 0.7773,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5973598566570373,
+      "grad_norm": 0.05575094862913427,
+      "learning_rate": 6.527788089568987e-06,
+      "loss": 0.8303,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5979449299740374,
+      "grad_norm": 0.04961953370443646,
+      "learning_rate": 6.525118142766527e-06,
+      "loss": 0.7392,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5985300032910374,
+      "grad_norm": 0.09625684788536548,
+      "learning_rate": 6.522446400657264e-06,
+      "loss": 0.9054,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5991150766080374,
+      "grad_norm": 0.05229897404990849,
+      "learning_rate": 6.519772865498291e-06,
+      "loss": 0.7961,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5997001499250375,
+      "grad_norm": 0.043742861050743945,
+      "learning_rate": 6.51709753954822e-06,
+      "loss": 0.8543,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6002852232420375,
+      "grad_norm": 0.04421257742872319,
+      "learning_rate": 6.514420425067179e-06,
+      "loss": 0.9387,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6008702965590376,
+      "grad_norm": 0.05074707736420169,
+      "learning_rate": 6.511741524316798e-06,
+      "loss": 0.6896,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6014553698760375,
+      "grad_norm": 0.038800133730654246,
+      "learning_rate": 6.509060839560223e-06,
+      "loss": 0.7429,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6020404431930376,
+      "grad_norm": 0.0672245152896617,
+      "learning_rate": 6.506378373062107e-06,
+      "loss": 0.8246,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6026255165100377,
+      "grad_norm": 0.04853749164209948,
+      "learning_rate": 6.503694127088604e-06,
+      "loss": 0.8006,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6032105898270377,
+      "grad_norm": 0.0423510296147158,
+      "learning_rate": 6.501008103907376e-06,
+      "loss": 0.8068,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6037956631440378,
+      "grad_norm": 0.04168402528442862,
+      "learning_rate": 6.498320305787583e-06,
+      "loss": 0.747,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6043807364610377,
+      "grad_norm": 0.05536933839665471,
+      "learning_rate": 6.495630734999885e-06,
+      "loss": 0.8214,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6049658097780378,
+      "grad_norm": 0.03706453349529763,
+      "learning_rate": 6.4929393938164425e-06,
+      "loss": 0.7507,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6055508830950378,
+      "grad_norm": 0.03569665816245634,
+      "learning_rate": 6.490246284510907e-06,
+      "loss": 0.8255,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6061359564120379,
+      "grad_norm": 0.04630173604608656,
+      "learning_rate": 6.487551409358428e-06,
+      "loss": 0.8046,
+      "step": 1036
+    },
+    {
+      "epoch": 0.606721029729038,
+      "grad_norm": 0.04723153392312322,
+      "learning_rate": 6.4848547706356444e-06,
+      "loss": 0.8256,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6073061030460379,
+      "grad_norm": 0.03827183486478635,
+      "learning_rate": 6.482156370620683e-06,
+      "loss": 0.7563,
+      "step": 1038
+    },
+    {
+      "epoch": 0.607891176363038,
+      "grad_norm": 0.0388015622476578,
+      "learning_rate": 6.479456211593165e-06,
+      "loss": 0.7176,
+      "step": 1039
+    },
+    {
+      "epoch": 0.608476249680038,
+      "grad_norm": 0.03999779366433317,
+      "learning_rate": 6.476754295834191e-06,
+      "loss": 0.8224,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6090613229970381,
+      "grad_norm": 0.039892311683218856,
+      "learning_rate": 6.47405062562635e-06,
+      "loss": 0.8236,
+      "step": 1041
+    },
+    {
+      "epoch": 0.609646396314038,
+      "grad_norm": 0.03973227428998665,
+      "learning_rate": 6.471345203253711e-06,
+      "loss": 0.8184,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6102314696310381,
+      "grad_norm": 0.04150106217436074,
+      "learning_rate": 6.468638031001823e-06,
+      "loss": 0.8804,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6108165429480382,
+      "grad_norm": 0.04239329779716534,
+      "learning_rate": 6.465929111157714e-06,
+      "loss": 0.7935,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6114016162650382,
+      "grad_norm": 0.05192716135637802,
+      "learning_rate": 6.463218446009888e-06,
+      "loss": 0.8526,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6119866895820383,
+      "grad_norm": 0.04095824415870296,
+      "learning_rate": 6.4605060378483255e-06,
+      "loss": 0.901,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6125717628990383,
+      "grad_norm": 0.03546640975061806,
+      "learning_rate": 6.457791888964478e-06,
+      "loss": 0.7811,
+      "step": 1047
+    },
+    {
+      "epoch": 0.6131568362160383,
+      "grad_norm": 0.040296634693903335,
+      "learning_rate": 6.455076001651265e-06,
+      "loss": 0.7403,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6137419095330383,
+      "grad_norm": 0.039361319712739894,
+      "learning_rate": 6.452358378203079e-06,
+      "loss": 0.8359,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6143269828500384,
+      "grad_norm": 0.04449717698536887,
+      "learning_rate": 6.449639020915777e-06,
+      "loss": 0.7877,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6149120561670385,
+      "grad_norm": 0.05075898809213348,
+      "learning_rate": 6.446917932086681e-06,
+      "loss": 0.7867,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6154971294840385,
+      "grad_norm": 0.05427508854998823,
+      "learning_rate": 6.444195114014573e-06,
+      "loss": 0.9079,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6160822028010385,
+      "grad_norm": 0.039038596627602784,
+      "learning_rate": 6.441470568999704e-06,
+      "loss": 0.8449,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6166672761180385,
+      "grad_norm": 0.05495139920274083,
+      "learning_rate": 6.438744299343774e-06,
+      "loss": 0.8195,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6172523494350386,
+      "grad_norm": 0.04036928882191201,
+      "learning_rate": 6.436016307349947e-06,
+      "loss": 0.9269,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6178374227520386,
+      "grad_norm": 0.03982823893586514,
+      "learning_rate": 6.4332865953228395e-06,
+      "loss": 0.7358,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6184224960690387,
+      "grad_norm": 0.06481474497909782,
+      "learning_rate": 6.430555165568521e-06,
+      "loss": 0.7267,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6190075693860387,
+      "grad_norm": 0.05875784906821891,
+      "learning_rate": 6.427822020394512e-06,
+      "loss": 0.8439,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6195926427030387,
+      "grad_norm": 0.040087218682744265,
+      "learning_rate": 6.425087162109781e-06,
+      "loss": 0.7822,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6201777160200388,
+      "grad_norm": 0.04225947062502592,
+      "learning_rate": 6.422350593024747e-06,
+      "loss": 0.8365,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6207627893370388,
+      "grad_norm": 0.03505701993198131,
+      "learning_rate": 6.419612315451275e-06,
+      "loss": 0.7465,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6213478626540389,
+      "grad_norm": 0.03617054781781428,
+      "learning_rate": 6.4168723317026655e-06,
+      "loss": 0.7628,
+      "step": 1062
+    },
+    {
+      "epoch": 0.6219329359710388,
+      "grad_norm": 0.04149790973744168,
+      "learning_rate": 6.414130644093669e-06,
+      "loss": 0.7846,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6225180092880389,
+      "grad_norm": 0.08318622079767567,
+      "learning_rate": 6.411387254940473e-06,
+      "loss": 0.8865,
+      "step": 1064
+    },
+    {
+      "epoch": 0.623103082605039,
+      "grad_norm": 0.04061744209355914,
+      "learning_rate": 6.4086421665607e-06,
+      "loss": 0.7673,
+      "step": 1065
+    },
+    {
+      "epoch": 0.623688155922039,
+      "grad_norm": 0.04460594650043864,
+      "learning_rate": 6.405895381273411e-06,
+      "loss": 0.7735,
+      "step": 1066
+    },
+    {
+      "epoch": 0.624273229239039,
+      "grad_norm": 0.05550280064698044,
+      "learning_rate": 6.403146901399098e-06,
+      "loss": 0.8569,
+      "step": 1067
+    },
+    {
+      "epoch": 0.624858302556039,
+      "grad_norm": 0.04163827309254167,
+      "learning_rate": 6.400396729259685e-06,
+      "loss": 0.8951,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6254433758730391,
+      "grad_norm": 0.1509844863070891,
+      "learning_rate": 6.39764486717853e-06,
+      "loss": 0.8082,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6260284491900391,
+      "grad_norm": 0.04492131767454413,
+      "learning_rate": 6.394891317480412e-06,
+      "loss": 0.7518,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6266135225070392,
+      "grad_norm": 0.0471963162082604,
+      "learning_rate": 6.39213608249154e-06,
+      "loss": 0.8127,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6271985958240393,
+      "grad_norm": 0.038627758596647896,
+      "learning_rate": 6.389379164539545e-06,
+      "loss": 0.7781,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6277836691410392,
+      "grad_norm": 0.04265598384978139,
+      "learning_rate": 6.386620565953482e-06,
+      "loss": 0.7698,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6283687424580393,
+      "grad_norm": 0.040866964196317926,
+      "learning_rate": 6.383860289063821e-06,
+      "loss": 0.741,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6289538157750393,
+      "grad_norm": 0.04488242542194653,
+      "learning_rate": 6.3810983362024575e-06,
+      "loss": 0.8618,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6295388890920394,
+      "grad_norm": 0.04181264245748804,
+      "learning_rate": 6.3783347097026935e-06,
+      "loss": 0.8318,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6301239624090393,
+      "grad_norm": 0.057019285327405245,
+      "learning_rate": 6.375569411899253e-06,
+      "loss": 0.822,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6307090357260394,
+      "grad_norm": 0.04075452383662032,
+      "learning_rate": 6.3728024451282675e-06,
+      "loss": 0.7854,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6312941090430395,
+      "grad_norm": 0.04369789370291459,
+      "learning_rate": 6.37003381172728e-06,
+      "loss": 0.7976,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6318791823600395,
+      "grad_norm": 0.05849968444823343,
+      "learning_rate": 6.367263514035242e-06,
+      "loss": 0.8309,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6324642556770396,
+      "grad_norm": 0.07398315736379607,
+      "learning_rate": 6.364491554392508e-06,
+      "loss": 0.8695,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6330493289940395,
+      "grad_norm": 0.14149329637182187,
+      "learning_rate": 6.36171793514084e-06,
+      "loss": 0.9035,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6336344023110396,
+      "grad_norm": 0.09431782156363094,
+      "learning_rate": 6.358942658623402e-06,
+      "loss": 0.7978,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6342194756280396,
+      "grad_norm": 0.11444686544474032,
+      "learning_rate": 6.356165727184753e-06,
+      "loss": 0.8075,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6348045489450397,
+      "grad_norm": 0.07683296171985984,
+      "learning_rate": 6.353387143170856e-06,
+      "loss": 0.7619,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6353896222620398,
+      "grad_norm": 0.03611601798693106,
+      "learning_rate": 6.3506069089290705e-06,
+      "loss": 0.7886,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6359746955790397,
+      "grad_norm": 0.055915745635355715,
+      "learning_rate": 6.3478250268081435e-06,
+      "loss": 0.8511,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6365597688960398,
+      "grad_norm": 0.048663447635497695,
+      "learning_rate": 6.34504149915822e-06,
+      "loss": 0.9001,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6371448422130398,
+      "grad_norm": 0.04226309960545497,
+      "learning_rate": 6.342256328330833e-06,
+      "loss": 0.7945,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6377299155300399,
+      "grad_norm": 0.0425747978871558,
+      "learning_rate": 6.339469516678903e-06,
+      "loss": 0.7945,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6383149888470399,
+      "grad_norm": 0.049540193290765706,
+      "learning_rate": 6.33668106655674e-06,
+      "loss": 0.8424,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6389000621640399,
+      "grad_norm": 0.04361089778932384,
+      "learning_rate": 6.333890980320033e-06,
+      "loss": 0.8775,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6394851354810399,
+      "grad_norm": 0.04952042923998371,
+      "learning_rate": 6.331099260325858e-06,
+      "loss": 0.7921,
+      "step": 1093
+    },
+    {
+      "epoch": 0.64007020879804,
+      "grad_norm": 0.08510241121137069,
+      "learning_rate": 6.32830590893267e-06,
+      "loss": 0.8449,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6406552821150401,
+      "grad_norm": 0.052581634750892337,
+      "learning_rate": 6.325510928500298e-06,
+      "loss": 0.8504,
+      "step": 1095
+    },
+    {
+      "epoch": 0.64124035543204,
+      "grad_norm": 0.06971844661189186,
+      "learning_rate": 6.322714321389955e-06,
+      "loss": 0.8049,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6418254287490401,
+      "grad_norm": 0.0425225993758099,
+      "learning_rate": 6.319916089964221e-06,
+      "loss": 0.7374,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6424105020660401,
+      "grad_norm": 0.04070882005149654,
+      "learning_rate": 6.317116236587052e-06,
+      "loss": 0.747,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6429955753830402,
+      "grad_norm": 0.04748840753374928,
+      "learning_rate": 6.314314763623775e-06,
+      "loss": 0.7728,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6435806487000402,
+      "grad_norm": 0.08783073493232298,
+      "learning_rate": 6.31151167344108e-06,
+      "loss": 0.8031,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6441657220170403,
+      "grad_norm": 0.12757918242470023,
+      "learning_rate": 6.308706968407029e-06,
+      "loss": 0.8275,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6447507953340403,
+      "grad_norm": 0.04265066246559783,
+      "learning_rate": 6.305900650891045e-06,
+      "loss": 0.8143,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6453358686510403,
+      "grad_norm": 0.05532213512913742,
+      "learning_rate": 6.303092723263917e-06,
+      "loss": 0.7623,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6459209419680404,
+      "grad_norm": 0.05575276568946241,
+      "learning_rate": 6.300283187897788e-06,
+      "loss": 0.7578,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6465060152850404,
+      "grad_norm": 0.04004488402472709,
+      "learning_rate": 6.297472047166164e-06,
+      "loss": 0.8287,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6470910886020405,
+      "grad_norm": 0.07040462606702534,
+      "learning_rate": 6.294659303443907e-06,
+      "loss": 0.802,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6476761619190404,
+      "grad_norm": 0.04991605035702622,
+      "learning_rate": 6.291844959107231e-06,
+      "loss": 0.78,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6482612352360405,
+      "grad_norm": 0.04810590450360042,
+      "learning_rate": 6.289029016533705e-06,
+      "loss": 0.7531,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6488463085530406,
+      "grad_norm": 0.04521072775635856,
+      "learning_rate": 6.286211478102243e-06,
+      "loss": 0.773,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6494313818700406,
+      "grad_norm": 0.04141710815776832,
+      "learning_rate": 6.283392346193114e-06,
+      "loss": 0.813,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6500164551870407,
+      "grad_norm": 0.043835473627139195,
+      "learning_rate": 6.280571623187929e-06,
+      "loss": 0.793,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6506015285040406,
+      "grad_norm": 0.05669397811371989,
+      "learning_rate": 6.277749311469643e-06,
+      "loss": 0.7317,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6511866018210407,
+      "grad_norm": 0.08646975004658393,
+      "learning_rate": 6.274925413422558e-06,
+      "loss": 0.8447,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6517716751380407,
+      "grad_norm": 0.039267068048581634,
+      "learning_rate": 6.272099931432308e-06,
+      "loss": 0.8171,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6523567484550408,
+      "grad_norm": 0.04626551882560527,
+      "learning_rate": 6.2692728678858705e-06,
+      "loss": 0.7719,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6529418217720409,
+      "grad_norm": 0.08786584886983898,
+      "learning_rate": 6.26644422517156e-06,
+      "loss": 0.7863,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6535268950890408,
+      "grad_norm": 0.07658703901409003,
+      "learning_rate": 6.26361400567902e-06,
+      "loss": 0.8893,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6541119684060409,
+      "grad_norm": 0.03799609243525315,
+      "learning_rate": 6.2607822117992326e-06,
+      "loss": 0.7575,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6546970417230409,
+      "grad_norm": 0.04335037636508533,
+      "learning_rate": 6.257948845924505e-06,
+      "loss": 0.7564,
+      "step": 1119
+    },
+    {
+      "epoch": 0.655282115040041,
+      "grad_norm": 0.04285395891669155,
+      "learning_rate": 6.2551139104484755e-06,
+      "loss": 0.8482,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6558671883570409,
+      "grad_norm": 0.03461985415221269,
+      "learning_rate": 6.252277407766103e-06,
+      "loss": 0.7411,
+      "step": 1121
+    },
+    {
+      "epoch": 0.656452261674041,
+      "grad_norm": 0.06081410937727396,
+      "learning_rate": 6.249439340273679e-06,
+      "loss": 0.8753,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6570373349910411,
+      "grad_norm": 0.08982630298329526,
+      "learning_rate": 6.246599710368809e-06,
+      "loss": 0.7514,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6576224083080411,
+      "grad_norm": 0.04564349868186195,
+      "learning_rate": 6.243758520450423e-06,
+      "loss": 0.7989,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6582074816250412,
+      "grad_norm": 0.043054553052619716,
+      "learning_rate": 6.240915772918768e-06,
+      "loss": 0.7454,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6587925549420411,
+      "grad_norm": 0.03677134636684232,
+      "learning_rate": 6.238071470175405e-06,
+      "loss": 0.8528,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6593776282590412,
+      "grad_norm": 0.04172483135818754,
+      "learning_rate": 6.235225614623212e-06,
+      "loss": 0.812,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6599627015760412,
+      "grad_norm": 0.05277877562040755,
+      "learning_rate": 6.232378208666376e-06,
+      "loss": 0.8283,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6605477748930413,
+      "grad_norm": 0.04454598931616706,
+      "learning_rate": 6.229529254710396e-06,
+      "loss": 0.8537,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6611328482100414,
+      "grad_norm": 0.047900592019452154,
+      "learning_rate": 6.226678755162076e-06,
+      "loss": 0.825,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6617179215270413,
+      "grad_norm": 0.05963183612092722,
+      "learning_rate": 6.223826712429529e-06,
+      "loss": 0.8042,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6623029948440414,
+      "grad_norm": 0.046383510242271296,
+      "learning_rate": 6.220973128922168e-06,
+      "loss": 0.865,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6628880681610414,
+      "grad_norm": 0.04417415078522508,
+      "learning_rate": 6.218118007050713e-06,
+      "loss": 0.8235,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6634731414780415,
+      "grad_norm": 0.04207639503924024,
+      "learning_rate": 6.215261349227178e-06,
+      "loss": 0.7858,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6640582147950415,
+      "grad_norm": 0.04649797867374506,
+      "learning_rate": 6.212403157864878e-06,
+      "loss": 0.868,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6646432881120415,
+      "grad_norm": 0.09407948533657494,
+      "learning_rate": 6.209543435378422e-06,
+      "loss": 0.8818,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6652283614290416,
+      "grad_norm": 0.037478109364168094,
+      "learning_rate": 6.206682184183712e-06,
+      "loss": 0.812,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6658134347460416,
+      "grad_norm": 0.05288676232169846,
+      "learning_rate": 6.203819406697945e-06,
+      "loss": 0.7548,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6663985080630417,
+      "grad_norm": 0.04535893610318677,
+      "learning_rate": 6.200955105339603e-06,
+      "loss": 0.8772,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6669835813800417,
+      "grad_norm": 0.04773693536693857,
+      "learning_rate": 6.198089282528456e-06,
+      "loss": 0.7763,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6675686546970417,
+      "grad_norm": 0.0518520434245037,
+      "learning_rate": 6.195221940685563e-06,
+      "loss": 0.7668,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6681537280140417,
+      "grad_norm": 0.05462892968699057,
+      "learning_rate": 6.192353082233263e-06,
+      "loss": 0.7096,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6687388013310418,
+      "grad_norm": 0.04122645056649732,
+      "learning_rate": 6.189482709595177e-06,
+      "loss": 0.7839,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6693238746480419,
+      "grad_norm": 0.058891294040750164,
+      "learning_rate": 6.186610825196204e-06,
+      "loss": 0.7504,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6699089479650419,
+      "grad_norm": 0.05296752593762354,
+      "learning_rate": 6.183737431462524e-06,
+      "loss": 0.7591,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6704940212820419,
+      "grad_norm": 0.04766421657837364,
+      "learning_rate": 6.180862530821588e-06,
+      "loss": 0.742,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6710790945990419,
+      "grad_norm": 0.0915137633830507,
+      "learning_rate": 6.177986125702121e-06,
+      "loss": 0.8167,
+      "step": 1147
+    },
+    {
+      "epoch": 0.671664167916042,
+      "grad_norm": 0.04064619759463224,
+      "learning_rate": 6.17510821853412e-06,
+      "loss": 0.755,
+      "step": 1148
+    },
+    {
+      "epoch": 0.672249241233042,
+      "grad_norm": 0.03962465637676519,
+      "learning_rate": 6.17222881174885e-06,
+      "loss": 0.7952,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6728343145500421,
+      "grad_norm": 0.0475379390885668,
+      "learning_rate": 6.169347907778846e-06,
+      "loss": 0.7889,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6734193878670421,
+      "grad_norm": 0.04260205050833479,
+      "learning_rate": 6.166465509057902e-06,
+      "loss": 0.7094,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6740044611840421,
+      "grad_norm": 0.0452503260334743,
+      "learning_rate": 6.163581618021079e-06,
+      "loss": 0.8137,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6745895345010422,
+      "grad_norm": 0.049045892950316486,
+      "learning_rate": 6.1606962371046975e-06,
+      "loss": 0.7476,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6751746078180422,
+      "grad_norm": 0.1839546472224546,
+      "learning_rate": 6.157809368746337e-06,
+      "loss": 0.7341,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6757596811350423,
+      "grad_norm": 0.04737636667405579,
+      "learning_rate": 6.154921015384833e-06,
+      "loss": 0.8772,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6763447544520422,
+      "grad_norm": 0.039827421277940374,
+      "learning_rate": 6.152031179460276e-06,
+      "loss": 0.8184,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6769298277690423,
+      "grad_norm": 0.03980204662807788,
+      "learning_rate": 6.14913986341401e-06,
+      "loss": 0.8504,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6775149010860424,
+      "grad_norm": 0.06377067252825656,
+      "learning_rate": 6.146247069688627e-06,
+      "loss": 0.8496,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6780999744030424,
+      "grad_norm": 0.5555285996614837,
+      "learning_rate": 6.14335280072797e-06,
+      "loss": 0.8131,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6786850477200425,
+      "grad_norm": 0.03909895393006111,
+      "learning_rate": 6.140457058977125e-06,
+      "loss": 0.7756,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6792701210370424,
+      "grad_norm": 0.03969052355186135,
+      "learning_rate": 6.137559846882426e-06,
+      "loss": 0.8209,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6798551943540425,
+      "grad_norm": 0.04103151559227765,
+      "learning_rate": 6.134661166891445e-06,
+      "loss": 0.8015,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6804402676710425,
+      "grad_norm": 0.05136493470713414,
+      "learning_rate": 6.131761021453e-06,
+      "loss": 0.8234,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6810253409880426,
+      "grad_norm": 0.04189728013078323,
+      "learning_rate": 6.128859413017141e-06,
+      "loss": 0.7964,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6816104143050427,
+      "grad_norm": 0.04337964675631273,
+      "learning_rate": 6.1259563440351564e-06,
+      "loss": 0.8434,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6821954876220426,
+      "grad_norm": 0.09451825900326687,
+      "learning_rate": 6.123051816959569e-06,
+      "loss": 0.8254,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6827805609390427,
+      "grad_norm": 0.045248262569469115,
+      "learning_rate": 6.120145834244133e-06,
+      "loss": 0.7613,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6833656342560427,
+      "grad_norm": 0.03837960459677211,
+      "learning_rate": 6.117238398343831e-06,
+      "loss": 0.7958,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6839507075730428,
+      "grad_norm": 0.05242659218061473,
+      "learning_rate": 6.114329511714876e-06,
+      "loss": 0.8356,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6845357808900427,
+      "grad_norm": 0.06860429447590187,
+      "learning_rate": 6.111419176814704e-06,
+      "loss": 0.802,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6851208542070428,
+      "grad_norm": 0.04508456559483142,
+      "learning_rate": 6.108507396101975e-06,
+      "loss": 0.8444,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6857059275240428,
+      "grad_norm": 0.06393026794280061,
+      "learning_rate": 6.105594172036572e-06,
+      "loss": 0.7585,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6862910008410429,
+      "grad_norm": 0.05435654372858688,
+      "learning_rate": 6.102679507079597e-06,
+      "loss": 0.9379,
+      "step": 1173
+    },
+    {
+      "epoch": 0.686876074158043,
+      "grad_norm": 0.057231278163684694,
+      "learning_rate": 6.099763403693366e-06,
+      "loss": 0.8431,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6874611474750429,
+      "grad_norm": 0.0414940103750233,
+      "learning_rate": 6.096845864341415e-06,
+      "loss": 0.8247,
+      "step": 1175
+    },
+    {
+      "epoch": 0.688046220792043,
+      "grad_norm": 0.03753965623304419,
+      "learning_rate": 6.09392689148849e-06,
+      "loss": 0.7384,
+      "step": 1176
+    },
+    {
+      "epoch": 0.688631294109043,
+      "grad_norm": 0.0445855037853274,
+      "learning_rate": 6.09100648760055e-06,
+      "loss": 0.8913,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6892163674260431,
+      "grad_norm": 0.03616019636034817,
+      "learning_rate": 6.08808465514476e-06,
+      "loss": 0.7741,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6898014407430431,
+      "grad_norm": 0.03958351633767278,
+      "learning_rate": 6.085161396589493e-06,
+      "loss": 0.6991,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6903865140600431,
+      "grad_norm": 0.03958973007993701,
+      "learning_rate": 6.082236714404331e-06,
+      "loss": 0.8455,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6909715873770432,
+      "grad_norm": 0.04165937145938639,
+      "learning_rate": 6.079310611060052e-06,
+      "loss": 0.8072,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6915566606940432,
+      "grad_norm": 0.045951396419511376,
+      "learning_rate": 6.07638308902864e-06,
+      "loss": 0.8313,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6921417340110433,
+      "grad_norm": 0.04752344885301804,
+      "learning_rate": 6.073454150783274e-06,
+      "loss": 0.8802,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6927268073280433,
+      "grad_norm": 0.05117696052721477,
+      "learning_rate": 6.070523798798329e-06,
+      "loss": 0.9472,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6933118806450433,
+      "grad_norm": 0.040955259256951086,
+      "learning_rate": 6.06759203554938e-06,
+      "loss": 0.8126,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6938969539620433,
+      "grad_norm": 0.034916342462867116,
+      "learning_rate": 6.064658863513186e-06,
+      "loss": 0.7091,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6944820272790434,
+      "grad_norm": 0.042140933363091035,
+      "learning_rate": 6.061724285167704e-06,
+      "loss": 0.8323,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6950671005960435,
+      "grad_norm": 0.04413794065710716,
+      "learning_rate": 6.058788302992072e-06,
+      "loss": 0.8419,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.042514536823805356,
+      "learning_rate": 6.055850919466621e-06,
+      "loss": 0.8863,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6962372472300435,
+      "grad_norm": 0.04885967786149231,
+      "learning_rate": 6.05291213707286e-06,
+      "loss": 0.7658,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6968223205470435,
+      "grad_norm": 0.04978541890148103,
+      "learning_rate": 6.0499719582934815e-06,
+      "loss": 0.7496,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6974073938640436,
+      "grad_norm": 0.05231245753746865,
+      "learning_rate": 6.047030385612362e-06,
+      "loss": 0.784,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6979924671810436,
+      "grad_norm": 0.04028385347084005,
+      "learning_rate": 6.0440874215145465e-06,
+      "loss": 0.6913,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6985775404980437,
+      "grad_norm": 0.062297108276536735,
+      "learning_rate": 6.041143068486264e-06,
+      "loss": 0.7337,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6991626138150437,
+      "grad_norm": 0.03928043528877265,
+      "learning_rate": 6.038197329014914e-06,
+      "loss": 0.8415,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6997476871320437,
+      "grad_norm": 0.046537841954999465,
+      "learning_rate": 6.035250205589064e-06,
+      "loss": 0.7873,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7003327604490438,
+      "grad_norm": 0.042222747639775425,
+      "learning_rate": 6.032301700698458e-06,
+      "loss": 0.7629,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7009178337660438,
+      "grad_norm": 0.05930943404435324,
+      "learning_rate": 6.029351816833998e-06,
+      "loss": 0.8273,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7015029070830439,
+      "grad_norm": 0.05874167359504905,
+      "learning_rate": 6.026400556487758e-06,
+      "loss": 0.7159,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7020879804000438,
+      "grad_norm": 0.052410757851244154,
+      "learning_rate": 6.023447922152972e-06,
+      "loss": 0.7887,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7026730537170439,
+      "grad_norm": 0.03785761302792468,
+      "learning_rate": 6.020493916324037e-06,
+      "loss": 0.7942,
+      "step": 1201
+    },
+    {
+      "epoch": 0.703258127034044,
+      "grad_norm": 0.0392431216621593,
+      "learning_rate": 6.017538541496503e-06,
+      "loss": 0.7885,
+      "step": 1202
+    },
+    {
+      "epoch": 0.703843200351044,
+      "grad_norm": 0.08892449717880838,
+      "learning_rate": 6.014581800167085e-06,
+      "loss": 0.8001,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7044282736680441,
+      "grad_norm": 0.06747014549686459,
+      "learning_rate": 6.011623694833644e-06,
+      "loss": 0.727,
+      "step": 1204
+    },
+    {
+      "epoch": 0.705013346985044,
+      "grad_norm": 0.04543757611973289,
+      "learning_rate": 6.008664227995198e-06,
+      "loss": 0.8129,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7055984203020441,
+      "grad_norm": 0.039104300131473785,
+      "learning_rate": 6.005703402151916e-06,
+      "loss": 0.8098,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7061834936190441,
+      "grad_norm": 0.038265510416551914,
+      "learning_rate": 6.0027412198051114e-06,
+      "loss": 0.8279,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7067685669360442,
+      "grad_norm": 0.03588239116909633,
+      "learning_rate": 5.999777683457247e-06,
+      "loss": 0.8096,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7073536402530443,
+      "grad_norm": 0.03971284808692058,
+      "learning_rate": 5.996812795611928e-06,
+      "loss": 0.8331,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7079387135700442,
+      "grad_norm": 0.03963917599976568,
+      "learning_rate": 5.9938465587739e-06,
+      "loss": 0.7908,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7085237868870443,
+      "grad_norm": 0.045273875457725724,
+      "learning_rate": 5.990878975449051e-06,
+      "loss": 0.8111,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7091088602040443,
+      "grad_norm": 0.04337467569823838,
+      "learning_rate": 5.9879100481444055e-06,
+      "loss": 0.8222,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7096939335210444,
+      "grad_norm": 0.039603085337812066,
+      "learning_rate": 5.984939779368122e-06,
+      "loss": 0.7721,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7102790068380443,
+      "grad_norm": 0.04156772105291293,
+      "learning_rate": 5.981968171629494e-06,
+      "loss": 0.8197,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7108640801550444,
+      "grad_norm": 0.04856521644044076,
+      "learning_rate": 5.978995227438944e-06,
+      "loss": 0.8677,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7114491534720445,
+      "grad_norm": 0.03956445076057253,
+      "learning_rate": 5.976020949308027e-06,
+      "loss": 0.7877,
+      "step": 1216
+    },
+    {
+      "epoch": 0.7120342267890445,
+      "grad_norm": 0.04862552856496891,
+      "learning_rate": 5.973045339749422e-06,
+      "loss": 0.8109,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7126193001060446,
+      "grad_norm": 0.045295501112575506,
+      "learning_rate": 5.970068401276935e-06,
+      "loss": 0.791,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7132043734230445,
+      "grad_norm": 0.03916519269752596,
+      "learning_rate": 5.967090136405491e-06,
+      "loss": 0.8131,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7137894467400446,
+      "grad_norm": 0.038383406339616045,
+      "learning_rate": 5.96411054765114e-06,
+      "loss": 0.8102,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7143745200570446,
+      "grad_norm": 0.06824384503876316,
+      "learning_rate": 5.961129637531047e-06,
+      "loss": 0.8104,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7149595933740447,
+      "grad_norm": 0.07316382930858244,
+      "learning_rate": 5.958147408563497e-06,
+      "loss": 0.7862,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7155446666910448,
+      "grad_norm": 0.04737080928686243,
+      "learning_rate": 5.9551638632678835e-06,
+      "loss": 0.8513,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7161297400080447,
+      "grad_norm": 0.0374937969952467,
+      "learning_rate": 5.952179004164718e-06,
+      "loss": 0.8002,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7167148133250448,
+      "grad_norm": 0.04053689173262939,
+      "learning_rate": 5.949192833775618e-06,
+      "loss": 0.7657,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7172998866420448,
+      "grad_norm": 0.0398778756693423,
+      "learning_rate": 5.946205354623312e-06,
+      "loss": 0.7725,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7178849599590449,
+      "grad_norm": 0.04506339056501446,
+      "learning_rate": 5.943216569231629e-06,
+      "loss": 0.7931,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7184700332760449,
+      "grad_norm": 0.04982830890931998,
+      "learning_rate": 5.940226480125508e-06,
+      "loss": 0.8265,
+      "step": 1228
+    },
+    {
+      "epoch": 0.719055106593045,
+      "grad_norm": 0.0397312539469508,
+      "learning_rate": 5.937235089830984e-06,
+      "loss": 0.769,
+      "step": 1229
+    },
+    {
+      "epoch": 0.719640179910045,
+      "grad_norm": 0.03837231212563197,
+      "learning_rate": 5.934242400875195e-06,
+      "loss": 0.8259,
+      "step": 1230
+    },
+    {
+      "epoch": 0.720225253227045,
+      "grad_norm": 0.03612472247568365,
+      "learning_rate": 5.931248415786371e-06,
+      "loss": 0.7637,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7208103265440451,
+      "grad_norm": 0.03537723098243959,
+      "learning_rate": 5.928253137093844e-06,
+      "loss": 0.738,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7213953998610451,
+      "grad_norm": 0.0345658269866166,
+      "learning_rate": 5.925256567328036e-06,
+      "loss": 0.7422,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7219804731780451,
+      "grad_norm": 0.041550218114552985,
+      "learning_rate": 5.922258709020456e-06,
+      "loss": 0.8158,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7225655464950451,
+      "grad_norm": 0.04738915180347003,
+      "learning_rate": 5.919259564703705e-06,
+      "loss": 0.7693,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7231506198120452,
+      "grad_norm": 0.22798373109574022,
+      "learning_rate": 5.916259136911472e-06,
+      "loss": 0.7964,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7237356931290453,
+      "grad_norm": 0.03960961306940275,
+      "learning_rate": 5.913257428178526e-06,
+      "loss": 0.8529,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7243207664460453,
+      "grad_norm": 0.03762982041843176,
+      "learning_rate": 5.910254441040723e-06,
+      "loss": 0.7978,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7249058397630453,
+      "grad_norm": 0.03967585434913458,
+      "learning_rate": 5.907250178034994e-06,
+      "loss": 0.7715,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7254909130800453,
+      "grad_norm": 0.035499686476307656,
+      "learning_rate": 5.904244641699352e-06,
+      "loss": 0.7821,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7260759863970454,
+      "grad_norm": 0.04146813756606358,
+      "learning_rate": 5.9012378345728824e-06,
+      "loss": 0.7832,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7266610597140454,
+      "grad_norm": 0.035835229174829814,
+      "learning_rate": 5.8982297591957465e-06,
+      "loss": 0.7951,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7272461330310455,
+      "grad_norm": 0.040752453651612,
+      "learning_rate": 5.8952204181091775e-06,
+      "loss": 0.7898,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7278312063480454,
+      "grad_norm": 0.04254122433979048,
+      "learning_rate": 5.8922098138554745e-06,
+      "loss": 0.8384,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7284162796650455,
+      "grad_norm": 0.048324438293500575,
+      "learning_rate": 5.889197948978008e-06,
+      "loss": 0.7273,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7290013529820456,
+      "grad_norm": 0.04876106635437417,
+      "learning_rate": 5.886184826021208e-06,
+      "loss": 0.7078,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7295864262990456,
+      "grad_norm": 0.047909628199116906,
+      "learning_rate": 5.883170447530575e-06,
+      "loss": 0.7293,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7301714996160457,
+      "grad_norm": 0.04759781973779755,
+      "learning_rate": 5.880154816052666e-06,
+      "loss": 0.8114,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7307565729330456,
+      "grad_norm": 0.03579815989872979,
+      "learning_rate": 5.8771379341350905e-06,
+      "loss": 0.7843,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7313416462500457,
+      "grad_norm": 0.048181696474966534,
+      "learning_rate": 5.874119804326525e-06,
+      "loss": 0.85,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7319267195670457,
+      "grad_norm": 0.040629567336692485,
+      "learning_rate": 5.871100429176694e-06,
+      "loss": 0.8121,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7325117928840458,
+      "grad_norm": 0.039483264479736525,
+      "learning_rate": 5.8680798112363784e-06,
+      "loss": 0.8365,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7330968662010459,
+      "grad_norm": 0.05089919878146447,
+      "learning_rate": 5.865057953057401e-06,
+      "loss": 0.8076,
+      "step": 1253
+    },
+    {
+      "epoch": 0.7336819395180458,
+      "grad_norm": 0.035397442355980586,
+      "learning_rate": 5.862034857192642e-06,
+      "loss": 0.7798,
+      "step": 1254
+    },
+    {
+      "epoch": 0.7342670128350459,
+      "grad_norm": 0.03687328400855783,
+      "learning_rate": 5.859010526196021e-06,
+      "loss": 0.7859,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7348520861520459,
+      "grad_norm": 0.06502490332457671,
+      "learning_rate": 5.855984962622504e-06,
+      "loss": 0.7927,
+      "step": 1256
+    },
+    {
+      "epoch": 0.735437159469046,
+      "grad_norm": 0.04157588146277028,
+      "learning_rate": 5.852958169028094e-06,
+      "loss": 0.7907,
+      "step": 1257
+    },
+    {
+      "epoch": 0.736022232786046,
+      "grad_norm": 0.04254658384010035,
+      "learning_rate": 5.849930147969839e-06,
+      "loss": 0.7903,
+      "step": 1258
+    },
+    {
+      "epoch": 0.736607306103046,
+      "grad_norm": 0.04207256800472794,
+      "learning_rate": 5.846900902005822e-06,
+      "loss": 0.7673,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7371923794200461,
+      "grad_norm": 0.03928632402273387,
+      "learning_rate": 5.843870433695156e-06,
+      "loss": 0.7548,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7377774527370461,
+      "grad_norm": 0.039954700127276935,
+      "learning_rate": 5.8408387455979946e-06,
+      "loss": 0.7986,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7383625260540462,
+      "grad_norm": 0.03783285103775797,
+      "learning_rate": 5.837805840275515e-06,
+      "loss": 0.8107,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7389475993710461,
+      "grad_norm": 0.040841957568584206,
+      "learning_rate": 5.834771720289929e-06,
+      "loss": 0.8005,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7395326726880462,
+      "grad_norm": 0.044981013127008924,
+      "learning_rate": 5.831736388204467e-06,
+      "loss": 0.7682,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7401177460050462,
+      "grad_norm": 0.038332859477558344,
+      "learning_rate": 5.828699846583389e-06,
+      "loss": 0.7548,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7407028193220463,
+      "grad_norm": 0.04385966631432292,
+      "learning_rate": 5.825662097991978e-06,
+      "loss": 0.8131,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7412878926390464,
+      "grad_norm": 0.038909792748055394,
+      "learning_rate": 5.82262314499653e-06,
+      "loss": 0.8741,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7418729659560463,
+      "grad_norm": 0.05612505051259805,
+      "learning_rate": 5.8195829901643655e-06,
+      "loss": 0.8022,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7424580392730464,
+      "grad_norm": 0.041230289913411614,
+      "learning_rate": 5.816541636063816e-06,
+      "loss": 0.7588,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7430431125900464,
+      "grad_norm": 0.053472936644260796,
+      "learning_rate": 5.813499085264229e-06,
+      "loss": 0.733,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7436281859070465,
+      "grad_norm": 0.05949960419609776,
+      "learning_rate": 5.8104553403359586e-06,
+      "loss": 0.8247,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7442132592240465,
+      "grad_norm": 0.04317467554051099,
+      "learning_rate": 5.807410403850371e-06,
+      "loss": 0.8015,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7447983325410465,
+      "grad_norm": 0.040008771003069384,
+      "learning_rate": 5.804364278379842e-06,
+      "loss": 0.7975,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7453834058580466,
+      "grad_norm": 0.039926332581406196,
+      "learning_rate": 5.801316966497744e-06,
+      "loss": 0.7404,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7459684791750466,
+      "grad_norm": 0.03922271741776474,
+      "learning_rate": 5.798268470778461e-06,
+      "loss": 0.8157,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7465535524920467,
+      "grad_norm": 0.03804854554685112,
+      "learning_rate": 5.795218793797367e-06,
+      "loss": 0.6987,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7471386258090467,
+      "grad_norm": 0.03876336199264868,
+      "learning_rate": 5.792167938130842e-06,
+      "loss": 0.8456,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7477236991260467,
+      "grad_norm": 0.03771896812051095,
+      "learning_rate": 5.78911590635626e-06,
+      "loss": 0.726,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7483087724430467,
+      "grad_norm": 0.036735705010982965,
+      "learning_rate": 5.786062701051983e-06,
+      "loss": 0.9274,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7488938457600468,
+      "grad_norm": 0.03943361612412777,
+      "learning_rate": 5.783008324797375e-06,
+      "loss": 0.7146,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7494789190770469,
+      "grad_norm": 0.03647462654976768,
+      "learning_rate": 5.779952780172777e-06,
+      "loss": 0.7589,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7500639923940469,
+      "grad_norm": 0.03462458239120716,
+      "learning_rate": 5.776896069759528e-06,
+      "loss": 0.7397,
+      "step": 1282
+    },
+    {
+      "epoch": 0.750649065711047,
+      "grad_norm": 0.03495722370436699,
+      "learning_rate": 5.773838196139946e-06,
+      "loss": 0.6993,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7512341390280469,
+      "grad_norm": 0.05321435758446983,
+      "learning_rate": 5.770779161897329e-06,
+      "loss": 0.8397,
+      "step": 1284
+    },
+    {
+      "epoch": 0.751819212345047,
+      "grad_norm": 0.07760562554265127,
+      "learning_rate": 5.767718969615965e-06,
+      "loss": 0.8,
+      "step": 1285
+    },
+    {
+      "epoch": 0.752404285662047,
+      "grad_norm": 0.044871732227994655,
+      "learning_rate": 5.7646576218811095e-06,
+      "loss": 0.7373,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7529893589790471,
+      "grad_norm": 0.040559575159430344,
+      "learning_rate": 5.761595121279002e-06,
+      "loss": 0.8097,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7535744322960471,
+      "grad_norm": 0.03837125490071528,
+      "learning_rate": 5.758531470396853e-06,
+      "loss": 0.8207,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7541595056130471,
+      "grad_norm": 0.0417380184844728,
+      "learning_rate": 5.7554666718228435e-06,
+      "loss": 0.7322,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7547445789300472,
+      "grad_norm": 0.04177068987632399,
+      "learning_rate": 5.752400728146127e-06,
+      "loss": 0.7949,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7553296522470472,
+      "grad_norm": 0.03750486470453801,
+      "learning_rate": 5.74933364195682e-06,
+      "loss": 0.713,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7559147255640473,
+      "grad_norm": 0.04456432301299169,
+      "learning_rate": 5.7462654158460105e-06,
+      "loss": 0.8051,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7564997988810472,
+      "grad_norm": 0.03721091596694818,
+      "learning_rate": 5.743196052405745e-06,
+      "loss": 0.8427,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7570848721980473,
+      "grad_norm": 0.04755935327200047,
+      "learning_rate": 5.74012555422903e-06,
+      "loss": 0.765,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7576699455150474,
+      "grad_norm": 0.038790617615815835,
+      "learning_rate": 5.737053923909832e-06,
+      "loss": 0.7368,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7582550188320474,
+      "grad_norm": 0.03835922133019433,
+      "learning_rate": 5.7339811640430805e-06,
+      "loss": 0.738,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7588400921490475,
+      "grad_norm": 0.04054793146664343,
+      "learning_rate": 5.730907277224645e-06,
+      "loss": 0.8177,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7594251654660474,
+      "grad_norm": 0.04247379928605068,
+      "learning_rate": 5.727832266051361e-06,
+      "loss": 0.8173,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7600102387830475,
+      "grad_norm": 0.0445025614285575,
+      "learning_rate": 5.724756133121007e-06,
+      "loss": 0.8927,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7605953121000475,
+      "grad_norm": 0.0926458972504,
+      "learning_rate": 5.721678881032307e-06,
+      "loss": 0.9005,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7611803854170476,
+      "grad_norm": 0.03742317171404953,
+      "learning_rate": 5.718600512384938e-06,
+      "loss": 0.8046,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7617654587340477,
+      "grad_norm": 0.03916116385336752,
+      "learning_rate": 5.715521029779511e-06,
+      "loss": 0.8223,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7623505320510476,
+      "grad_norm": 0.05168394725564997,
+      "learning_rate": 5.71244043581759e-06,
+      "loss": 0.8097,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7629356053680477,
+      "grad_norm": 0.04369686184899199,
+      "learning_rate": 5.709358733101665e-06,
+      "loss": 0.7624,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7635206786850477,
+      "grad_norm": 0.052165784587932,
+      "learning_rate": 5.7062759242351705e-06,
+      "loss": 0.8156,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7641057520020478,
+      "grad_norm": 0.04082442609405472,
+      "learning_rate": 5.703192011822474e-06,
+      "loss": 0.8779,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7646908253190478,
+      "grad_norm": 0.0416953556866423,
+      "learning_rate": 5.700106998468874e-06,
+      "loss": 0.8806,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7652758986360478,
+      "grad_norm": 0.03820299209890676,
+      "learning_rate": 5.697020886780601e-06,
+      "loss": 0.7473,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7658609719530479,
+      "grad_norm": 0.036763238071425314,
+      "learning_rate": 5.693933679364812e-06,
+      "loss": 0.7509,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7664460452700479,
+      "grad_norm": 0.03779190565018098,
+      "learning_rate": 5.690845378829589e-06,
+      "loss": 0.8345,
+      "step": 1310
+    },
+    {
+      "epoch": 0.767031118587048,
+      "grad_norm": 0.03632799411852302,
+      "learning_rate": 5.68775598778394e-06,
+      "loss": 0.7592,
+      "step": 1311
+    },
+    {
+      "epoch": 0.767616191904048,
+      "grad_norm": 0.04011020299884528,
+      "learning_rate": 5.684665508837792e-06,
+      "loss": 0.7438,
+      "step": 1312
+    },
+    {
+      "epoch": 0.768201265221048,
+      "grad_norm": 0.03991606487653607,
+      "learning_rate": 5.681573944601991e-06,
+      "loss": 0.8176,
+      "step": 1313
+    },
+    {
+      "epoch": 0.768786338538048,
+      "grad_norm": 0.03597560209452427,
+      "learning_rate": 5.678481297688301e-06,
+      "loss": 0.7547,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7693714118550481,
+      "grad_norm": 0.04169549636775327,
+      "learning_rate": 5.675387570709403e-06,
+      "loss": 0.8889,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7699564851720482,
+      "grad_norm": 0.04170869376840914,
+      "learning_rate": 5.672292766278882e-06,
+      "loss": 0.8606,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7705415584890482,
+      "grad_norm": 0.044938956861301685,
+      "learning_rate": 5.669196887011245e-06,
+      "loss": 0.8402,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7711266318060482,
+      "grad_norm": 0.035153707331906876,
+      "learning_rate": 5.666099935521897e-06,
+      "loss": 0.7729,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7717117051230482,
+      "grad_norm": 0.06118327685992558,
+      "learning_rate": 5.663001914427155e-06,
+      "loss": 0.7555,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7722967784400483,
+      "grad_norm": 0.03968655862023267,
+      "learning_rate": 5.659902826344237e-06,
+      "loss": 0.7255,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7728818517570483,
+      "grad_norm": 0.04086333995138192,
+      "learning_rate": 5.656802673891261e-06,
+      "loss": 0.7239,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7734669250740484,
+      "grad_norm": 0.04463158217642648,
+      "learning_rate": 5.6537014596872496e-06,
+      "loss": 0.7722,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7740519983910483,
+      "grad_norm": 0.06012694877055056,
+      "learning_rate": 5.650599186352116e-06,
+      "loss": 0.8534,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7746370717080484,
+      "grad_norm": 0.03696498183729998,
+      "learning_rate": 5.6474958565066705e-06,
+      "loss": 0.8262,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7752221450250485,
+      "grad_norm": 0.0413037337153161,
+      "learning_rate": 5.644391472772617e-06,
+      "loss": 0.7777,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7758072183420485,
+      "grad_norm": 0.03916694037724885,
+      "learning_rate": 5.641286037772551e-06,
+      "loss": 0.7261,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7763922916590486,
+      "grad_norm": 0.042884361024367525,
+      "learning_rate": 5.638179554129953e-06,
+      "loss": 0.8306,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7769773649760485,
+      "grad_norm": 0.04318783701390697,
+      "learning_rate": 5.635072024469189e-06,
+      "loss": 0.7502,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7775624382930486,
+      "grad_norm": 0.04238539765681867,
+      "learning_rate": 5.631963451415512e-06,
+      "loss": 0.8651,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7781475116100486,
+      "grad_norm": 0.0606324772137342,
+      "learning_rate": 5.628853837595055e-06,
+      "loss": 0.8407,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7787325849270487,
+      "grad_norm": 0.03973726867422112,
+      "learning_rate": 5.62574318563483e-06,
+      "loss": 0.7456,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7793176582440487,
+      "grad_norm": 0.04553233971802566,
+      "learning_rate": 5.622631498162723e-06,
+      "loss": 0.7778,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7799027315610487,
+      "grad_norm": 0.045327548192169206,
+      "learning_rate": 5.619518777807502e-06,
+      "loss": 0.7996,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 0.05867372409357833,
+      "learning_rate": 5.616405027198802e-06,
+      "loss": 0.8378,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7810728781950488,
+      "grad_norm": 0.03626782443560853,
+      "learning_rate": 5.613290248967128e-06,
+      "loss": 0.7915,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7816579515120489,
+      "grad_norm": 0.03991616366449104,
+      "learning_rate": 5.610174445743857e-06,
+      "loss": 0.7521,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7822430248290488,
+      "grad_norm": 0.03644002616250411,
+      "learning_rate": 5.6070576201612286e-06,
+      "loss": 0.8306,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7828280981460489,
+      "grad_norm": 0.03526533642183784,
+      "learning_rate": 5.603939774852349e-06,
+      "loss": 0.7778,
+      "step": 1338
+    },
+    {
+      "epoch": 0.783413171463049,
+      "grad_norm": 0.05592830536988608,
+      "learning_rate": 5.6008209124511825e-06,
+      "loss": 0.804,
+      "step": 1339
+    },
+    {
+      "epoch": 0.783998244780049,
+      "grad_norm": 0.04740521258730855,
+      "learning_rate": 5.597701035592555e-06,
+      "loss": 0.8071,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7845833180970491,
+      "grad_norm": 0.22162476710778273,
+      "learning_rate": 5.59458014691215e-06,
+      "loss": 0.8066,
+      "step": 1341
+    },
+    {
+      "epoch": 0.785168391414049,
+      "grad_norm": 0.09169014540583856,
+      "learning_rate": 5.5914582490465015e-06,
+      "loss": 0.8464,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7857534647310491,
+      "grad_norm": 0.039007767079512076,
+      "learning_rate": 5.588335344633001e-06,
+      "loss": 0.779,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7863385380480491,
+      "grad_norm": 0.04485954456072006,
+      "learning_rate": 5.585211436309887e-06,
+      "loss": 0.777,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7869236113650492,
+      "grad_norm": 0.039288691984984164,
+      "learning_rate": 5.58208652671625e-06,
+      "loss": 0.8205,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7875086846820493,
+      "grad_norm": 0.04480005712178459,
+      "learning_rate": 5.578960618492022e-06,
+      "loss": 0.8083,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7880937579990492,
+      "grad_norm": 0.03846704640576893,
+      "learning_rate": 5.575833714277983e-06,
+      "loss": 0.7769,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7886788313160493,
+      "grad_norm": 0.0414560185138673,
+      "learning_rate": 5.572705816715748e-06,
+      "loss": 0.8654,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7892639046330493,
+      "grad_norm": 0.0501606429389028,
+      "learning_rate": 5.569576928447778e-06,
+      "loss": 0.7904,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7898489779500494,
+      "grad_norm": 0.041891454878211096,
+      "learning_rate": 5.566447052117368e-06,
+      "loss": 0.7469,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7904340512670494,
+      "grad_norm": 0.04101885937248216,
+      "learning_rate": 5.563316190368648e-06,
+      "loss": 0.759,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7910191245840494,
+      "grad_norm": 0.036683416502670804,
+      "learning_rate": 5.560184345846577e-06,
+      "loss": 0.8344,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7916041979010495,
+      "grad_norm": 0.045836001062305856,
+      "learning_rate": 5.557051521196952e-06,
+      "loss": 0.8326,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7921892712180495,
+      "grad_norm": 0.0397925205994133,
+      "learning_rate": 5.553917719066391e-06,
+      "loss": 0.8104,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7927743445350496,
+      "grad_norm": 0.041024359015282515,
+      "learning_rate": 5.5507829421023415e-06,
+      "loss": 0.8124,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7933594178520496,
+      "grad_norm": 0.03779172917164305,
+      "learning_rate": 5.547647192953073e-06,
+      "loss": 0.788,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7939444911690496,
+      "grad_norm": 0.03788055346324226,
+      "learning_rate": 5.544510474267677e-06,
+      "loss": 0.7715,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7945295644860496,
+      "grad_norm": 0.03853302326944159,
+      "learning_rate": 5.541372788696063e-06,
+      "loss": 0.7159,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7951146378030497,
+      "grad_norm": 0.03754302624204286,
+      "learning_rate": 5.538234138888959e-06,
+      "loss": 0.8184,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7956997111200498,
+      "grad_norm": 0.0394348592493888,
+      "learning_rate": 5.535094527497905e-06,
+      "loss": 0.8367,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7962847844370498,
+      "grad_norm": 0.03906040463168052,
+      "learning_rate": 5.531953957175258e-06,
+      "loss": 0.7447,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7968698577540498,
+      "grad_norm": 0.03608720224204406,
+      "learning_rate": 5.528812430574179e-06,
+      "loss": 0.7342,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7974549310710498,
+      "grad_norm": 0.04229914698861472,
+      "learning_rate": 5.525669950348644e-06,
+      "loss": 0.7747,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7980400043880499,
+      "grad_norm": 0.037880293992017665,
+      "learning_rate": 5.522526519153425e-06,
+      "loss": 0.806,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7986250777050499,
+      "grad_norm": 0.03984285750551966,
+      "learning_rate": 5.519382139644108e-06,
+      "loss": 0.7931,
+      "step": 1365
+    },
+    {
+      "epoch": 0.79921015102205,
+      "grad_norm": 0.08742568964292106,
+      "learning_rate": 5.516236814477073e-06,
+      "loss": 0.783,
+      "step": 1366
+    },
+    {
+      "epoch": 0.79979522433905,
+      "grad_norm": 0.05924902432803569,
+      "learning_rate": 5.513090546309501e-06,
+      "loss": 0.8574,
+      "step": 1367
+    },
+    {
+      "epoch": 0.80038029765605,
+      "grad_norm": 0.03846930309455261,
+      "learning_rate": 5.50994333779937e-06,
+      "loss": 0.7829,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8009653709730501,
+      "grad_norm": 0.03668377181993435,
+      "learning_rate": 5.506795191605454e-06,
+      "loss": 0.7873,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8015504442900501,
+      "grad_norm": 0.04173707671452499,
+      "learning_rate": 5.503646110387313e-06,
+      "loss": 0.8333,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8021355176070502,
+      "grad_norm": 0.03842538857985326,
+      "learning_rate": 5.500496096805305e-06,
+      "loss": 0.6669,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8027205909240501,
+      "grad_norm": 0.045318530527092046,
+      "learning_rate": 5.49734515352057e-06,
+      "loss": 0.7893,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8033056642410502,
+      "grad_norm": 0.042283011239155184,
+      "learning_rate": 5.494193283195036e-06,
+      "loss": 0.7635,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8038907375580503,
+      "grad_norm": 0.042565824937185165,
+      "learning_rate": 5.491040488491412e-06,
+      "loss": 0.7387,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8044758108750503,
+      "grad_norm": 0.040462074923080384,
+      "learning_rate": 5.487886772073193e-06,
+      "loss": 0.7739,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8050608841920504,
+      "grad_norm": 0.03792367430136744,
+      "learning_rate": 5.484732136604644e-06,
+      "loss": 0.7566,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8056459575090503,
+      "grad_norm": 0.03692580293915857,
+      "learning_rate": 5.481576584750816e-06,
+      "loss": 0.8481,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8062310308260504,
+      "grad_norm": 0.046451787891487024,
+      "learning_rate": 5.478420119177527e-06,
+      "loss": 0.8445,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8068161041430504,
+      "grad_norm": 0.04084326621384236,
+      "learning_rate": 5.47526274255137e-06,
+      "loss": 0.7751,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8074011774600505,
+      "grad_norm": 0.04059410095244892,
+      "learning_rate": 5.4721044575397074e-06,
+      "loss": 0.8631,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8079862507770506,
+      "grad_norm": 0.04272865593875444,
+      "learning_rate": 5.468945266810668e-06,
+      "loss": 0.7363,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8085713240940505,
+      "grad_norm": 0.040838508223671875,
+      "learning_rate": 5.4657851730331475e-06,
+      "loss": 0.7909,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8091563974110506,
+      "grad_norm": 0.03888480720535378,
+      "learning_rate": 5.462624178876804e-06,
+      "loss": 0.7638,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8097414707280506,
+      "grad_norm": 0.04210282976595793,
+      "learning_rate": 5.459462287012054e-06,
+      "loss": 0.7131,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8103265440450507,
+      "grad_norm": 0.041245050616825804,
+      "learning_rate": 5.456299500110075e-06,
+      "loss": 0.798,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8109116173620506,
+      "grad_norm": 0.03765154790140343,
+      "learning_rate": 5.453135820842801e-06,
+      "loss": 0.7585,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8114966906790507,
+      "grad_norm": 0.04000694111877368,
+      "learning_rate": 5.449971251882917e-06,
+      "loss": 0.7911,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8120817639960508,
+      "grad_norm": 0.03476195859648224,
+      "learning_rate": 5.4468057959038604e-06,
+      "loss": 0.761,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8126668373130508,
+      "grad_norm": 0.03902654854791974,
+      "learning_rate": 5.4436394555798215e-06,
+      "loss": 0.7841,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8132519106300509,
+      "grad_norm": 0.04263837806916918,
+      "learning_rate": 5.4404722335857345e-06,
+      "loss": 0.8001,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8138369839470508,
+      "grad_norm": 0.04109922825698143,
+      "learning_rate": 5.437304132597278e-06,
+      "loss": 0.7333,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8144220572640509,
+      "grad_norm": 0.06739804898618297,
+      "learning_rate": 5.4341351552908744e-06,
+      "loss": 0.7495,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8150071305810509,
+      "grad_norm": 0.04932649676543138,
+      "learning_rate": 5.430965304343685e-06,
+      "loss": 0.8227,
+      "step": 1393
+    },
+    {
+      "epoch": 0.815592203898051,
+      "grad_norm": 0.0394930305814403,
+      "learning_rate": 5.427794582433614e-06,
+      "loss": 0.7627,
+      "step": 1394
+    },
+    {
+      "epoch": 0.816177277215051,
+      "grad_norm": 0.03667171398926786,
+      "learning_rate": 5.424622992239294e-06,
+      "loss": 0.7338,
+      "step": 1395
+    },
+    {
+      "epoch": 0.816762350532051,
+      "grad_norm": 0.03845361319571211,
+      "learning_rate": 5.421450536440095e-06,
+      "loss": 0.8347,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8173474238490511,
+      "grad_norm": 0.042037246421225145,
+      "learning_rate": 5.41827721771612e-06,
+      "loss": 0.749,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8179324971660511,
+      "grad_norm": 0.03891125248418732,
+      "learning_rate": 5.415103038748198e-06,
+      "loss": 0.744,
+      "step": 1398
+    },
+    {
+      "epoch": 0.8185175704830512,
+      "grad_norm": 0.06129253436260354,
+      "learning_rate": 5.411928002217888e-06,
+      "loss": 0.7449,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8191026438000512,
+      "grad_norm": 0.038946849210413866,
+      "learning_rate": 5.408752110807467e-06,
+      "loss": 0.7716,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8196877171170512,
+      "grad_norm": 0.033497315808280095,
+      "learning_rate": 5.405575367199942e-06,
+      "loss": 0.8,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8202727904340512,
+      "grad_norm": 0.042207165872255904,
+      "learning_rate": 5.402397774079036e-06,
+      "loss": 0.7393,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8208578637510513,
+      "grad_norm": 0.03955442411585693,
+      "learning_rate": 5.3992193341291886e-06,
+      "loss": 0.7736,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8214429370680514,
+      "grad_norm": 0.05003446882200215,
+      "learning_rate": 5.396040050035557e-06,
+      "loss": 0.7875,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8220280103850514,
+      "grad_norm": 1.585774424689007,
+      "learning_rate": 5.392859924484012e-06,
+      "loss": 0.7529,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8226130837020514,
+      "grad_norm": 0.03926754895363517,
+      "learning_rate": 5.389678960161134e-06,
+      "loss": 0.7126,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8231981570190514,
+      "grad_norm": 0.045548787620002675,
+      "learning_rate": 5.3864971597542095e-06,
+      "loss": 0.8089,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8237832303360515,
+      "grad_norm": 0.050549309003184176,
+      "learning_rate": 5.383314525951239e-06,
+      "loss": 0.7794,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8243683036530515,
+      "grad_norm": 0.03761672158310273,
+      "learning_rate": 5.380131061440916e-06,
+      "loss": 0.813,
+      "step": 1409
+    },
+    {
+      "epoch": 0.8249533769700516,
+      "grad_norm": 0.06445609012820583,
+      "learning_rate": 5.376946768912645e-06,
+      "loss": 0.7872,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8255384502870516,
+      "grad_norm": 0.04256126719536265,
+      "learning_rate": 5.3737616510565285e-06,
+      "loss": 0.8119,
+      "step": 1411
+    },
+    {
+      "epoch": 0.8261235236040516,
+      "grad_norm": 0.041183658130345274,
+      "learning_rate": 5.370575710563362e-06,
+      "loss": 0.6822,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8267085969210517,
+      "grad_norm": 0.04394347764957121,
+      "learning_rate": 5.367388950124638e-06,
+      "loss": 0.8018,
+      "step": 1413
+    },
+    {
+      "epoch": 0.8272936702380517,
+      "grad_norm": 0.0489504060626976,
+      "learning_rate": 5.364201372432544e-06,
+      "loss": 0.7585,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8278787435550518,
+      "grad_norm": 0.03696771639000695,
+      "learning_rate": 5.361012980179955e-06,
+      "loss": 0.7659,
+      "step": 1415
+    },
+    {
+      "epoch": 0.8284638168720517,
+      "grad_norm": 0.03914887066611441,
+      "learning_rate": 5.357823776060436e-06,
+      "loss": 0.7116,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8290488901890518,
+      "grad_norm": 0.0509069459164667,
+      "learning_rate": 5.3546337627682375e-06,
+      "loss": 0.8464,
+      "step": 1417
+    },
+    {
+      "epoch": 0.8296339635060519,
+      "grad_norm": 0.03774518281104574,
+      "learning_rate": 5.351442942998291e-06,
+      "loss": 0.7202,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8302190368230519,
+      "grad_norm": 0.04736561860882581,
+      "learning_rate": 5.348251319446215e-06,
+      "loss": 0.7846,
+      "step": 1419
+    },
+    {
+      "epoch": 0.830804110140052,
+      "grad_norm": 0.05148202522600593,
+      "learning_rate": 5.3450588948083e-06,
+      "loss": 0.8305,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8313891834570519,
+      "grad_norm": 0.043173058947526516,
+      "learning_rate": 5.341865671781521e-06,
+      "loss": 0.7154,
+      "step": 1421
+    },
+    {
+      "epoch": 0.831974256774052,
+      "grad_norm": 0.07628502093896318,
+      "learning_rate": 5.338671653063519e-06,
+      "loss": 0.7989,
+      "step": 1422
+    },
+    {
+      "epoch": 0.832559330091052,
+      "grad_norm": 0.055710961638667414,
+      "learning_rate": 5.335476841352614e-06,
+      "loss": 0.8053,
+      "step": 1423
+    },
+    {
+      "epoch": 0.8331444034080521,
+      "grad_norm": 0.03907449804352977,
+      "learning_rate": 5.3322812393477926e-06,
+      "loss": 0.768,
+      "step": 1424
+    },
+    {
+      "epoch": 0.8337294767250522,
+      "grad_norm": 0.05390227526387696,
+      "learning_rate": 5.32908484974871e-06,
+      "loss": 0.8034,
+      "step": 1425
+    },
+    {
+      "epoch": 0.8343145500420521,
+      "grad_norm": 0.0398118255725142,
+      "learning_rate": 5.325887675255688e-06,
+      "loss": 0.813,
+      "step": 1426
+    },
+    {
+      "epoch": 0.8348996233590522,
+      "grad_norm": 0.04145492964896595,
+      "learning_rate": 5.3226897185697085e-06,
+      "loss": 0.7782,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8354846966760522,
+      "grad_norm": 0.045142573535406814,
+      "learning_rate": 5.319490982392416e-06,
+      "loss": 0.7923,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8360697699930523,
+      "grad_norm": 0.0417853187675979,
+      "learning_rate": 5.316291469426113e-06,
+      "loss": 0.7188,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8366548433100522,
+      "grad_norm": 0.03824273821943993,
+      "learning_rate": 5.313091182373761e-06,
+      "loss": 0.8268,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8372399166270523,
+      "grad_norm": 0.04265515858265742,
+      "learning_rate": 5.309890123938971e-06,
+      "loss": 0.7841,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8378249899440524,
+      "grad_norm": 0.04319663661457448,
+      "learning_rate": 5.306688296826007e-06,
+      "loss": 0.8278,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8384100632610524,
+      "grad_norm": 0.04289615774239121,
+      "learning_rate": 5.3034857037397855e-06,
+      "loss": 0.7259,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8389951365780525,
+      "grad_norm": 0.042931577411094914,
+      "learning_rate": 5.3002823473858645e-06,
+      "loss": 0.8723,
+      "step": 1434
+    },
+    {
+      "epoch": 0.8395802098950524,
+      "grad_norm": 0.06290471800979577,
+      "learning_rate": 5.297078230470454e-06,
+      "loss": 0.8477,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8401652832120525,
+      "grad_norm": 0.03920031661089491,
+      "learning_rate": 5.293873355700399e-06,
+      "loss": 0.7391,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8407503565290525,
+      "grad_norm": 0.04019727257985879,
+      "learning_rate": 5.290667725783189e-06,
+      "loss": 0.7166,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8413354298460526,
+      "grad_norm": 0.05482096600060075,
+      "learning_rate": 5.287461343426953e-06,
+      "loss": 0.7527,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8419205031630527,
+      "grad_norm": 0.08758101157201069,
+      "learning_rate": 5.28425421134045e-06,
+      "loss": 0.8409,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8425055764800526,
+      "grad_norm": 0.042702212748573325,
+      "learning_rate": 5.281046332233078e-06,
+      "loss": 0.7841,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8430906497970527,
+      "grad_norm": 0.058403994145729214,
+      "learning_rate": 5.277837708814864e-06,
+      "loss": 0.8108,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8436757231140527,
+      "grad_norm": 0.04768258856736286,
+      "learning_rate": 5.274628343796462e-06,
+      "loss": 0.8224,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8442607964310528,
+      "grad_norm": 0.036438567596678934,
+      "learning_rate": 5.271418239889155e-06,
+      "loss": 0.8098,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8448458697480528,
+      "grad_norm": 0.038906276549651175,
+      "learning_rate": 5.268207399804851e-06,
+      "loss": 0.7441,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8454309430650528,
+      "grad_norm": 0.04178256613424276,
+      "learning_rate": 5.264995826256074e-06,
+      "loss": 0.8095,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8460160163820529,
+      "grad_norm": 0.0567861747607305,
+      "learning_rate": 5.2617835219559775e-06,
+      "loss": 0.7407,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8466010896990529,
+      "grad_norm": 0.04554919492814466,
+      "learning_rate": 5.258570489618321e-06,
+      "loss": 0.7694,
+      "step": 1447
+    },
+    {
+      "epoch": 0.847186163016053,
+      "grad_norm": 0.048208046364352455,
+      "learning_rate": 5.255356731957488e-06,
+      "loss": 0.7255,
+      "step": 1448
+    },
+    {
+      "epoch": 0.847771236333053,
+      "grad_norm": 0.040443603231303464,
+      "learning_rate": 5.252142251688474e-06,
+      "loss": 0.8018,
+      "step": 1449
+    },
+    {
+      "epoch": 0.848356309650053,
+      "grad_norm": 0.06530684120775691,
+      "learning_rate": 5.248927051526877e-06,
+      "loss": 0.7677,
+      "step": 1450
+    },
+    {
+      "epoch": 0.848941382967053,
+      "grad_norm": 0.039867530196060805,
+      "learning_rate": 5.245711134188912e-06,
+      "loss": 0.769,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8495264562840531,
+      "grad_norm": 0.04207996814905851,
+      "learning_rate": 5.242494502391395e-06,
+      "loss": 0.7442,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8501115296010532,
+      "grad_norm": 0.04561472811714227,
+      "learning_rate": 5.239277158851749e-06,
+      "loss": 0.7115,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8506966029180532,
+      "grad_norm": 0.04360996606853927,
+      "learning_rate": 5.236059106287996e-06,
+      "loss": 0.8547,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8512816762350532,
+      "grad_norm": 0.04267227083806566,
+      "learning_rate": 5.232840347418756e-06,
+      "loss": 0.7316,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8518667495520532,
+      "grad_norm": 0.04424954078490177,
+      "learning_rate": 5.229620884963249e-06,
+      "loss": 0.7864,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8524518228690533,
+      "grad_norm": 0.043831224967070745,
+      "learning_rate": 5.226400721641285e-06,
+      "loss": 0.7831,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8530368961860533,
+      "grad_norm": 0.041068348295436305,
+      "learning_rate": 5.223179860173271e-06,
+      "loss": 0.864,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8536219695030534,
+      "grad_norm": 0.03891142553337627,
+      "learning_rate": 5.219958303280201e-06,
+      "loss": 0.8002,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8542070428200534,
+      "grad_norm": 0.038910550390794815,
+      "learning_rate": 5.216736053683655e-06,
+      "loss": 0.6932,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8547921161370534,
+      "grad_norm": 0.08030189721000423,
+      "learning_rate": 5.213513114105802e-06,
+      "loss": 0.7286,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8553771894540535,
+      "grad_norm": 0.0376740189158807,
+      "learning_rate": 5.210289487269392e-06,
+      "loss": 0.7832,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8559622627710535,
+      "grad_norm": 0.03877582322418199,
+      "learning_rate": 5.207065175897753e-06,
+      "loss": 0.8055,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8565473360880536,
+      "grad_norm": 0.041349052303152146,
+      "learning_rate": 5.203840182714796e-06,
+      "loss": 0.6915,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8571324094050535,
+      "grad_norm": 0.04301186186037183,
+      "learning_rate": 5.200614510445007e-06,
+      "loss": 0.7925,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8577174827220536,
+      "grad_norm": 0.04787283684151834,
+      "learning_rate": 5.197388161813443e-06,
+      "loss": 0.7505,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8583025560390537,
+      "grad_norm": 0.09574674570840018,
+      "learning_rate": 5.194161139545733e-06,
+      "loss": 0.8076,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8588876293560537,
+      "grad_norm": 0.04155223740576836,
+      "learning_rate": 5.190933446368078e-06,
+      "loss": 0.7496,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8594727026730538,
+      "grad_norm": 0.03701742254247571,
+      "learning_rate": 5.187705085007243e-06,
+      "loss": 0.7563,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8600577759900537,
+      "grad_norm": 0.04678171662949799,
+      "learning_rate": 5.184476058190558e-06,
+      "loss": 0.7159,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8606428493070538,
+      "grad_norm": 0.04884408311526059,
+      "learning_rate": 5.181246368645916e-06,
+      "loss": 0.9046,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8612279226240538,
+      "grad_norm": 0.04557695869573877,
+      "learning_rate": 5.1780160191017685e-06,
+      "loss": 0.8452,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8618129959410539,
+      "grad_norm": 0.04072103320834028,
+      "learning_rate": 5.1747850122871275e-06,
+      "loss": 0.7661,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8623980692580538,
+      "grad_norm": 0.062070335429827074,
+      "learning_rate": 5.171553350931555e-06,
+      "loss": 0.7456,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8629831425750539,
+      "grad_norm": 0.04041423716100811,
+      "learning_rate": 5.1683210377651716e-06,
+      "loss": 0.6993,
+      "step": 1475
+    },
+    {
+      "epoch": 0.863568215892054,
+      "grad_norm": 0.037496335851542203,
+      "learning_rate": 5.165088075518646e-06,
+      "loss": 0.7977,
+      "step": 1476
+    },
+    {
+      "epoch": 0.864153289209054,
+      "grad_norm": 0.04125529850667817,
+      "learning_rate": 5.161854466923194e-06,
+      "loss": 0.803,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8647383625260541,
+      "grad_norm": 0.06648674433353301,
+      "learning_rate": 5.15862021471058e-06,
+      "loss": 0.7707,
+      "step": 1478
+    },
+    {
+      "epoch": 0.865323435843054,
+      "grad_norm": 0.04189383564593303,
+      "learning_rate": 5.155385321613111e-06,
+      "loss": 0.8043,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8659085091600541,
+      "grad_norm": 0.04233513271829965,
+      "learning_rate": 5.1521497903636335e-06,
+      "loss": 0.7909,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8664935824770541,
+      "grad_norm": 0.03892645909618685,
+      "learning_rate": 5.148913623695536e-06,
+      "loss": 0.7712,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8670786557940542,
+      "grad_norm": 0.042788389662538735,
+      "learning_rate": 5.145676824342745e-06,
+      "loss": 0.8628,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8676637291110543,
+      "grad_norm": 0.03339414866019395,
+      "learning_rate": 5.142439395039715e-06,
+      "loss": 0.8101,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8682488024280542,
+      "grad_norm": 0.04123013617638013,
+      "learning_rate": 5.139201338521442e-06,
+      "loss": 0.7808,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8688338757450543,
+      "grad_norm": 0.15209820057805395,
+      "learning_rate": 5.135962657523442e-06,
+      "loss": 0.8155,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8694189490620543,
+      "grad_norm": 0.038333113007498906,
+      "learning_rate": 5.1327233547817655e-06,
+      "loss": 0.8114,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8700040223790544,
+      "grad_norm": 0.1091230339415769,
+      "learning_rate": 5.129483433032988e-06,
+      "loss": 0.7413,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8705890956960544,
+      "grad_norm": 0.04419253223096684,
+      "learning_rate": 5.126242895014202e-06,
+      "loss": 0.8002,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8711741690130544,
+      "grad_norm": 0.03647244403949355,
+      "learning_rate": 5.123001743463026e-06,
+      "loss": 0.7514,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8717592423300545,
+      "grad_norm": 0.040974797817733756,
+      "learning_rate": 5.119759981117599e-06,
+      "loss": 0.7391,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8723443156470545,
+      "grad_norm": 0.03756329508778396,
+      "learning_rate": 5.116517610716568e-06,
+      "loss": 0.7665,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8729293889640546,
+      "grad_norm": 0.03838778975594434,
+      "learning_rate": 5.113274634999101e-06,
+      "loss": 0.7946,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8735144622810546,
+      "grad_norm": 0.04331410260349596,
+      "learning_rate": 5.110031056704872e-06,
+      "loss": 0.8715,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8740995355980546,
+      "grad_norm": 0.0394075894195183,
+      "learning_rate": 5.10678687857407e-06,
+      "loss": 0.7063,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8746846089150546,
+      "grad_norm": 0.04199482290559279,
+      "learning_rate": 5.103542103347385e-06,
+      "loss": 0.7819,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8752696822320547,
+      "grad_norm": 0.03998823252954996,
+      "learning_rate": 5.100296733766013e-06,
+      "loss": 0.6424,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8758547555490548,
+      "grad_norm": 0.04740849060105482,
+      "learning_rate": 5.0970507725716545e-06,
+      "loss": 0.8076,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8764398288660548,
+      "grad_norm": 0.03967591518671566,
+      "learning_rate": 5.09380422250651e-06,
+      "loss": 0.7159,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8770249021830548,
+      "grad_norm": 0.040850522852139234,
+      "learning_rate": 5.090557086313272e-06,
+      "loss": 0.767,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8776099755000548,
+      "grad_norm": 0.15770863380367606,
+      "learning_rate": 5.087309366735134e-06,
+      "loss": 0.8364,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8781950488170549,
+      "grad_norm": 0.03898410562227323,
+      "learning_rate": 5.0840610665157806e-06,
+      "loss": 0.7527,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8787801221340549,
+      "grad_norm": 0.040414642012746985,
+      "learning_rate": 5.080812188399384e-06,
+      "loss": 0.7949,
+      "step": 1502
+    },
+    {
+      "epoch": 0.879365195451055,
+      "grad_norm": 0.04083305581647033,
+      "learning_rate": 5.0775627351306094e-06,
+      "loss": 0.7524,
+      "step": 1503
+    },
+    {
+      "epoch": 0.879950268768055,
+      "grad_norm": 0.032514118714502355,
+      "learning_rate": 5.074312709454604e-06,
+      "loss": 0.729,
+      "step": 1504
+    },
+    {
+      "epoch": 0.880535342085055,
+      "grad_norm": 0.038019163282452334,
+      "learning_rate": 5.0710621141170015e-06,
+      "loss": 0.7649,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8811204154020551,
+      "grad_norm": 0.03465271277149766,
+      "learning_rate": 5.0678109518639164e-06,
+      "loss": 0.7869,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8817054887190551,
+      "grad_norm": 0.03772455496689458,
+      "learning_rate": 5.06455922544194e-06,
+      "loss": 0.7613,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8822905620360552,
+      "grad_norm": 0.037177660173090675,
+      "learning_rate": 5.06130693759814e-06,
+      "loss": 0.7243,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8828756353530551,
+      "grad_norm": 0.039221859770761806,
+      "learning_rate": 5.058054091080063e-06,
+      "loss": 0.7136,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8834607086700552,
+      "grad_norm": 0.048694373606749476,
+      "learning_rate": 5.054800688635721e-06,
+      "loss": 0.7674,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8840457819870553,
+      "grad_norm": 0.04046266225913142,
+      "learning_rate": 5.051546733013602e-06,
+      "loss": 0.7526,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8846308553040553,
+      "grad_norm": 0.04165158889305381,
+      "learning_rate": 5.048292226962659e-06,
+      "loss": 0.7713,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8852159286210554,
+      "grad_norm": 0.03903357348513953,
+      "learning_rate": 5.045037173232306e-06,
+      "loss": 0.7989,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8858010019380553,
+      "grad_norm": 0.03893629709023498,
+      "learning_rate": 5.041781574572425e-06,
+      "loss": 0.7602,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8863860752550554,
+      "grad_norm": 0.053074781240617146,
+      "learning_rate": 5.038525433733358e-06,
+      "loss": 0.7197,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8869711485720554,
+      "grad_norm": 0.03986757723797883,
+      "learning_rate": 5.035268753465901e-06,
+      "loss": 0.7613,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8875562218890555,
+      "grad_norm": 0.03768647901063803,
+      "learning_rate": 5.032011536521309e-06,
+      "loss": 0.8723,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8881412952060556,
+      "grad_norm": 0.04233597893650536,
+      "learning_rate": 5.0287537856512874e-06,
+      "loss": 0.829,
+      "step": 1518
+    },
+    {
+      "epoch": 0.8887263685230555,
+      "grad_norm": 0.03833944818184599,
+      "learning_rate": 5.025495503608001e-06,
+      "loss": 0.7874,
+      "step": 1519
+    },
+    {
+      "epoch": 0.8893114418400556,
+      "grad_norm": 0.038092427816242835,
+      "learning_rate": 5.0222366931440515e-06,
+      "loss": 0.7448,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8898965151570556,
+      "grad_norm": 0.03305492326165736,
+      "learning_rate": 5.018977357012494e-06,
+      "loss": 0.6638,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8904815884740557,
+      "grad_norm": 0.04283989577661302,
+      "learning_rate": 5.015717497966829e-06,
+      "loss": 0.773,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8910666617910556,
+      "grad_norm": 0.03767830677413742,
+      "learning_rate": 5.012457118760993e-06,
+      "loss": 0.746,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8916517351080557,
+      "grad_norm": 0.31951539666885187,
+      "learning_rate": 5.009196222149366e-06,
+      "loss": 0.8181,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8922368084250558,
+      "grad_norm": 0.04141411707140542,
+      "learning_rate": 5.005934810886765e-06,
+      "loss": 0.7699,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8928218817420558,
+      "grad_norm": 0.04045968293380357,
+      "learning_rate": 5.0026728877284415e-06,
+      "loss": 0.7258,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8934069550590559,
+      "grad_norm": 0.039404297272624296,
+      "learning_rate": 4.999410455430076e-06,
+      "loss": 0.7537,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8939920283760558,
+      "grad_norm": 0.041014176601729714,
+      "learning_rate": 4.9961475167477856e-06,
+      "loss": 0.7699,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8945771016930559,
+      "grad_norm": 0.038058343537700656,
+      "learning_rate": 4.992884074438108e-06,
+      "loss": 0.7459,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8951621750100559,
+      "grad_norm": 0.038810739375882135,
+      "learning_rate": 4.989620131258013e-06,
+      "loss": 0.7668,
+      "step": 1530
+    },
+    {
+      "epoch": 0.895747248327056,
+      "grad_norm": 0.042807540000545895,
+      "learning_rate": 4.98635568996489e-06,
+      "loss": 0.7196,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8963323216440561,
+      "grad_norm": 0.04859312187849364,
+      "learning_rate": 4.9830907533165485e-06,
+      "loss": 0.743,
+      "step": 1532
+    },
+    {
+      "epoch": 0.896917394961056,
+      "grad_norm": 0.04633256062984912,
+      "learning_rate": 4.979825324071219e-06,
+      "loss": 0.788,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8975024682780561,
+      "grad_norm": 0.03792342373881924,
+      "learning_rate": 4.976559404987544e-06,
+      "loss": 0.8347,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8980875415950561,
+      "grad_norm": 0.043360533681012677,
+      "learning_rate": 4.973292998824588e-06,
+      "loss": 0.7931,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8986726149120562,
+      "grad_norm": 0.03665588059872115,
+      "learning_rate": 4.970026108341818e-06,
+      "loss": 0.8341,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8992576882290562,
+      "grad_norm": 0.03911321895257535,
+      "learning_rate": 4.966758736299114e-06,
+      "loss": 0.7022,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8998427615460562,
+      "grad_norm": 0.03807679640234228,
+      "learning_rate": 4.963490885456764e-06,
+      "loss": 0.7606,
+      "step": 1538
+    },
+    {
+      "epoch": 0.9004278348630563,
+      "grad_norm": 0.047813376297185586,
+      "learning_rate": 4.9602225585754575e-06,
+      "loss": 0.851,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9010129081800563,
+      "grad_norm": 0.04059937455713067,
+      "learning_rate": 4.95695375841629e-06,
+      "loss": 0.8127,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9015979814970564,
+      "grad_norm": 0.05021922046354935,
+      "learning_rate": 4.953684487740751e-06,
+      "loss": 0.8058,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9021830548140564,
+      "grad_norm": 0.036688823164468184,
+      "learning_rate": 4.950414749310736e-06,
+      "loss": 0.7953,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9027681281310564,
+      "grad_norm": 0.07379482387202636,
+      "learning_rate": 4.947144545888528e-06,
+      "loss": 0.7334,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9033532014480564,
+      "grad_norm": 0.0436885604988347,
+      "learning_rate": 4.943873880236805e-06,
+      "loss": 0.7843,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9039382747650565,
+      "grad_norm": 0.037138950603287514,
+      "learning_rate": 4.940602755118637e-06,
+      "loss": 0.7426,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9045233480820565,
+      "grad_norm": 0.06299316081615561,
+      "learning_rate": 4.93733117329748e-06,
+      "loss": 0.7667,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9051084213990566,
+      "grad_norm": 0.14368489621412747,
+      "learning_rate": 4.934059137537182e-06,
+      "loss": 0.7187,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9056934947160566,
+      "grad_norm": 0.04204980565272586,
+      "learning_rate": 4.9307866506019625e-06,
+      "loss": 0.775,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9062785680330566,
+      "grad_norm": 0.04379760553699639,
+      "learning_rate": 4.927513715256434e-06,
+      "loss": 0.7758,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9068636413500567,
+      "grad_norm": 0.03874093982602192,
+      "learning_rate": 4.9242403342655815e-06,
+      "loss": 0.8522,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9074487146670567,
+      "grad_norm": 0.16908575016119626,
+      "learning_rate": 4.920966510394769e-06,
+      "loss": 0.7786,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9080337879840568,
+      "grad_norm": 0.03659836060994663,
+      "learning_rate": 4.917692246409733e-06,
+      "loss": 0.7795,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9086188613010567,
+      "grad_norm": 0.040782376168304955,
+      "learning_rate": 4.9144175450765814e-06,
+      "loss": 0.7682,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9092039346180568,
+      "grad_norm": 0.04146338306008847,
+      "learning_rate": 4.911142409161795e-06,
+      "loss": 0.7366,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9097890079350569,
+      "grad_norm": 0.04286248778324747,
+      "learning_rate": 4.9078668414322176e-06,
+      "loss": 0.7368,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9103740812520569,
+      "grad_norm": 0.04888887313290673,
+      "learning_rate": 4.9045908446550595e-06,
+      "loss": 0.7896,
+      "step": 1556
+    },
+    {
+      "epoch": 0.910959154569057,
+      "grad_norm": 0.03802578544146469,
+      "learning_rate": 4.901314421597895e-06,
+      "loss": 0.7867,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9115442278860569,
+      "grad_norm": 0.0697904135176372,
+      "learning_rate": 4.898037575028657e-06,
+      "loss": 0.7006,
+      "step": 1558
+    },
+    {
+      "epoch": 0.912129301203057,
+      "grad_norm": 0.03992583542856418,
+      "learning_rate": 4.894760307715634e-06,
+      "loss": 0.7847,
+      "step": 1559
+    },
+    {
+      "epoch": 0.912714374520057,
+      "grad_norm": 0.04168078061737128,
+      "learning_rate": 4.891482622427473e-06,
+      "loss": 0.7257,
+      "step": 1560
+    },
+    {
+      "epoch": 0.9132994478370571,
+      "grad_norm": 0.03886703909354667,
+      "learning_rate": 4.888204521933178e-06,
+      "loss": 0.7405,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9138845211540572,
+      "grad_norm": 0.038571266137394804,
+      "learning_rate": 4.8849260090020916e-06,
+      "loss": 0.7847,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9144695944710571,
+      "grad_norm": 0.039942483875795645,
+      "learning_rate": 4.881647086403918e-06,
+      "loss": 0.7541,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9150546677880572,
+      "grad_norm": 0.04050391982359178,
+      "learning_rate": 4.878367756908698e-06,
+      "loss": 0.8032,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9156397411050572,
+      "grad_norm": 0.06213525861224018,
+      "learning_rate": 4.875088023286823e-06,
+      "loss": 0.7917,
+      "step": 1565
+    },
+    {
+      "epoch": 0.9162248144220573,
+      "grad_norm": 0.04040488630209474,
+      "learning_rate": 4.871807888309021e-06,
+      "loss": 0.7906,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9168098877390572,
+      "grad_norm": 0.04159089864338253,
+      "learning_rate": 4.868527354746361e-06,
+      "loss": 0.8193,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9173949610560573,
+      "grad_norm": 0.06139536834412755,
+      "learning_rate": 4.86524642537025e-06,
+      "loss": 0.7674,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9179800343730574,
+      "grad_norm": 0.03947767749509195,
+      "learning_rate": 4.861965102952426e-06,
+      "loss": 0.8447,
+      "step": 1569
+    },
+    {
+      "epoch": 0.9185651076900574,
+      "grad_norm": 0.043695480643704855,
+      "learning_rate": 4.858683390264963e-06,
+      "loss": 0.87,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9191501810070575,
+      "grad_norm": 0.03962500281019953,
+      "learning_rate": 4.855401290080259e-06,
+      "loss": 0.7799,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9197352543240574,
+      "grad_norm": 0.03883602861594149,
+      "learning_rate": 4.8521188051710465e-06,
+      "loss": 0.7198,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9203203276410575,
+      "grad_norm": 0.038996303115135654,
+      "learning_rate": 4.84883593831038e-06,
+      "loss": 0.8369,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9209054009580575,
+      "grad_norm": 0.036243906078929175,
+      "learning_rate": 4.845552692271633e-06,
+      "loss": 0.7819,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9214904742750576,
+      "grad_norm": 0.03598026113530175,
+      "learning_rate": 4.8422690698285055e-06,
+      "loss": 0.7848,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9220755475920577,
+      "grad_norm": 0.03927177230794752,
+      "learning_rate": 4.83898507375501e-06,
+      "loss": 0.8389,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9226606209090576,
+      "grad_norm": 0.0385535598643195,
+      "learning_rate": 4.83570070682548e-06,
+      "loss": 0.8231,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9232456942260577,
+      "grad_norm": 0.03748485100626454,
+      "learning_rate": 4.832415971814558e-06,
+      "loss": 0.7769,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9238307675430577,
+      "grad_norm": 0.04406376389798333,
+      "learning_rate": 4.829130871497198e-06,
+      "loss": 0.8501,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9244158408600578,
+      "grad_norm": 0.04181053014039125,
+      "learning_rate": 4.825845408648666e-06,
+      "loss": 0.7151,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9250009141770578,
+      "grad_norm": 0.04012827948350568,
+      "learning_rate": 4.82255958604453e-06,
+      "loss": 0.7475,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9255859874940578,
+      "grad_norm": 0.04066339251456573,
+      "learning_rate": 4.819273406460664e-06,
+      "loss": 0.7764,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9261710608110579,
+      "grad_norm": 0.03654201108348175,
+      "learning_rate": 4.815986872673245e-06,
+      "loss": 0.7002,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9267561341280579,
+      "grad_norm": 0.05574027310447493,
+      "learning_rate": 4.812699987458744e-06,
+      "loss": 0.7396,
+      "step": 1584
+    },
+    {
+      "epoch": 0.927341207445058,
+      "grad_norm": 0.043906342122207766,
+      "learning_rate": 4.809412753593935e-06,
+      "loss": 0.842,
+      "step": 1585
+    },
+    {
+      "epoch": 0.927926280762058,
+      "grad_norm": 0.03608562466923883,
+      "learning_rate": 4.806125173855884e-06,
+      "loss": 0.879,
+      "step": 1586
+    },
+    {
+      "epoch": 0.928511354079058,
+      "grad_norm": 0.053877179770256106,
+      "learning_rate": 4.802837251021948e-06,
+      "loss": 0.7109,
+      "step": 1587
+    },
+    {
+      "epoch": 0.929096427396058,
+      "grad_norm": 0.044680943459160405,
+      "learning_rate": 4.799548987869778e-06,
+      "loss": 0.7893,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9296815007130581,
+      "grad_norm": 0.051152646979938876,
+      "learning_rate": 4.7962603871773066e-06,
+      "loss": 0.8483,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9302665740300582,
+      "grad_norm": 0.03887470736846696,
+      "learning_rate": 4.7929714517227524e-06,
+      "loss": 0.7322,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9308516473470582,
+      "grad_norm": 0.037211850030407195,
+      "learning_rate": 4.789682184284625e-06,
+      "loss": 0.7299,
+      "step": 1591
+    },
+    {
+      "epoch": 0.9314367206640582,
+      "grad_norm": 0.037429526440719174,
+      "learning_rate": 4.786392587641704e-06,
+      "loss": 0.8137,
+      "step": 1592
+    },
+    {
+      "epoch": 0.9320217939810582,
+      "grad_norm": 0.16993450311970792,
+      "learning_rate": 4.783102664573052e-06,
+      "loss": 0.7883,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9326068672980583,
+      "grad_norm": 0.04297358226397087,
+      "learning_rate": 4.779812417858008e-06,
+      "loss": 0.8226,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9331919406150583,
+      "grad_norm": 0.03667744575467755,
+      "learning_rate": 4.776521850276182e-06,
+      "loss": 0.7385,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9337770139320584,
+      "grad_norm": 0.03913765169009846,
+      "learning_rate": 4.7732309646074585e-06,
+      "loss": 0.8733,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9343620872490584,
+      "grad_norm": 0.03736629949079511,
+      "learning_rate": 4.769939763631986e-06,
+      "loss": 0.7932,
+      "step": 1597
+    },
+    {
+      "epoch": 0.9349471605660584,
+      "grad_norm": 0.037356612963331204,
+      "learning_rate": 4.766648250130183e-06,
+      "loss": 0.8071,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9355322338830585,
+      "grad_norm": 0.04829265429843901,
+      "learning_rate": 4.76335642688273e-06,
+      "loss": 0.7536,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9361173072000585,
+      "grad_norm": 0.03663643287071603,
+      "learning_rate": 4.760064296670573e-06,
+      "loss": 0.7419,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9367023805170586,
+      "grad_norm": 0.03851622839931266,
+      "learning_rate": 4.756771862274911e-06,
+      "loss": 0.7391,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9372874538340585,
+      "grad_norm": 0.058935243933442134,
+      "learning_rate": 4.753479126477207e-06,
+      "loss": 0.8036,
+      "step": 1602
+    },
+    {
+      "epoch": 0.9378725271510586,
+      "grad_norm": 0.05643207099843358,
+      "learning_rate": 4.750186092059171e-06,
+      "loss": 0.7764,
+      "step": 1603
+    },
+    {
+      "epoch": 0.9384576004680587,
+      "grad_norm": 0.038227121653592495,
+      "learning_rate": 4.746892761802771e-06,
+      "loss": 0.6986,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9390426737850587,
+      "grad_norm": 0.041109384181244574,
+      "learning_rate": 4.743599138490224e-06,
+      "loss": 0.6453,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9396277471020588,
+      "grad_norm": 0.04596157162267912,
+      "learning_rate": 4.740305224903993e-06,
+      "loss": 0.8228,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9402128204190587,
+      "grad_norm": 0.04749314030842733,
+      "learning_rate": 4.73701102382679e-06,
+      "loss": 0.8627,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9407978937360588,
+      "grad_norm": 0.04073293983059644,
+      "learning_rate": 4.733716538041561e-06,
+      "loss": 0.7932,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9413829670530588,
+      "grad_norm": 0.03695351421055921,
+      "learning_rate": 4.730421770331503e-06,
+      "loss": 0.8384,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9419680403700589,
+      "grad_norm": 0.041338479689365416,
+      "learning_rate": 4.727126723480046e-06,
+      "loss": 0.7826,
+      "step": 1610
+    },
+    {
+      "epoch": 0.942553113687059,
+      "grad_norm": 0.039567218912003696,
+      "learning_rate": 4.723831400270854e-06,
+      "loss": 0.71,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9431381870040589,
+      "grad_norm": 0.038765181273196586,
+      "learning_rate": 4.7205358034878295e-06,
+      "loss": 0.7528,
+      "step": 1612
+    },
+    {
+      "epoch": 0.943723260321059,
+      "grad_norm": 0.03857144409516923,
+      "learning_rate": 4.717239935915105e-06,
+      "loss": 0.8085,
+      "step": 1613
+    },
+    {
+      "epoch": 0.944308333638059,
+      "grad_norm": 0.050546795531220844,
+      "learning_rate": 4.713943800337034e-06,
+      "loss": 0.814,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9448934069550591,
+      "grad_norm": 0.03877844831137139,
+      "learning_rate": 4.710647399538209e-06,
+      "loss": 0.8016,
+      "step": 1615
+    },
+    {
+      "epoch": 0.945478480272059,
+      "grad_norm": 0.04286623826991479,
+      "learning_rate": 4.707350736303439e-06,
+      "loss": 0.7046,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9460635535890591,
+      "grad_norm": 0.041695250387251796,
+      "learning_rate": 4.704053813417753e-06,
+      "loss": 0.7558,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9466486269060592,
+      "grad_norm": 0.03686599980639858,
+      "learning_rate": 4.700756633666404e-06,
+      "loss": 0.7133,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9472337002230592,
+      "grad_norm": 0.05197679285294168,
+      "learning_rate": 4.6974591998348605e-06,
+      "loss": 0.7537,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9478187735400593,
+      "grad_norm": 0.04448660325795372,
+      "learning_rate": 4.694161514708806e-06,
+      "loss": 0.7808,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9484038468570593,
+      "grad_norm": 0.04186213228118794,
+      "learning_rate": 4.690863581074137e-06,
+      "loss": 0.7767,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9489889201740593,
+      "grad_norm": 0.0400854810264849,
+      "learning_rate": 4.687565401716953e-06,
+      "loss": 0.7603,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9495739934910593,
+      "grad_norm": 0.03701384913843474,
+      "learning_rate": 4.684266979423573e-06,
+      "loss": 0.633,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9501590668080594,
+      "grad_norm": 0.04591450855041653,
+      "learning_rate": 4.680968316980514e-06,
+      "loss": 0.7848,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9507441401250594,
+      "grad_norm": 0.04476228996175277,
+      "learning_rate": 4.677669417174493e-06,
+      "loss": 0.7866,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9513292134420595,
+      "grad_norm": 0.03707636228363127,
+      "learning_rate": 4.674370282792435e-06,
+      "loss": 0.7799,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9519142867590595,
+      "grad_norm": 0.03969252104571136,
+      "learning_rate": 4.671070916621458e-06,
+      "loss": 0.7713,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9524993600760595,
+      "grad_norm": 0.07811310064791632,
+      "learning_rate": 4.66777132144888e-06,
+      "loss": 0.72,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9530844333930596,
+      "grad_norm": 0.03719654474999105,
+      "learning_rate": 4.6644715000622065e-06,
+      "loss": 0.7672,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9536695067100596,
+      "grad_norm": 0.041072520733934526,
+      "learning_rate": 4.6611714552491385e-06,
+      "loss": 0.7603,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9542545800270597,
+      "grad_norm": 0.07853658068992475,
+      "learning_rate": 4.657871189797565e-06,
+      "loss": 0.8341,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9548396533440596,
+      "grad_norm": 0.0388996416456516,
+      "learning_rate": 4.654570706495562e-06,
+      "loss": 0.6744,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9554247266610597,
+      "grad_norm": 0.04381785661377394,
+      "learning_rate": 4.651270008131388e-06,
+      "loss": 0.7378,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9560097999780598,
+      "grad_norm": 0.04201885634367334,
+      "learning_rate": 4.647969097493483e-06,
+      "loss": 0.7474,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9565948732950598,
+      "grad_norm": 0.03726848394227385,
+      "learning_rate": 4.644667977370468e-06,
+      "loss": 0.832,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9571799466120599,
+      "grad_norm": 0.04564574185484768,
+      "learning_rate": 4.64136665055114e-06,
+      "loss": 0.754,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9577650199290598,
+      "grad_norm": 0.03922811287637288,
+      "learning_rate": 4.638065119824471e-06,
+      "loss": 0.7385,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9583500932460599,
+      "grad_norm": 0.04617849099423774,
+      "learning_rate": 4.634763387979606e-06,
+      "loss": 0.8023,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9589351665630599,
+      "grad_norm": 0.051954956199202075,
+      "learning_rate": 4.631461457805855e-06,
+      "loss": 0.731,
+      "step": 1639
+    },
+    {
+      "epoch": 0.95952023988006,
+      "grad_norm": 0.03670055716355514,
+      "learning_rate": 4.628159332092705e-06,
+      "loss": 0.7279,
+      "step": 1640
+    },
+    {
+      "epoch": 0.96010531319706,
+      "grad_norm": 0.03879162525758375,
+      "learning_rate": 4.624857013629797e-06,
+      "loss": 0.7359,
+      "step": 1641
+    },
+    {
+      "epoch": 0.96069038651406,
+      "grad_norm": 0.03884240382534952,
+      "learning_rate": 4.621554505206946e-06,
+      "loss": 0.7002,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9612754598310601,
+      "grad_norm": 0.03949437881641414,
+      "learning_rate": 4.618251809614119e-06,
+      "loss": 0.8259,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9618605331480601,
+      "grad_norm": 0.03808396641461775,
+      "learning_rate": 4.614948929641444e-06,
+      "loss": 0.7026,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9624456064650602,
+      "grad_norm": 0.040652640698166743,
+      "learning_rate": 4.611645868079205e-06,
+      "loss": 0.7399,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9630306797820601,
+      "grad_norm": 0.041448998916834554,
+      "learning_rate": 4.60834262771784e-06,
+      "loss": 0.7924,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9636157530990602,
+      "grad_norm": 0.03414951748294008,
+      "learning_rate": 4.605039211347937e-06,
+      "loss": 0.694,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9642008264160603,
+      "grad_norm": 0.042150214675898114,
+      "learning_rate": 4.601735621760234e-06,
+      "loss": 0.8276,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9647858997330603,
+      "grad_norm": 0.04420133355554904,
+      "learning_rate": 4.598431861745613e-06,
+      "loss": 0.7827,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9653709730500604,
+      "grad_norm": 0.04509650606205026,
+      "learning_rate": 4.5951279340951e-06,
+      "loss": 0.7904,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9659560463670603,
+      "grad_norm": 0.038633078744749694,
+      "learning_rate": 4.591823841599868e-06,
+      "loss": 0.7758,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9665411196840604,
+      "grad_norm": 0.038236691470263345,
+      "learning_rate": 4.588519587051222e-06,
+      "loss": 0.8076,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9671261930010604,
+      "grad_norm": 0.03695431719529444,
+      "learning_rate": 4.585215173240608e-06,
+      "loss": 0.7743,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9677112663180605,
+      "grad_norm": 0.036468585815911905,
+      "learning_rate": 4.581910602959607e-06,
+      "loss": 0.8153,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9682963396350606,
+      "grad_norm": 0.03926208587753877,
+      "learning_rate": 4.5786058789999275e-06,
+      "loss": 0.7298,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9688814129520605,
+      "grad_norm": 0.04187645286136086,
+      "learning_rate": 4.575301004153415e-06,
+      "loss": 0.7198,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9694664862690606,
+      "grad_norm": 0.0413499471995496,
+      "learning_rate": 4.571995981212035e-06,
+      "loss": 0.7575,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9700515595860606,
+      "grad_norm": 0.03415914885523872,
+      "learning_rate": 4.568690812967886e-06,
+      "loss": 0.8051,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9706366329030607,
+      "grad_norm": 0.03725068359340925,
+      "learning_rate": 4.565385502213181e-06,
+      "loss": 0.735,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9712217062200607,
+      "grad_norm": 0.03893390573113403,
+      "learning_rate": 4.562080051740261e-06,
+      "loss": 0.7236,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9718067795370607,
+      "grad_norm": 0.03672862046599132,
+      "learning_rate": 4.558774464341577e-06,
+      "loss": 0.8214,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9723918528540608,
+      "grad_norm": 0.035620814744529494,
+      "learning_rate": 4.555468742809705e-06,
+      "loss": 0.8607,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9729769261710608,
+      "grad_norm": 0.045219348615134776,
+      "learning_rate": 4.552162889937326e-06,
+      "loss": 0.793,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9735619994880609,
+      "grad_norm": 0.04120637349080747,
+      "learning_rate": 4.548856908517236e-06,
+      "loss": 0.761,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9741470728050609,
+      "grad_norm": 0.037426001969961896,
+      "learning_rate": 4.545550801342341e-06,
+      "loss": 0.7146,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9747321461220609,
+      "grad_norm": 0.04095353340190115,
+      "learning_rate": 4.542244571205649e-06,
+      "loss": 0.7821,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9753172194390609,
+      "grad_norm": 0.035000026858081075,
+      "learning_rate": 4.538938220900275e-06,
+      "loss": 0.7288,
+      "step": 1667
+    },
+    {
+      "epoch": 0.975902292756061,
+      "grad_norm": 0.03646938717887608,
+      "learning_rate": 4.535631753219433e-06,
+      "loss": 0.7056,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9764873660730611,
+      "grad_norm": 0.14152781507488307,
+      "learning_rate": 4.53232517095644e-06,
+      "loss": 0.8503,
+      "step": 1669
+    },
+    {
+      "epoch": 0.977072439390061,
+      "grad_norm": 0.040871359096417335,
+      "learning_rate": 4.529018476904708e-06,
+      "loss": 0.6784,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9776575127070611,
+      "grad_norm": 0.03608347625330109,
+      "learning_rate": 4.52571167385774e-06,
+      "loss": 0.7103,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9782425860240611,
+      "grad_norm": 0.038805523132996685,
+      "learning_rate": 4.522404764609138e-06,
+      "loss": 0.7475,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9788276593410612,
+      "grad_norm": 0.05674110098536558,
+      "learning_rate": 4.5190977519525864e-06,
+      "loss": 0.8209,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9794127326580612,
+      "grad_norm": 0.03773221772259324,
+      "learning_rate": 4.5157906386818625e-06,
+      "loss": 0.8054,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9799978059750613,
+      "grad_norm": 0.03641333096994463,
+      "learning_rate": 4.512483427590826e-06,
+      "loss": 0.8017,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9805828792920613,
+      "grad_norm": 0.03773409152086726,
+      "learning_rate": 4.509176121473418e-06,
+      "loss": 0.7771,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9811679526090613,
+      "grad_norm": 0.03675860145260396,
+      "learning_rate": 4.505868723123664e-06,
+      "loss": 0.7667,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9817530259260614,
+      "grad_norm": 0.03759372778875698,
+      "learning_rate": 4.502561235335665e-06,
+      "loss": 0.7969,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9823380992430614,
+      "grad_norm": 0.0375414007747319,
+      "learning_rate": 4.499253660903595e-06,
+      "loss": 0.7952,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9829231725600615,
+      "grad_norm": 0.03422344712918065,
+      "learning_rate": 4.495946002621706e-06,
+      "loss": 0.7594,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9835082458770614,
+      "grad_norm": 0.056821433139655465,
+      "learning_rate": 4.492638263284318e-06,
+      "loss": 0.749,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9840933191940615,
+      "grad_norm": 0.11252954598380907,
+      "learning_rate": 4.489330445685821e-06,
+      "loss": 0.7391,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9846783925110616,
+      "grad_norm": 0.042469382100373515,
+      "learning_rate": 4.486022552620667e-06,
+      "loss": 0.7923,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9852634658280616,
+      "grad_norm": 0.048834820135742855,
+      "learning_rate": 4.482714586883381e-06,
+      "loss": 0.7472,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9858485391450617,
+      "grad_norm": 0.03586075302565133,
+      "learning_rate": 4.479406551268536e-06,
+      "loss": 0.7405,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9864336124620616,
+      "grad_norm": 0.033808844435251076,
+      "learning_rate": 4.476098448570778e-06,
+      "loss": 0.7668,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9870186857790617,
+      "grad_norm": 0.03839203972551903,
+      "learning_rate": 4.472790281584798e-06,
+      "loss": 0.7892,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9876037590960617,
+      "grad_norm": 0.041714468819081474,
+      "learning_rate": 4.469482053105348e-06,
+      "loss": 0.6825,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9881888324130618,
+      "grad_norm": 0.04639986807865863,
+      "learning_rate": 4.4661737659272335e-06,
+      "loss": 0.77,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9887739057300619,
+      "grad_norm": 0.03542864575710173,
+      "learning_rate": 4.462865422845301e-06,
+      "loss": 0.7578,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9893589790470618,
+      "grad_norm": 0.03475434478064272,
+      "learning_rate": 4.459557026654453e-06,
+      "loss": 0.7783,
+      "step": 1691
+    },
+    {
+      "epoch": 0.9899440523640619,
+      "grad_norm": 0.15800557603413382,
+      "learning_rate": 4.456248580149634e-06,
+      "loss": 0.8381,
+      "step": 1692
+    },
+    {
+      "epoch": 0.9905291256810619,
+      "grad_norm": 0.03764241378382179,
+      "learning_rate": 4.452940086125827e-06,
+      "loss": 0.7773,
+      "step": 1693
+    },
+    {
+      "epoch": 0.991114198998062,
+      "grad_norm": 0.036921463016592035,
+      "learning_rate": 4.449631547378062e-06,
+      "loss": 0.7178,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9916992723150619,
+      "grad_norm": 0.04206493996925596,
+      "learning_rate": 4.446322966701402e-06,
+      "loss": 0.8461,
+      "step": 1695
+    },
+    {
+      "epoch": 0.992284345632062,
+      "grad_norm": 0.03684149770322195,
+      "learning_rate": 4.443014346890949e-06,
+      "loss": 0.7421,
+      "step": 1696
+    },
+    {
+      "epoch": 0.992869418949062,
+      "grad_norm": 0.03706235629788984,
+      "learning_rate": 4.439705690741832e-06,
+      "loss": 0.747,
+      "step": 1697
+    },
+    {
+      "epoch": 0.9934544922660621,
+      "grad_norm": 0.03593179042253731,
+      "learning_rate": 4.436397001049217e-06,
+      "loss": 0.7973,
+      "step": 1698
+    },
+    {
+      "epoch": 0.9940395655830622,
+      "grad_norm": 0.03758870754317747,
+      "learning_rate": 4.4330882806082965e-06,
+      "loss": 0.752,
+      "step": 1699
+    },
+    {
+      "epoch": 0.9946246389000621,
+      "grad_norm": 0.039890034089472524,
+      "learning_rate": 4.429779532214288e-06,
+      "loss": 0.8217,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9952097122170622,
+      "grad_norm": 0.038591012254334464,
+      "learning_rate": 4.4264707586624305e-06,
+      "loss": 0.7909,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9957947855340622,
+      "grad_norm": 0.037417734290621826,
+      "learning_rate": 4.4231619627479895e-06,
+      "loss": 0.8227,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9963798588510623,
+      "grad_norm": 0.036754661863113706,
+      "learning_rate": 4.419853147266248e-06,
+      "loss": 0.8273,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9969649321680623,
+      "grad_norm": 0.03545112711834771,
+      "learning_rate": 4.4165443150125e-06,
+      "loss": 0.7733,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9975500054850623,
+      "grad_norm": 0.0453520557159223,
+      "learning_rate": 4.413235468782063e-06,
+      "loss": 0.7291,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9981350788020624,
+      "grad_norm": 0.03434680179451016,
+      "learning_rate": 4.409926611370255e-06,
+      "loss": 0.7253,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9987201521190624,
+      "grad_norm": 0.03386357517901744,
+      "learning_rate": 4.406617745572415e-06,
+      "loss": 0.7843,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9993052254360625,
+      "grad_norm": 0.03935904245669636,
+      "learning_rate": 4.403308874183881e-06,
+      "loss": 0.7553,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9998902987530625,
+      "grad_norm": 0.03529453299196854,
+      "learning_rate": 4.4e-06,
+      "loss": 0.8177,
+      "step": 1709
+    },
+    {
+      "epoch": 1.0003291037408126,
+      "grad_norm": 0.0618709966791903,
+      "learning_rate": 4.3966911258161184e-06,
+      "loss": 1.2009,
+      "step": 1710
+    },
+    {
+      "epoch": 1.0009141770578125,
+      "grad_norm": 0.05054959075407372,
+      "learning_rate": 4.393382254427584e-06,
+      "loss": 0.7522,
+      "step": 1711
+    },
+    {
+      "epoch": 1.0014992503748126,
+      "grad_norm": 0.046546989029328134,
+      "learning_rate": 4.390073388629746e-06,
+      "loss": 0.6742,
+      "step": 1712
+    },
+    {
+      "epoch": 1.0020843236918127,
+      "grad_norm": 0.0394482669118922,
+      "learning_rate": 4.386764531217939e-06,
+      "loss": 0.6234,
+      "step": 1713
+    },
+    {
+      "epoch": 1.0026693970088127,
+      "grad_norm": 0.04036631292187251,
+      "learning_rate": 4.3834556849875e-06,
+      "loss": 0.5347,
+      "step": 1714
+    },
+    {
+      "epoch": 1.0032544703258126,
+      "grad_norm": 0.0683784483426072,
+      "learning_rate": 4.380146852733752e-06,
+      "loss": 0.6644,
+      "step": 1715
+    },
+    {
+      "epoch": 1.0038395436428127,
+      "grad_norm": 0.04024887608738653,
+      "learning_rate": 4.376838037252009e-06,
+      "loss": 0.5429,
+      "step": 1716
+    },
+    {
+      "epoch": 1.0044246169598128,
+      "grad_norm": 0.04001998958059999,
+      "learning_rate": 4.37352924133757e-06,
+      "loss": 0.6346,
+      "step": 1717
+    },
+    {
+      "epoch": 1.0050096902768129,
+      "grad_norm": 0.03981740214221622,
+      "learning_rate": 4.370220467785714e-06,
+      "loss": 0.6179,
+      "step": 1718
+    },
+    {
+      "epoch": 1.005594763593813,
+      "grad_norm": 0.04221085502304135,
+      "learning_rate": 4.366911719391705e-06,
+      "loss": 0.6728,
+      "step": 1719
+    },
+    {
+      "epoch": 1.0061798369108128,
+      "grad_norm": 0.041008201601693474,
+      "learning_rate": 4.363602998950783e-06,
+      "loss": 0.664,
+      "step": 1720
+    },
+    {
+      "epoch": 1.006764910227813,
+      "grad_norm": 0.04267520538484461,
+      "learning_rate": 4.360294309258169e-06,
+      "loss": 0.5861,
+      "step": 1721
+    },
+    {
+      "epoch": 1.007349983544813,
+      "grad_norm": 0.0380878408560243,
+      "learning_rate": 4.3569856531090524e-06,
+      "loss": 0.6581,
+      "step": 1722
+    },
+    {
+      "epoch": 1.007935056861813,
+      "grad_norm": 0.044581500184632424,
+      "learning_rate": 4.353677033298599e-06,
+      "loss": 0.6328,
+      "step": 1723
+    },
+    {
+      "epoch": 1.008520130178813,
+      "grad_norm": 0.04959235942274387,
+      "learning_rate": 4.350368452621938e-06,
+      "loss": 0.7268,
+      "step": 1724
+    },
+    {
+      "epoch": 1.009105203495813,
+      "grad_norm": 0.2678728322882004,
+      "learning_rate": 4.347059913874174e-06,
+      "loss": 0.5815,
+      "step": 1725
+    },
+    {
+      "epoch": 1.009690276812813,
+      "grad_norm": 0.04227602325404652,
+      "learning_rate": 4.343751419850368e-06,
+      "loss": 0.6351,
+      "step": 1726
+    },
+    {
+      "epoch": 1.0102753501298132,
+      "grad_norm": 0.04538776787193514,
+      "learning_rate": 4.340442973345546e-06,
+      "loss": 0.5471,
+      "step": 1727
+    },
+    {
+      "epoch": 1.0108604234468133,
+      "grad_norm": 0.06267951676327517,
+      "learning_rate": 4.337134577154699e-06,
+      "loss": 0.6463,
+      "step": 1728
+    },
+    {
+      "epoch": 1.0114454967638131,
+      "grad_norm": 0.04037151798930933,
+      "learning_rate": 4.333826234072769e-06,
+      "loss": 0.6353,
+      "step": 1729
+    },
+    {
+      "epoch": 1.0120305700808132,
+      "grad_norm": 0.03689888346419047,
+      "learning_rate": 4.330517946894651e-06,
+      "loss": 0.5699,
+      "step": 1730
+    },
+    {
+      "epoch": 1.0126156433978133,
+      "grad_norm": 0.03497818177140738,
+      "learning_rate": 4.3272097184152026e-06,
+      "loss": 0.5889,
+      "step": 1731
+    },
+    {
+      "epoch": 1.0132007167148134,
+      "grad_norm": 0.11544161330826186,
+      "learning_rate": 4.3239015514292225e-06,
+      "loss": 0.6106,
+      "step": 1732
+    },
+    {
+      "epoch": 1.0137857900318135,
+      "grad_norm": 0.04017346774193498,
+      "learning_rate": 4.320593448731463e-06,
+      "loss": 0.6093,
+      "step": 1733
+    },
+    {
+      "epoch": 1.0143708633488133,
+      "grad_norm": 0.06993118467504732,
+      "learning_rate": 4.317285413116621e-06,
+      "loss": 0.6178,
+      "step": 1734
+    },
+    {
+      "epoch": 1.0149559366658134,
+      "grad_norm": 0.042065264751811854,
+      "learning_rate": 4.313977447379332e-06,
+      "loss": 0.6502,
+      "step": 1735
+    },
+    {
+      "epoch": 1.0155410099828135,
+      "grad_norm": 0.041084510994879864,
+      "learning_rate": 4.310669554314179e-06,
+      "loss": 0.5617,
+      "step": 1736
+    },
+    {
+      "epoch": 1.0161260832998136,
+      "grad_norm": 0.049710466173704275,
+      "learning_rate": 4.307361736715682e-06,
+      "loss": 0.6016,
+      "step": 1737
+    },
+    {
+      "epoch": 1.0167111566168134,
+      "grad_norm": 0.03634779224936381,
+      "learning_rate": 4.304053997378294e-06,
+      "loss": 0.6092,
+      "step": 1738
+    },
+    {
+      "epoch": 1.0172962299338135,
+      "grad_norm": 0.05093471107840585,
+      "learning_rate": 4.300746339096404e-06,
+      "loss": 0.6604,
+      "step": 1739
+    },
+    {
+      "epoch": 1.0178813032508136,
+      "grad_norm": 0.052249007194382246,
+      "learning_rate": 4.2974387646643366e-06,
+      "loss": 0.6268,
+      "step": 1740
+    },
+    {
+      "epoch": 1.0184663765678137,
+      "grad_norm": 0.04168400074209779,
+      "learning_rate": 4.294131276876336e-06,
+      "loss": 0.6615,
+      "step": 1741
+    },
+    {
+      "epoch": 1.0190514498848138,
+      "grad_norm": 0.10284332944913842,
+      "learning_rate": 4.290823878526582e-06,
+      "loss": 0.6481,
+      "step": 1742
+    },
+    {
+      "epoch": 1.0196365232018136,
+      "grad_norm": 0.06107877656234674,
+      "learning_rate": 4.287516572409174e-06,
+      "loss": 0.5273,
+      "step": 1743
+    },
+    {
+      "epoch": 1.0202215965188137,
+      "grad_norm": 0.039593505867561385,
+      "learning_rate": 4.284209361318137e-06,
+      "loss": 0.5541,
+      "step": 1744
+    },
+    {
+      "epoch": 1.0208066698358138,
+      "grad_norm": 0.03815708858382624,
+      "learning_rate": 4.280902248047414e-06,
+      "loss": 0.6568,
+      "step": 1745
+    },
+    {
+      "epoch": 1.021391743152814,
+      "grad_norm": 0.045350052500293725,
+      "learning_rate": 4.277595235390863e-06,
+      "loss": 0.4949,
+      "step": 1746
+    },
+    {
+      "epoch": 1.021976816469814,
+      "grad_norm": 0.062366644080380174,
+      "learning_rate": 4.27428832614226e-06,
+      "loss": 0.6412,
+      "step": 1747
+    },
+    {
+      "epoch": 1.0225618897868138,
+      "grad_norm": 0.038514364508854086,
+      "learning_rate": 4.270981523095292e-06,
+      "loss": 0.5954,
+      "step": 1748
+    },
+    {
+      "epoch": 1.023146963103814,
+      "grad_norm": 0.040934523714329187,
+      "learning_rate": 4.267674829043559e-06,
+      "loss": 0.6202,
+      "step": 1749
+    },
+    {
+      "epoch": 1.023732036420814,
+      "grad_norm": 0.04396368047960114,
+      "learning_rate": 4.2643682467805675e-06,
+      "loss": 0.5798,
+      "step": 1750
+    },
+    {
+      "epoch": 1.024317109737814,
+      "grad_norm": 0.05011201210029028,
+      "learning_rate": 4.261061779099726e-06,
+      "loss": 0.6321,
+      "step": 1751
+    },
+    {
+      "epoch": 1.024902183054814,
+      "grad_norm": 0.037782483274636854,
+      "learning_rate": 4.2577554287943515e-06,
+      "loss": 0.6632,
+      "step": 1752
+    },
+    {
+      "epoch": 1.025487256371814,
+      "grad_norm": 0.04152834895888912,
+      "learning_rate": 4.2544491986576595e-06,
+      "loss": 0.5945,
+      "step": 1753
+    },
+    {
+      "epoch": 1.0260723296888141,
+      "grad_norm": 0.05806333477040265,
+      "learning_rate": 4.2511430914827634e-06,
+      "loss": 0.6637,
+      "step": 1754
+    },
+    {
+      "epoch": 1.0266574030058142,
+      "grad_norm": 0.10117835297951212,
+      "learning_rate": 4.247837110062674e-06,
+      "loss": 0.6018,
+      "step": 1755
+    },
+    {
+      "epoch": 1.0272424763228143,
+      "grad_norm": 0.03885375577058434,
+      "learning_rate": 4.244531257190297e-06,
+      "loss": 0.5957,
+      "step": 1756
+    },
+    {
+      "epoch": 1.0278275496398142,
+      "grad_norm": 0.03754400735025208,
+      "learning_rate": 4.2412255356584226e-06,
+      "loss": 0.6496,
+      "step": 1757
+    },
+    {
+      "epoch": 1.0284126229568142,
+      "grad_norm": 0.04039354810616503,
+      "learning_rate": 4.23791994825974e-06,
+      "loss": 0.636,
+      "step": 1758
+    },
+    {
+      "epoch": 1.0289976962738143,
+      "grad_norm": 0.03960082134430345,
+      "learning_rate": 4.2346144977868195e-06,
+      "loss": 0.6303,
+      "step": 1759
+    },
+    {
+      "epoch": 1.0295827695908144,
+      "grad_norm": 0.038169466661217184,
+      "learning_rate": 4.231309187032114e-06,
+      "loss": 0.667,
+      "step": 1760
+    },
+    {
+      "epoch": 1.0301678429078145,
+      "grad_norm": 0.036760609517840784,
+      "learning_rate": 4.228004018787964e-06,
+      "loss": 0.5326,
+      "step": 1761
+    },
+    {
+      "epoch": 1.0307529162248144,
+      "grad_norm": 0.05224287627733055,
+      "learning_rate": 4.224698995846585e-06,
+      "loss": 0.5847,
+      "step": 1762
+    },
+    {
+      "epoch": 1.0313379895418144,
+      "grad_norm": 0.04647643862134824,
+      "learning_rate": 4.221394121000072e-06,
+      "loss": 0.6364,
+      "step": 1763
+    },
+    {
+      "epoch": 1.0319230628588145,
+      "grad_norm": 0.039427816196321165,
+      "learning_rate": 4.218089397040393e-06,
+      "loss": 0.6066,
+      "step": 1764
+    },
+    {
+      "epoch": 1.0325081361758146,
+      "grad_norm": 0.03480024025349239,
+      "learning_rate": 4.214784826759391e-06,
+      "loss": 0.5554,
+      "step": 1765
+    },
+    {
+      "epoch": 1.0330932094928145,
+      "grad_norm": 0.08040264519097115,
+      "learning_rate": 4.211480412948778e-06,
+      "loss": 0.5786,
+      "step": 1766
+    },
+    {
+      "epoch": 1.0336782828098146,
+      "grad_norm": 0.04884829861187939,
+      "learning_rate": 4.208176158400132e-06,
+      "loss": 0.6147,
+      "step": 1767
+    },
+    {
+      "epoch": 1.0342633561268146,
+      "grad_norm": 0.03732626269644399,
+      "learning_rate": 4.2048720659049e-06,
+      "loss": 0.6238,
+      "step": 1768
+    },
+    {
+      "epoch": 1.0348484294438147,
+      "grad_norm": 0.03742496299770894,
+      "learning_rate": 4.201568138254388e-06,
+      "loss": 0.6473,
+      "step": 1769
+    },
+    {
+      "epoch": 1.0354335027608148,
+      "grad_norm": 0.037259663998191445,
+      "learning_rate": 4.1982643782397665e-06,
+      "loss": 0.5993,
+      "step": 1770
+    },
+    {
+      "epoch": 1.0360185760778147,
+      "grad_norm": 0.03806318673198438,
+      "learning_rate": 4.194960788652061e-06,
+      "loss": 0.5744,
+      "step": 1771
+    },
+    {
+      "epoch": 1.0366036493948148,
+      "grad_norm": 0.04271776837607931,
+      "learning_rate": 4.191657372282161e-06,
+      "loss": 0.6405,
+      "step": 1772
+    },
+    {
+      "epoch": 1.0371887227118148,
+      "grad_norm": 0.1193247723794853,
+      "learning_rate": 4.188354131920795e-06,
+      "loss": 0.6305,
+      "step": 1773
+    },
+    {
+      "epoch": 1.037773796028815,
+      "grad_norm": 0.041408678944841584,
+      "learning_rate": 4.185051070358556e-06,
+      "loss": 0.672,
+      "step": 1774
+    },
+    {
+      "epoch": 1.038358869345815,
+      "grad_norm": 0.04305727303228828,
+      "learning_rate": 4.181748190385882e-06,
+      "loss": 0.574,
+      "step": 1775
+    },
+    {
+      "epoch": 1.0389439426628149,
+      "grad_norm": 0.03600148051856665,
+      "learning_rate": 4.178445494793054e-06,
+      "loss": 0.5927,
+      "step": 1776
+    },
+    {
+      "epoch": 1.039529015979815,
+      "grad_norm": 0.036257379509770316,
+      "learning_rate": 4.1751429863702024e-06,
+      "loss": 0.6351,
+      "step": 1777
+    },
+    {
+      "epoch": 1.040114089296815,
+      "grad_norm": 0.05508170928413832,
+      "learning_rate": 4.171840667907297e-06,
+      "loss": 0.6693,
+      "step": 1778
+    },
+    {
+      "epoch": 1.0406991626138151,
+      "grad_norm": 0.0359589616622991,
+      "learning_rate": 4.168538542194145e-06,
+      "loss": 0.643,
+      "step": 1779
+    },
+    {
+      "epoch": 1.041284235930815,
+      "grad_norm": 0.04574135624876753,
+      "learning_rate": 4.1652366120203955e-06,
+      "loss": 0.6592,
+      "step": 1780
+    },
+    {
+      "epoch": 1.041869309247815,
+      "grad_norm": 0.04576629116297331,
+      "learning_rate": 4.161934880175529e-06,
+      "loss": 0.6442,
+      "step": 1781
+    },
+    {
+      "epoch": 1.0424543825648152,
+      "grad_norm": 0.0390727224308255,
+      "learning_rate": 4.15863334944886e-06,
+      "loss": 0.6548,
+      "step": 1782
+    },
+    {
+      "epoch": 1.0430394558818152,
+      "grad_norm": 0.041392206592811995,
+      "learning_rate": 4.155332022629533e-06,
+      "loss": 0.6029,
+      "step": 1783
+    },
+    {
+      "epoch": 1.0436245291988153,
+      "grad_norm": 0.03762867728092406,
+      "learning_rate": 4.152030902506518e-06,
+      "loss": 0.6379,
+      "step": 1784
+    },
+    {
+      "epoch": 1.0442096025158152,
+      "grad_norm": 0.04399197262339416,
+      "learning_rate": 4.148729991868613e-06,
+      "loss": 0.6502,
+      "step": 1785
+    },
+    {
+      "epoch": 1.0447946758328153,
+      "grad_norm": 0.034378949993079645,
+      "learning_rate": 4.145429293504438e-06,
+      "loss": 0.5196,
+      "step": 1786
+    },
+    {
+      "epoch": 1.0453797491498154,
+      "grad_norm": 0.08443203067098491,
+      "learning_rate": 4.142128810202434e-06,
+      "loss": 0.5683,
+      "step": 1787
+    },
+    {
+      "epoch": 1.0459648224668154,
+      "grad_norm": 0.0379632191291091,
+      "learning_rate": 4.138828544750861e-06,
+      "loss": 0.592,
+      "step": 1788
+    },
+    {
+      "epoch": 1.0465498957838153,
+      "grad_norm": 0.038094993376842946,
+      "learning_rate": 4.135528499937794e-06,
+      "loss": 0.6709,
+      "step": 1789
+    },
+    {
+      "epoch": 1.0471349691008154,
+      "grad_norm": 0.04377752430482512,
+      "learning_rate": 4.13222867855112e-06,
+      "loss": 0.5478,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0477200424178155,
+      "grad_norm": 0.040127517533422315,
+      "learning_rate": 4.128929083378542e-06,
+      "loss": 0.6338,
+      "step": 1791
+    },
+    {
+      "epoch": 1.0483051157348156,
+      "grad_norm": 0.041540955383264945,
+      "learning_rate": 4.125629717207565e-06,
+      "loss": 0.6282,
+      "step": 1792
+    },
+    {
+      "epoch": 1.0488901890518156,
+      "grad_norm": 0.04313573791852679,
+      "learning_rate": 4.122330582825507e-06,
+      "loss": 0.5929,
+      "step": 1793
+    },
+    {
+      "epoch": 1.0494752623688155,
+      "grad_norm": 0.04920349769302968,
+      "learning_rate": 4.119031683019487e-06,
+      "loss": 0.642,
+      "step": 1794
+    },
+    {
+      "epoch": 1.0500603356858156,
+      "grad_norm": 0.04118717172286108,
+      "learning_rate": 4.115733020576427e-06,
+      "loss": 0.6701,
+      "step": 1795
+    },
+    {
+      "epoch": 1.0506454090028157,
+      "grad_norm": 0.03762242324561743,
+      "learning_rate": 4.112434598283046e-06,
+      "loss": 0.5557,
+      "step": 1796
+    },
+    {
+      "epoch": 1.0512304823198158,
+      "grad_norm": 0.04863139641918998,
+      "learning_rate": 4.1091364189258645e-06,
+      "loss": 0.6696,
+      "step": 1797
+    },
+    {
+      "epoch": 1.0518155556368158,
+      "grad_norm": 0.03800325798941253,
+      "learning_rate": 4.105838485291194e-06,
+      "loss": 0.5444,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0524006289538157,
+      "grad_norm": 0.04250406776849721,
+      "learning_rate": 4.102540800165137e-06,
+      "loss": 0.6006,
+      "step": 1799
+    },
+    {
+      "epoch": 1.0529857022708158,
+      "grad_norm": 0.03566004409727574,
+      "learning_rate": 4.099243366333598e-06,
+      "loss": 0.5474,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0535707755878159,
+      "grad_norm": 0.03920829695942249,
+      "learning_rate": 4.0959461865822476e-06,
+      "loss": 0.6428,
+      "step": 1801
+    },
+    {
+      "epoch": 1.054155848904816,
+      "grad_norm": 0.04622201693753423,
+      "learning_rate": 4.092649263696562e-06,
+      "loss": 0.5943,
+      "step": 1802
+    },
+    {
+      "epoch": 1.054740922221816,
+      "grad_norm": 0.05615949708259005,
+      "learning_rate": 4.089352600461791e-06,
+      "loss": 0.5458,
+      "step": 1803
+    },
+    {
+      "epoch": 1.055325995538816,
+      "grad_norm": 0.04045765925529075,
+      "learning_rate": 4.086056199662964e-06,
+      "loss": 0.635,
+      "step": 1804
+    },
+    {
+      "epoch": 1.055911068855816,
+      "grad_norm": 0.04630492862429359,
+      "learning_rate": 4.0827600640848965e-06,
+      "loss": 0.6339,
+      "step": 1805
+    },
+    {
+      "epoch": 1.056496142172816,
+      "grad_norm": 0.04703035260473936,
+      "learning_rate": 4.079464196512169e-06,
+      "loss": 0.6374,
+      "step": 1806
+    },
+    {
+      "epoch": 1.0570812154898162,
+      "grad_norm": 0.03739049283716097,
+      "learning_rate": 4.0761685997291465e-06,
+      "loss": 0.6206,
+      "step": 1807
+    },
+    {
+      "epoch": 1.057666288806816,
+      "grad_norm": 0.036569806915412575,
+      "learning_rate": 4.072873276519955e-06,
+      "loss": 0.5138,
+      "step": 1808
+    },
+    {
+      "epoch": 1.058251362123816,
+      "grad_norm": 0.04051229683477904,
+      "learning_rate": 4.069578229668497e-06,
+      "loss": 0.6494,
+      "step": 1809
+    },
+    {
+      "epoch": 1.0588364354408162,
+      "grad_norm": 0.035018447172935556,
+      "learning_rate": 4.0662834619584385e-06,
+      "loss": 0.5486,
+      "step": 1810
+    },
+    {
+      "epoch": 1.0594215087578163,
+      "grad_norm": 0.04259663408302132,
+      "learning_rate": 4.062988976173212e-06,
+      "loss": 0.5984,
+      "step": 1811
+    },
+    {
+      "epoch": 1.0600065820748164,
+      "grad_norm": 0.04168057194830636,
+      "learning_rate": 4.059694775096007e-06,
+      "loss": 0.6659,
+      "step": 1812
+    },
+    {
+      "epoch": 1.0605916553918162,
+      "grad_norm": 0.03900712312055412,
+      "learning_rate": 4.056400861509775e-06,
+      "loss": 0.6419,
+      "step": 1813
+    },
+    {
+      "epoch": 1.0611767287088163,
+      "grad_norm": 0.034201541655564785,
+      "learning_rate": 4.053107238197229e-06,
+      "loss": 0.5947,
+      "step": 1814
+    },
+    {
+      "epoch": 1.0617618020258164,
+      "grad_norm": 0.03574462040512478,
+      "learning_rate": 4.0498139079408295e-06,
+      "loss": 0.6378,
+      "step": 1815
+    },
+    {
+      "epoch": 1.0623468753428165,
+      "grad_norm": 0.038067071655048315,
+      "learning_rate": 4.046520873522795e-06,
+      "loss": 0.5626,
+      "step": 1816
+    },
+    {
+      "epoch": 1.0629319486598163,
+      "grad_norm": 0.04383692370100588,
+      "learning_rate": 4.043228137725088e-06,
+      "loss": 0.6189,
+      "step": 1817
+    },
+    {
+      "epoch": 1.0635170219768164,
+      "grad_norm": 0.03887409023229734,
+      "learning_rate": 4.039935703329427e-06,
+      "loss": 0.6488,
+      "step": 1818
+    },
+    {
+      "epoch": 1.0641020952938165,
+      "grad_norm": 0.037631106644317916,
+      "learning_rate": 4.03664357311727e-06,
+      "loss": 0.5685,
+      "step": 1819
+    },
+    {
+      "epoch": 1.0646871686108166,
+      "grad_norm": 0.03915394810150019,
+      "learning_rate": 4.033351749869816e-06,
+      "loss": 0.5143,
+      "step": 1820
+    },
+    {
+      "epoch": 1.0652722419278167,
+      "grad_norm": 0.23381999090629155,
+      "learning_rate": 4.030060236368015e-06,
+      "loss": 0.748,
+      "step": 1821
+    },
+    {
+      "epoch": 1.0658573152448165,
+      "grad_norm": 0.03673527214266167,
+      "learning_rate": 4.026769035392542e-06,
+      "loss": 0.6073,
+      "step": 1822
+    },
+    {
+      "epoch": 1.0664423885618166,
+      "grad_norm": 0.04415310297747642,
+      "learning_rate": 4.0234781497238175e-06,
+      "loss": 0.5535,
+      "step": 1823
+    },
+    {
+      "epoch": 1.0670274618788167,
+      "grad_norm": 0.03714005735947303,
+      "learning_rate": 4.020187582141993e-06,
+      "loss": 0.5971,
+      "step": 1824
+    },
+    {
+      "epoch": 1.0676125351958168,
+      "grad_norm": 0.03766490691087146,
+      "learning_rate": 4.016897335426948e-06,
+      "loss": 0.6188,
+      "step": 1825
+    },
+    {
+      "epoch": 1.0681976085128169,
+      "grad_norm": 0.04033316810065516,
+      "learning_rate": 4.013607412358296e-06,
+      "loss": 0.6111,
+      "step": 1826
+    },
+    {
+      "epoch": 1.0687826818298167,
+      "grad_norm": 0.05353718654581134,
+      "learning_rate": 4.010317815715376e-06,
+      "loss": 0.6179,
+      "step": 1827
+    },
+    {
+      "epoch": 1.0693677551468168,
+      "grad_norm": 0.046608435545504985,
+      "learning_rate": 4.007028548277247e-06,
+      "loss": 0.6532,
+      "step": 1828
+    },
+    {
+      "epoch": 1.069952828463817,
+      "grad_norm": 0.03697488633055468,
+      "learning_rate": 4.003739612822695e-06,
+      "loss": 0.5901,
+      "step": 1829
+    },
+    {
+      "epoch": 1.070537901780817,
+      "grad_norm": 0.036722760608521324,
+      "learning_rate": 4.000451012130223e-06,
+      "loss": 0.605,
+      "step": 1830
+    },
+    {
+      "epoch": 1.0711229750978168,
+      "grad_norm": 0.038328830780971296,
+      "learning_rate": 3.997162748978052e-06,
+      "loss": 0.6481,
+      "step": 1831
+    },
+    {
+      "epoch": 1.071708048414817,
+      "grad_norm": 0.03873004486583279,
+      "learning_rate": 3.993874826144117e-06,
+      "loss": 0.636,
+      "step": 1832
+    },
+    {
+      "epoch": 1.072293121731817,
+      "grad_norm": 0.03337836188489294,
+      "learning_rate": 3.990587246406065e-06,
+      "loss": 0.5294,
+      "step": 1833
+    },
+    {
+      "epoch": 1.072878195048817,
+      "grad_norm": 0.039178449513301655,
+      "learning_rate": 3.987300012541256e-06,
+      "loss": 0.6828,
+      "step": 1834
+    },
+    {
+      "epoch": 1.0734632683658172,
+      "grad_norm": 0.03482979527366474,
+      "learning_rate": 3.984013127326756e-06,
+      "loss": 0.6274,
+      "step": 1835
+    },
+    {
+      "epoch": 1.074048341682817,
+      "grad_norm": 0.03869433482507194,
+      "learning_rate": 3.980726593539336e-06,
+      "loss": 0.6027,
+      "step": 1836
+    },
+    {
+      "epoch": 1.0746334149998171,
+      "grad_norm": 0.04937610492549535,
+      "learning_rate": 3.9774404139554695e-06,
+      "loss": 0.6049,
+      "step": 1837
+    },
+    {
+      "epoch": 1.0752184883168172,
+      "grad_norm": 0.04799441643569765,
+      "learning_rate": 3.974154591351334e-06,
+      "loss": 0.6649,
+      "step": 1838
+    },
+    {
+      "epoch": 1.0758035616338173,
+      "grad_norm": 0.05456302001307194,
+      "learning_rate": 3.9708691285028015e-06,
+      "loss": 0.5912,
+      "step": 1839
+    },
+    {
+      "epoch": 1.0763886349508174,
+      "grad_norm": 0.03519964787380995,
+      "learning_rate": 3.967584028185443e-06,
+      "loss": 0.5268,
+      "step": 1840
+    },
+    {
+      "epoch": 1.0769737082678172,
+      "grad_norm": 0.038771677450854726,
+      "learning_rate": 3.9642992931745195e-06,
+      "loss": 0.6219,
+      "step": 1841
+    },
+    {
+      "epoch": 1.0775587815848173,
+      "grad_norm": 0.03752406445751352,
+      "learning_rate": 3.961014926244989e-06,
+      "loss": 0.5636,
+      "step": 1842
+    },
+    {
+      "epoch": 1.0781438549018174,
+      "grad_norm": 0.04781643644337222,
+      "learning_rate": 3.957730930171496e-06,
+      "loss": 0.5356,
+      "step": 1843
+    },
+    {
+      "epoch": 1.0787289282188175,
+      "grad_norm": 0.03852154632487492,
+      "learning_rate": 3.954447307728367e-06,
+      "loss": 0.7087,
+      "step": 1844
+    },
+    {
+      "epoch": 1.0793140015358174,
+      "grad_norm": 0.03440089125130851,
+      "learning_rate": 3.951164061689621e-06,
+      "loss": 0.5732,
+      "step": 1845
+    },
+    {
+      "epoch": 1.0798990748528174,
+      "grad_norm": 0.10561960370834975,
+      "learning_rate": 3.947881194828953e-06,
+      "loss": 0.5871,
+      "step": 1846
+    },
+    {
+      "epoch": 1.0804841481698175,
+      "grad_norm": 0.03365711719667564,
+      "learning_rate": 3.944598709919741e-06,
+      "loss": 0.5541,
+      "step": 1847
+    },
+    {
+      "epoch": 1.0810692214868176,
+      "grad_norm": 0.04080392556587059,
+      "learning_rate": 3.941316609735038e-06,
+      "loss": 0.6567,
+      "step": 1848
+    },
+    {
+      "epoch": 1.0816542948038177,
+      "grad_norm": 0.03668175589524903,
+      "learning_rate": 3.9380348970475745e-06,
+      "loss": 0.6026,
+      "step": 1849
+    },
+    {
+      "epoch": 1.0822393681208176,
+      "grad_norm": 0.039068033523461096,
+      "learning_rate": 3.934753574629751e-06,
+      "loss": 0.6756,
+      "step": 1850
+    },
+    {
+      "epoch": 1.0828244414378176,
+      "grad_norm": 0.03600473203520014,
+      "learning_rate": 3.931472645253638e-06,
+      "loss": 0.5998,
+      "step": 1851
+    },
+    {
+      "epoch": 1.0834095147548177,
+      "grad_norm": 0.04025386130635245,
+      "learning_rate": 3.928192111690978e-06,
+      "loss": 0.5597,
+      "step": 1852
+    },
+    {
+      "epoch": 1.0839945880718178,
+      "grad_norm": 0.036383871440146785,
+      "learning_rate": 3.924911976713177e-06,
+      "loss": 0.6221,
+      "step": 1853
+    },
+    {
+      "epoch": 1.0845796613888177,
+      "grad_norm": 0.04167477133917602,
+      "learning_rate": 3.9216322430913026e-06,
+      "loss": 0.6218,
+      "step": 1854
+    },
+    {
+      "epoch": 1.0851647347058178,
+      "grad_norm": 0.04451384993527383,
+      "learning_rate": 3.9183529135960836e-06,
+      "loss": 0.6512,
+      "step": 1855
+    },
+    {
+      "epoch": 1.0857498080228178,
+      "grad_norm": 0.043002906810256446,
+      "learning_rate": 3.915073990997909e-06,
+      "loss": 0.6139,
+      "step": 1856
+    },
+    {
+      "epoch": 1.086334881339818,
+      "grad_norm": 0.04885435791726515,
+      "learning_rate": 3.911795478066823e-06,
+      "loss": 0.5525,
+      "step": 1857
+    },
+    {
+      "epoch": 1.086919954656818,
+      "grad_norm": 0.04165286964749901,
+      "learning_rate": 3.908517377572525e-06,
+      "loss": 0.5547,
+      "step": 1858
+    },
+    {
+      "epoch": 1.0875050279738179,
+      "grad_norm": 0.1274643153042942,
+      "learning_rate": 3.905239692284366e-06,
+      "loss": 0.7246,
+      "step": 1859
+    },
+    {
+      "epoch": 1.088090101290818,
+      "grad_norm": 0.03908126994536125,
+      "learning_rate": 3.9019624249713445e-06,
+      "loss": 0.5542,
+      "step": 1860
+    },
+    {
+      "epoch": 1.088675174607818,
+      "grad_norm": 0.03773075708542036,
+      "learning_rate": 3.898685578402105e-06,
+      "loss": 0.5707,
+      "step": 1861
+    },
+    {
+      "epoch": 1.0892602479248181,
+      "grad_norm": 0.036829088849046586,
+      "learning_rate": 3.89540915534494e-06,
+      "loss": 0.5886,
+      "step": 1862
+    },
+    {
+      "epoch": 1.0898453212418182,
+      "grad_norm": 0.04130199746948449,
+      "learning_rate": 3.892133158567783e-06,
+      "loss": 0.6131,
+      "step": 1863
+    },
+    {
+      "epoch": 1.090430394558818,
+      "grad_norm": 0.04746832213288144,
+      "learning_rate": 3.888857590838205e-06,
+      "loss": 0.6648,
+      "step": 1864
+    },
+    {
+      "epoch": 1.0910154678758182,
+      "grad_norm": 0.041508017388942386,
+      "learning_rate": 3.885582454923419e-06,
+      "loss": 0.6277,
+      "step": 1865
+    },
+    {
+      "epoch": 1.0916005411928182,
+      "grad_norm": 0.0409505246603534,
+      "learning_rate": 3.882307753590267e-06,
+      "loss": 0.6125,
+      "step": 1866
+    },
+    {
+      "epoch": 1.0921856145098183,
+      "grad_norm": 0.03566935377869805,
+      "learning_rate": 3.879033489605231e-06,
+      "loss": 0.5809,
+      "step": 1867
+    },
+    {
+      "epoch": 1.0927706878268184,
+      "grad_norm": 0.04208144075756873,
+      "learning_rate": 3.875759665734418e-06,
+      "loss": 0.6503,
+      "step": 1868
+    },
+    {
+      "epoch": 1.0933557611438183,
+      "grad_norm": 0.05319025544123129,
+      "learning_rate": 3.872486284743565e-06,
+      "loss": 0.6612,
+      "step": 1869
+    },
+    {
+      "epoch": 1.0939408344608184,
+      "grad_norm": 0.036973295820488555,
+      "learning_rate": 3.869213349398037e-06,
+      "loss": 0.6477,
+      "step": 1870
+    },
+    {
+      "epoch": 1.0945259077778184,
+      "grad_norm": 0.03394449055816878,
+      "learning_rate": 3.8659408624628196e-06,
+      "loss": 0.5772,
+      "step": 1871
+    },
+    {
+      "epoch": 1.0951109810948185,
+      "grad_norm": 0.036674864070854414,
+      "learning_rate": 3.8626688267025195e-06,
+      "loss": 0.6398,
+      "step": 1872
+    },
+    {
+      "epoch": 1.0956960544118184,
+      "grad_norm": 0.03929339407053973,
+      "learning_rate": 3.859397244881364e-06,
+      "loss": 0.5085,
+      "step": 1873
+    },
+    {
+      "epoch": 1.0962811277288185,
+      "grad_norm": 0.040046901855836564,
+      "learning_rate": 3.856126119763195e-06,
+      "loss": 0.6364,
+      "step": 1874
+    },
+    {
+      "epoch": 1.0968662010458186,
+      "grad_norm": 0.035486379801793984,
+      "learning_rate": 3.8528554541114725e-06,
+      "loss": 0.5641,
+      "step": 1875
+    },
+    {
+      "epoch": 1.0974512743628186,
+      "grad_norm": 0.03936545902658793,
+      "learning_rate": 3.849585250689265e-06,
+      "loss": 0.5571,
+      "step": 1876
+    },
+    {
+      "epoch": 1.0980363476798187,
+      "grad_norm": 0.03913539174688499,
+      "learning_rate": 3.846315512259249e-06,
+      "loss": 0.5749,
+      "step": 1877
+    },
+    {
+      "epoch": 1.0986214209968186,
+      "grad_norm": 0.05225621206411416,
+      "learning_rate": 3.843046241583711e-06,
+      "loss": 0.6485,
+      "step": 1878
+    },
+    {
+      "epoch": 1.0992064943138187,
+      "grad_norm": 0.038252002337340885,
+      "learning_rate": 3.839777441424543e-06,
+      "loss": 0.5985,
+      "step": 1879
+    },
+    {
+      "epoch": 1.0997915676308188,
+      "grad_norm": 0.07992087638883832,
+      "learning_rate": 3.8365091145432355e-06,
+      "loss": 0.5751,
+      "step": 1880
+    },
+    {
+      "epoch": 1.1003766409478188,
+      "grad_norm": 0.041079963315258525,
+      "learning_rate": 3.833241263700887e-06,
+      "loss": 0.5941,
+      "step": 1881
+    },
+    {
+      "epoch": 1.1009617142648187,
+      "grad_norm": 0.03572906188540022,
+      "learning_rate": 3.829973891658183e-06,
+      "loss": 0.597,
+      "step": 1882
+    },
+    {
+      "epoch": 1.1015467875818188,
+      "grad_norm": 0.03454484109641548,
+      "learning_rate": 3.8267070011754125e-06,
+      "loss": 0.522,
+      "step": 1883
+    },
+    {
+      "epoch": 1.1021318608988189,
+      "grad_norm": 0.03529581537504829,
+      "learning_rate": 3.823440595012456e-06,
+      "loss": 0.5534,
+      "step": 1884
+    },
+    {
+      "epoch": 1.102716934215819,
+      "grad_norm": 0.03860812108580398,
+      "learning_rate": 3.820174675928782e-06,
+      "loss": 0.6171,
+      "step": 1885
+    },
+    {
+      "epoch": 1.103302007532819,
+      "grad_norm": 0.03891069294549701,
+      "learning_rate": 3.816909246683452e-06,
+      "loss": 0.6313,
+      "step": 1886
+    },
+    {
+      "epoch": 1.103887080849819,
+      "grad_norm": 0.04460994600621541,
+      "learning_rate": 3.8136443100351114e-06,
+      "loss": 0.6144,
+      "step": 1887
+    },
+    {
+      "epoch": 1.104472154166819,
+      "grad_norm": 0.03793462372864508,
+      "learning_rate": 3.8103798687419875e-06,
+      "loss": 0.6043,
+      "step": 1888
+    },
+    {
+      "epoch": 1.105057227483819,
+      "grad_norm": 0.03789544727466346,
+      "learning_rate": 3.8071159255618916e-06,
+      "loss": 0.5585,
+      "step": 1889
+    },
+    {
+      "epoch": 1.1056423008008192,
+      "grad_norm": 0.03902301020371876,
+      "learning_rate": 3.803852483252215e-06,
+      "loss": 0.5883,
+      "step": 1890
+    },
+    {
+      "epoch": 1.1062273741178192,
+      "grad_norm": 0.04123331452349562,
+      "learning_rate": 3.800589544569924e-06,
+      "loss": 0.5997,
+      "step": 1891
+    },
+    {
+      "epoch": 1.106812447434819,
+      "grad_norm": 0.0384499004027348,
+      "learning_rate": 3.79732711227156e-06,
+      "loss": 0.5425,
+      "step": 1892
+    },
+    {
+      "epoch": 1.1073975207518192,
+      "grad_norm": 0.03931159943079186,
+      "learning_rate": 3.794065189113235e-06,
+      "loss": 0.6912,
+      "step": 1893
+    },
+    {
+      "epoch": 1.1079825940688193,
+      "grad_norm": 0.042235818217412,
+      "learning_rate": 3.790803777850633e-06,
+      "loss": 0.536,
+      "step": 1894
+    },
+    {
+      "epoch": 1.1085676673858194,
+      "grad_norm": 0.07854013217181985,
+      "learning_rate": 3.787542881239007e-06,
+      "loss": 0.6006,
+      "step": 1895
+    },
+    {
+      "epoch": 1.1091527407028194,
+      "grad_norm": 0.03802962654136962,
+      "learning_rate": 3.7842825020331704e-06,
+      "loss": 0.5405,
+      "step": 1896
+    },
+    {
+      "epoch": 1.1097378140198193,
+      "grad_norm": 0.03899078017028906,
+      "learning_rate": 3.781022642987504e-06,
+      "loss": 0.606,
+      "step": 1897
+    },
+    {
+      "epoch": 1.1103228873368194,
+      "grad_norm": 0.05928727770359705,
+      "learning_rate": 3.7777633068559485e-06,
+      "loss": 0.625,
+      "step": 1898
+    },
+    {
+      "epoch": 1.1109079606538195,
+      "grad_norm": 0.036332098995253276,
+      "learning_rate": 3.7745044963919995e-06,
+      "loss": 0.563,
+      "step": 1899
+    },
+    {
+      "epoch": 1.1114930339708196,
+      "grad_norm": 0.04254812999743139,
+      "learning_rate": 3.7712462143487117e-06,
+      "loss": 0.5929,
+      "step": 1900
+    },
+    {
+      "epoch": 1.1120781072878194,
+      "grad_norm": 0.047049542101719036,
+      "learning_rate": 3.7679884634786917e-06,
+      "loss": 0.6332,
+      "step": 1901
+    },
+    {
+      "epoch": 1.1126631806048195,
+      "grad_norm": 0.05287536643154779,
+      "learning_rate": 3.7647312465340998e-06,
+      "loss": 0.6421,
+      "step": 1902
+    },
+    {
+      "epoch": 1.1132482539218196,
+      "grad_norm": 0.04037132747201506,
+      "learning_rate": 3.7614745662666428e-06,
+      "loss": 0.6103,
+      "step": 1903
+    },
+    {
+      "epoch": 1.1138333272388197,
+      "grad_norm": 0.04432953106580269,
+      "learning_rate": 3.758218425427575e-06,
+      "loss": 0.614,
+      "step": 1904
+    },
+    {
+      "epoch": 1.1144184005558198,
+      "grad_norm": 0.040942366129102785,
+      "learning_rate": 3.754962826767695e-06,
+      "loss": 0.5639,
+      "step": 1905
+    },
+    {
+      "epoch": 1.1150034738728196,
+      "grad_norm": 0.043564413168081004,
+      "learning_rate": 3.751707773037341e-06,
+      "loss": 0.6363,
+      "step": 1906
+    },
+    {
+      "epoch": 1.1155885471898197,
+      "grad_norm": 0.04136755699281491,
+      "learning_rate": 3.748453266986397e-06,
+      "loss": 0.573,
+      "step": 1907
+    },
+    {
+      "epoch": 1.1161736205068198,
+      "grad_norm": 0.03824782580749499,
+      "learning_rate": 3.745199311364278e-06,
+      "loss": 0.5565,
+      "step": 1908
+    },
+    {
+      "epoch": 1.1167586938238199,
+      "grad_norm": 0.03974476800325834,
+      "learning_rate": 3.7419459089199386e-06,
+      "loss": 0.5743,
+      "step": 1909
+    },
+    {
+      "epoch": 1.1173437671408197,
+      "grad_norm": 0.040366618049435816,
+      "learning_rate": 3.73869306240186e-06,
+      "loss": 0.6937,
+      "step": 1910
+    },
+    {
+      "epoch": 1.1179288404578198,
+      "grad_norm": 0.05249900276814925,
+      "learning_rate": 3.7354407745580606e-06,
+      "loss": 0.6519,
+      "step": 1911
+    },
+    {
+      "epoch": 1.11851391377482,
+      "grad_norm": 0.036071772850342874,
+      "learning_rate": 3.7321890481360836e-06,
+      "loss": 0.6146,
+      "step": 1912
+    },
+    {
+      "epoch": 1.11909898709182,
+      "grad_norm": 0.03497854387051239,
+      "learning_rate": 3.7289378858829972e-06,
+      "loss": 0.5502,
+      "step": 1913
+    },
+    {
+      "epoch": 1.11968406040882,
+      "grad_norm": 0.03656290869848137,
+      "learning_rate": 3.725687290545396e-06,
+      "loss": 0.5734,
+      "step": 1914
+    },
+    {
+      "epoch": 1.12026913372582,
+      "grad_norm": 0.04595919088109381,
+      "learning_rate": 3.7224372648693914e-06,
+      "loss": 0.6477,
+      "step": 1915
+    },
+    {
+      "epoch": 1.12085420704282,
+      "grad_norm": 0.03665383639665251,
+      "learning_rate": 3.7191878116006165e-06,
+      "loss": 0.6453,
+      "step": 1916
+    },
+    {
+      "epoch": 1.12143928035982,
+      "grad_norm": 0.036312925482254364,
+      "learning_rate": 3.71593893348422e-06,
+      "loss": 0.5398,
+      "step": 1917
+    },
+    {
+      "epoch": 1.1220243536768202,
+      "grad_norm": 0.040072606927414364,
+      "learning_rate": 3.7126906332648654e-06,
+      "loss": 0.6523,
+      "step": 1918
+    },
+    {
+      "epoch": 1.12260942699382,
+      "grad_norm": 0.039060685599701495,
+      "learning_rate": 3.709442913686727e-06,
+      "loss": 0.5473,
+      "step": 1919
+    },
+    {
+      "epoch": 1.1231945003108201,
+      "grad_norm": 0.039666654769551724,
+      "learning_rate": 3.7061957774934907e-06,
+      "loss": 0.594,
+      "step": 1920
+    },
+    {
+      "epoch": 1.1237795736278202,
+      "grad_norm": 0.043648362600625844,
+      "learning_rate": 3.7029492274283447e-06,
+      "loss": 0.5124,
+      "step": 1921
+    },
+    {
+      "epoch": 1.1243646469448203,
+      "grad_norm": 0.03845451561120115,
+      "learning_rate": 3.699703266233986e-06,
+      "loss": 0.6418,
+      "step": 1922
+    },
+    {
+      "epoch": 1.1249497202618204,
+      "grad_norm": 0.0376236779158915,
+      "learning_rate": 3.696457896652616e-06,
+      "loss": 0.6533,
+      "step": 1923
+    },
+    {
+      "epoch": 1.1255347935788205,
+      "grad_norm": 0.05306744881879573,
+      "learning_rate": 3.6932131214259306e-06,
+      "loss": 0.6145,
+      "step": 1924
+    },
+    {
+      "epoch": 1.1261198668958203,
+      "grad_norm": 0.042646944784004735,
+      "learning_rate": 3.6899689432951282e-06,
+      "loss": 0.5696,
+      "step": 1925
+    },
+    {
+      "epoch": 1.1267049402128204,
+      "grad_norm": 0.038706359647602315,
+      "learning_rate": 3.6867253650008998e-06,
+      "loss": 0.6775,
+      "step": 1926
+    },
+    {
+      "epoch": 1.1272900135298205,
+      "grad_norm": 0.03479027884450534,
+      "learning_rate": 3.6834823892834316e-06,
+      "loss": 0.5519,
+      "step": 1927
+    },
+    {
+      "epoch": 1.1278750868468206,
+      "grad_norm": 0.038169378097629214,
+      "learning_rate": 3.6802400188824017e-06,
+      "loss": 0.62,
+      "step": 1928
+    },
+    {
+      "epoch": 1.1284601601638204,
+      "grad_norm": 0.04283058271473719,
+      "learning_rate": 3.676998256536972e-06,
+      "loss": 0.6568,
+      "step": 1929
+    },
+    {
+      "epoch": 1.1290452334808205,
+      "grad_norm": 0.03536662278612379,
+      "learning_rate": 3.673757104985798e-06,
+      "loss": 0.5256,
+      "step": 1930
+    },
+    {
+      "epoch": 1.1296303067978206,
+      "grad_norm": 0.052506764522120546,
+      "learning_rate": 3.6705165669670126e-06,
+      "loss": 0.6033,
+      "step": 1931
+    },
+    {
+      "epoch": 1.1302153801148207,
+      "grad_norm": 0.034900074971183304,
+      "learning_rate": 3.6672766452182346e-06,
+      "loss": 0.5209,
+      "step": 1932
+    },
+    {
+      "epoch": 1.1308004534318208,
+      "grad_norm": 0.035938101531711,
+      "learning_rate": 3.6640373424765587e-06,
+      "loss": 0.6279,
+      "step": 1933
+    },
+    {
+      "epoch": 1.1313855267488206,
+      "grad_norm": 0.0392912667983424,
+      "learning_rate": 3.660798661478558e-06,
+      "loss": 0.5815,
+      "step": 1934
+    },
+    {
+      "epoch": 1.1319706000658207,
+      "grad_norm": 0.0398303511009714,
+      "learning_rate": 3.6575606049602835e-06,
+      "loss": 0.626,
+      "step": 1935
+    },
+    {
+      "epoch": 1.1325556733828208,
+      "grad_norm": 0.03611889807550921,
+      "learning_rate": 3.6543231756572554e-06,
+      "loss": 0.6771,
+      "step": 1936
+    },
+    {
+      "epoch": 1.133140746699821,
+      "grad_norm": 0.03532028188902697,
+      "learning_rate": 3.651086376304464e-06,
+      "loss": 0.5152,
+      "step": 1937
+    },
+    {
+      "epoch": 1.1337258200168208,
+      "grad_norm": 0.03967164811503248,
+      "learning_rate": 3.647850209636367e-06,
+      "loss": 0.6211,
+      "step": 1938
+    },
+    {
+      "epoch": 1.1343108933338208,
+      "grad_norm": 0.038476418430884396,
+      "learning_rate": 3.6446146783868902e-06,
+      "loss": 0.6209,
+      "step": 1939
+    },
+    {
+      "epoch": 1.134895966650821,
+      "grad_norm": 0.04068455632446945,
+      "learning_rate": 3.6413797852894204e-06,
+      "loss": 0.5772,
+      "step": 1940
+    },
+    {
+      "epoch": 1.135481039967821,
+      "grad_norm": 0.04555373439429878,
+      "learning_rate": 3.6381455330768054e-06,
+      "loss": 0.7082,
+      "step": 1941
+    },
+    {
+      "epoch": 1.136066113284821,
+      "grad_norm": 0.03521342334256349,
+      "learning_rate": 3.634911924481355e-06,
+      "loss": 0.5159,
+      "step": 1942
+    },
+    {
+      "epoch": 1.136651186601821,
+      "grad_norm": 0.04258017514314027,
+      "learning_rate": 3.6316789622348285e-06,
+      "loss": 0.6753,
+      "step": 1943
+    },
+    {
+      "epoch": 1.137236259918821,
+      "grad_norm": 0.03516384606881102,
+      "learning_rate": 3.6284466490684453e-06,
+      "loss": 0.6492,
+      "step": 1944
+    },
+    {
+      "epoch": 1.1378213332358211,
+      "grad_norm": 0.04116371229255566,
+      "learning_rate": 3.625214987712873e-06,
+      "loss": 0.6333,
+      "step": 1945
+    },
+    {
+      "epoch": 1.1384064065528212,
+      "grad_norm": 0.04111092828869013,
+      "learning_rate": 3.621983980898231e-06,
+      "loss": 0.5967,
+      "step": 1946
+    },
+    {
+      "epoch": 1.138991479869821,
+      "grad_norm": 0.04510171654268671,
+      "learning_rate": 3.618753631354085e-06,
+      "loss": 0.6787,
+      "step": 1947
+    },
+    {
+      "epoch": 1.1395765531868212,
+      "grad_norm": 0.03695584919253051,
+      "learning_rate": 3.6155239418094423e-06,
+      "loss": 0.6041,
+      "step": 1948
+    },
+    {
+      "epoch": 1.1401616265038212,
+      "grad_norm": 0.03536563522774939,
+      "learning_rate": 3.6122949149927583e-06,
+      "loss": 0.5044,
+      "step": 1949
+    },
+    {
+      "epoch": 1.1407466998208213,
+      "grad_norm": 0.03650916289855313,
+      "learning_rate": 3.6090665536319217e-06,
+      "loss": 0.6112,
+      "step": 1950
+    },
+    {
+      "epoch": 1.1413317731378214,
+      "grad_norm": 0.039407880387507706,
+      "learning_rate": 3.605838860454266e-06,
+      "loss": 0.6436,
+      "step": 1951
+    },
+    {
+      "epoch": 1.1419168464548213,
+      "grad_norm": 0.039216656690735256,
+      "learning_rate": 3.6026118381865577e-06,
+      "loss": 0.7106,
+      "step": 1952
+    },
+    {
+      "epoch": 1.1425019197718214,
+      "grad_norm": 0.03688297014066584,
+      "learning_rate": 3.5993854895549933e-06,
+      "loss": 0.6011,
+      "step": 1953
+    },
+    {
+      "epoch": 1.1430869930888214,
+      "grad_norm": 0.042465561939299803,
+      "learning_rate": 3.596159817285203e-06,
+      "loss": 0.5543,
+      "step": 1954
+    },
+    {
+      "epoch": 1.1436720664058215,
+      "grad_norm": 0.03618763455291692,
+      "learning_rate": 3.592934824102247e-06,
+      "loss": 0.5578,
+      "step": 1955
+    },
+    {
+      "epoch": 1.1442571397228214,
+      "grad_norm": 0.036402015787460676,
+      "learning_rate": 3.5897105127306093e-06,
+      "loss": 0.6402,
+      "step": 1956
+    },
+    {
+      "epoch": 1.1448422130398215,
+      "grad_norm": 0.06377588782844341,
+      "learning_rate": 3.5864868858941975e-06,
+      "loss": 0.5915,
+      "step": 1957
+    },
+    {
+      "epoch": 1.1454272863568216,
+      "grad_norm": 0.03955675423854017,
+      "learning_rate": 3.5832639463163456e-06,
+      "loss": 0.5695,
+      "step": 1958
+    },
+    {
+      "epoch": 1.1460123596738216,
+      "grad_norm": 0.03883728695837122,
+      "learning_rate": 3.5800416967197994e-06,
+      "loss": 0.6688,
+      "step": 1959
+    },
+    {
+      "epoch": 1.1465974329908217,
+      "grad_norm": 0.0398645229902971,
+      "learning_rate": 3.5768201398267292e-06,
+      "loss": 0.5659,
+      "step": 1960
+    },
+    {
+      "epoch": 1.1471825063078218,
+      "grad_norm": 0.0411957993109104,
+      "learning_rate": 3.573599278358715e-06,
+      "loss": 0.5577,
+      "step": 1961
+    },
+    {
+      "epoch": 1.1477675796248217,
+      "grad_norm": 0.034151574742763646,
+      "learning_rate": 3.5703791150367505e-06,
+      "loss": 0.594,
+      "step": 1962
+    },
+    {
+      "epoch": 1.1483526529418218,
+      "grad_norm": 0.03855527081580829,
+      "learning_rate": 3.5671596525812443e-06,
+      "loss": 0.5414,
+      "step": 1963
+    },
+    {
+      "epoch": 1.1489377262588218,
+      "grad_norm": 0.0409230119543168,
+      "learning_rate": 3.5639408937120043e-06,
+      "loss": 0.6178,
+      "step": 1964
+    },
+    {
+      "epoch": 1.149522799575822,
+      "grad_norm": 0.042664078843859005,
+      "learning_rate": 3.5607228411482505e-06,
+      "loss": 0.6063,
+      "step": 1965
+    },
+    {
+      "epoch": 1.1501078728928218,
+      "grad_norm": 0.04213966156831369,
+      "learning_rate": 3.557505497608604e-06,
+      "loss": 0.5389,
+      "step": 1966
+    },
+    {
+      "epoch": 1.1506929462098219,
+      "grad_norm": 0.037379317263132615,
+      "learning_rate": 3.554288865811088e-06,
+      "loss": 0.6255,
+      "step": 1967
+    },
+    {
+      "epoch": 1.151278019526822,
+      "grad_norm": 0.03742555979907474,
+      "learning_rate": 3.5510729484731233e-06,
+      "loss": 0.5866,
+      "step": 1968
+    },
+    {
+      "epoch": 1.151863092843822,
+      "grad_norm": 0.03923026388121522,
+      "learning_rate": 3.5478577483115277e-06,
+      "loss": 0.6159,
+      "step": 1969
+    },
+    {
+      "epoch": 1.1524481661608221,
+      "grad_norm": 0.04249118193800539,
+      "learning_rate": 3.5446432680425115e-06,
+      "loss": 0.6039,
+      "step": 1970
+    },
+    {
+      "epoch": 1.153033239477822,
+      "grad_norm": 0.07675275707771452,
+      "learning_rate": 3.5414295103816788e-06,
+      "loss": 0.5494,
+      "step": 1971
+    },
+    {
+      "epoch": 1.153618312794822,
+      "grad_norm": 0.03437489090682111,
+      "learning_rate": 3.538216478044024e-06,
+      "loss": 0.5604,
+      "step": 1972
+    },
+    {
+      "epoch": 1.1542033861118222,
+      "grad_norm": 0.10483435145121636,
+      "learning_rate": 3.5350041737439247e-06,
+      "loss": 0.6421,
+      "step": 1973
+    },
+    {
+      "epoch": 1.1547884594288222,
+      "grad_norm": 0.047824790606816193,
+      "learning_rate": 3.5317926001951505e-06,
+      "loss": 0.6048,
+      "step": 1974
+    },
+    {
+      "epoch": 1.155373532745822,
+      "grad_norm": 0.03972268704341939,
+      "learning_rate": 3.5285817601108443e-06,
+      "loss": 0.5955,
+      "step": 1975
+    },
+    {
+      "epoch": 1.1559586060628222,
+      "grad_norm": 0.043115977687794516,
+      "learning_rate": 3.5253716562035385e-06,
+      "loss": 0.6081,
+      "step": 1976
+    },
+    {
+      "epoch": 1.1565436793798223,
+      "grad_norm": 0.039959761660668336,
+      "learning_rate": 3.522162291185136e-06,
+      "loss": 0.648,
+      "step": 1977
+    },
+    {
+      "epoch": 1.1571287526968224,
+      "grad_norm": 0.3381087189100133,
+      "learning_rate": 3.5189536677669213e-06,
+      "loss": 0.6018,
+      "step": 1978
+    },
+    {
+      "epoch": 1.1577138260138224,
+      "grad_norm": 0.03675281433214776,
+      "learning_rate": 3.515745788659549e-06,
+      "loss": 0.5625,
+      "step": 1979
+    },
+    {
+      "epoch": 1.1582988993308223,
+      "grad_norm": 0.03556127214421104,
+      "learning_rate": 3.5125386565730473e-06,
+      "loss": 0.5644,
+      "step": 1980
+    },
+    {
+      "epoch": 1.1588839726478224,
+      "grad_norm": 0.03888401533955742,
+      "learning_rate": 3.5093322742168103e-06,
+      "loss": 0.6228,
+      "step": 1981
+    },
+    {
+      "epoch": 1.1594690459648225,
+      "grad_norm": 0.07138672279489928,
+      "learning_rate": 3.5061266442996013e-06,
+      "loss": 0.6444,
+      "step": 1982
+    },
+    {
+      "epoch": 1.1600541192818226,
+      "grad_norm": 0.04367774811390162,
+      "learning_rate": 3.5029217695295464e-06,
+      "loss": 0.5675,
+      "step": 1983
+    },
+    {
+      "epoch": 1.1606391925988224,
+      "grad_norm": 0.033577525885149805,
+      "learning_rate": 3.4997176526141347e-06,
+      "loss": 0.5912,
+      "step": 1984
+    },
+    {
+      "epoch": 1.1612242659158225,
+      "grad_norm": 0.041101043350749655,
+      "learning_rate": 3.496514296260216e-06,
+      "loss": 0.5663,
+      "step": 1985
+    },
+    {
+      "epoch": 1.1618093392328226,
+      "grad_norm": 0.036838996439545135,
+      "learning_rate": 3.493311703173993e-06,
+      "loss": 0.6567,
+      "step": 1986
+    },
+    {
+      "epoch": 1.1623944125498227,
+      "grad_norm": 0.04300438023283441,
+      "learning_rate": 3.4901098760610292e-06,
+      "loss": 0.5611,
+      "step": 1987
+    },
+    {
+      "epoch": 1.1629794858668228,
+      "grad_norm": 0.03514099210411253,
+      "learning_rate": 3.4869088176262383e-06,
+      "loss": 0.5295,
+      "step": 1988
+    },
+    {
+      "epoch": 1.1635645591838228,
+      "grad_norm": 0.04249663621333313,
+      "learning_rate": 3.483708530573886e-06,
+      "loss": 0.6283,
+      "step": 1989
+    },
+    {
+      "epoch": 1.1641496325008227,
+      "grad_norm": 0.03613780881600909,
+      "learning_rate": 3.4805090176075835e-06,
+      "loss": 0.6074,
+      "step": 1990
+    },
+    {
+      "epoch": 1.1647347058178228,
+      "grad_norm": 0.04959791048909453,
+      "learning_rate": 3.4773102814302915e-06,
+      "loss": 0.6174,
+      "step": 1991
+    },
+    {
+      "epoch": 1.1653197791348229,
+      "grad_norm": 0.037749992452740126,
+      "learning_rate": 3.474112324744312e-06,
+      "loss": 0.5907,
+      "step": 1992
+    },
+    {
+      "epoch": 1.165904852451823,
+      "grad_norm": 0.0351608550697518,
+      "learning_rate": 3.470915150251289e-06,
+      "loss": 0.5905,
+      "step": 1993
+    },
+    {
+      "epoch": 1.1664899257688228,
+      "grad_norm": 0.03893625789488356,
+      "learning_rate": 3.4677187606522066e-06,
+      "loss": 0.5675,
+      "step": 1994
+    },
+    {
+      "epoch": 1.167074999085823,
+      "grad_norm": 0.04059108907664171,
+      "learning_rate": 3.4645231586473856e-06,
+      "loss": 0.5728,
+      "step": 1995
+    },
+    {
+      "epoch": 1.167660072402823,
+      "grad_norm": 0.04113287183937823,
+      "learning_rate": 3.4613283469364812e-06,
+      "loss": 0.6118,
+      "step": 1996
+    },
+    {
+      "epoch": 1.168245145719823,
+      "grad_norm": 0.03719646254418526,
+      "learning_rate": 3.45813432821848e-06,
+      "loss": 0.6406,
+      "step": 1997
+    },
+    {
+      "epoch": 1.1688302190368232,
+      "grad_norm": 0.06040366792983708,
+      "learning_rate": 3.4549411051916993e-06,
+      "loss": 0.6843,
+      "step": 1998
+    },
+    {
+      "epoch": 1.169415292353823,
+      "grad_norm": 0.039091200896895856,
+      "learning_rate": 3.451748680553785e-06,
+      "loss": 0.5895,
+      "step": 1999
+    },
+    {
+      "epoch": 1.170000365670823,
+      "grad_norm": 0.03719442419223529,
+      "learning_rate": 3.4485570570017083e-06,
+      "loss": 0.6165,
+      "step": 2000
+    },
+    {
+      "epoch": 1.1705854389878232,
+      "grad_norm": 0.036207964808247124,
+      "learning_rate": 3.445366237231762e-06,
+      "loss": 0.5825,
+      "step": 2001
+    },
+    {
+      "epoch": 1.1711705123048233,
+      "grad_norm": 0.037815350275091156,
+      "learning_rate": 3.4421762239395646e-06,
+      "loss": 0.5934,
+      "step": 2002
+    },
+    {
+      "epoch": 1.1717555856218231,
+      "grad_norm": 0.054653596037471275,
+      "learning_rate": 3.438987019820045e-06,
+      "loss": 0.6826,
+      "step": 2003
+    },
+    {
+      "epoch": 1.1723406589388232,
+      "grad_norm": 0.03400992901791061,
+      "learning_rate": 3.4357986275674566e-06,
+      "loss": 0.4815,
+      "step": 2004
+    },
+    {
+      "epoch": 1.1729257322558233,
+      "grad_norm": 0.0466833061216275,
+      "learning_rate": 3.432611049875362e-06,
+      "loss": 0.6306,
+      "step": 2005
+    },
+    {
+      "epoch": 1.1735108055728234,
+      "grad_norm": 0.03423491044593052,
+      "learning_rate": 3.429424289436638e-06,
+      "loss": 0.5756,
+      "step": 2006
+    },
+    {
+      "epoch": 1.1740958788898235,
+      "grad_norm": 0.04055240441842592,
+      "learning_rate": 3.426238348943472e-06,
+      "loss": 0.5772,
+      "step": 2007
+    },
+    {
+      "epoch": 1.1746809522068233,
+      "grad_norm": 0.1689145341911383,
+      "learning_rate": 3.423053231087354e-06,
+      "loss": 0.6136,
+      "step": 2008
+    },
+    {
+      "epoch": 1.1752660255238234,
+      "grad_norm": 0.036623976167559434,
+      "learning_rate": 3.4198689385590848e-06,
+      "loss": 0.5839,
+      "step": 2009
+    },
+    {
+      "epoch": 1.1758510988408235,
+      "grad_norm": 0.038221379263121774,
+      "learning_rate": 3.416685474048761e-06,
+      "loss": 0.7034,
+      "step": 2010
+    },
+    {
+      "epoch": 1.1764361721578236,
+      "grad_norm": 0.03539988785174503,
+      "learning_rate": 3.4135028402457888e-06,
+      "loss": 0.5319,
+      "step": 2011
+    },
+    {
+      "epoch": 1.1770212454748235,
+      "grad_norm": 0.03800403001773324,
+      "learning_rate": 3.410321039838866e-06,
+      "loss": 0.6349,
+      "step": 2012
+    },
+    {
+      "epoch": 1.1776063187918235,
+      "grad_norm": 0.04227520585992831,
+      "learning_rate": 3.407140075515988e-06,
+      "loss": 0.6129,
+      "step": 2013
+    },
+    {
+      "epoch": 1.1781913921088236,
+      "grad_norm": 0.03831788774127032,
+      "learning_rate": 3.4039599499644434e-06,
+      "loss": 0.5848,
+      "step": 2014
+    },
+    {
+      "epoch": 1.1787764654258237,
+      "grad_norm": 0.04284762288768014,
+      "learning_rate": 3.400780665870811e-06,
+      "loss": 0.6488,
+      "step": 2015
+    },
+    {
+      "epoch": 1.1793615387428238,
+      "grad_norm": 0.041288161978273966,
+      "learning_rate": 3.3976022259209643e-06,
+      "loss": 0.627,
+      "step": 2016
+    },
+    {
+      "epoch": 1.1799466120598237,
+      "grad_norm": 0.039481075603532444,
+      "learning_rate": 3.3944246328000574e-06,
+      "loss": 0.5888,
+      "step": 2017
+    },
+    {
+      "epoch": 1.1805316853768237,
+      "grad_norm": 0.04335101317156415,
+      "learning_rate": 3.3912478891925338e-06,
+      "loss": 0.6734,
+      "step": 2018
+    },
+    {
+      "epoch": 1.1811167586938238,
+      "grad_norm": 0.04116415093116544,
+      "learning_rate": 3.3880719977821127e-06,
+      "loss": 0.6432,
+      "step": 2019
+    },
+    {
+      "epoch": 1.181701832010824,
+      "grad_norm": 0.036848265334549324,
+      "learning_rate": 3.384896961251802e-06,
+      "loss": 0.5763,
+      "step": 2020
+    },
+    {
+      "epoch": 1.182286905327824,
+      "grad_norm": 0.042651651929545335,
+      "learning_rate": 3.38172278228388e-06,
+      "loss": 0.5687,
+      "step": 2021
+    },
+    {
+      "epoch": 1.1828719786448239,
+      "grad_norm": 0.03635888340073088,
+      "learning_rate": 3.3785494635599048e-06,
+      "loss": 0.5954,
+      "step": 2022
+    },
+    {
+      "epoch": 1.183457051961824,
+      "grad_norm": 0.039839665329704434,
+      "learning_rate": 3.3753770077607076e-06,
+      "loss": 0.6383,
+      "step": 2023
+    },
+    {
+      "epoch": 1.184042125278824,
+      "grad_norm": 0.043332517759083845,
+      "learning_rate": 3.372205417566387e-06,
+      "loss": 0.6532,
+      "step": 2024
+    },
+    {
+      "epoch": 1.184627198595824,
+      "grad_norm": 0.03562735039321648,
+      "learning_rate": 3.3690346956563146e-06,
+      "loss": 0.6788,
+      "step": 2025
+    },
+    {
+      "epoch": 1.1852122719128242,
+      "grad_norm": 0.04660842986654955,
+      "learning_rate": 3.365864844709127e-06,
+      "loss": 0.6477,
+      "step": 2026
+    },
+    {
+      "epoch": 1.185797345229824,
+      "grad_norm": 0.03429494053167668,
+      "learning_rate": 3.3626958674027213e-06,
+      "loss": 0.5504,
+      "step": 2027
+    },
+    {
+      "epoch": 1.1863824185468241,
+      "grad_norm": 0.0680307436226198,
+      "learning_rate": 3.359527766414265e-06,
+      "loss": 0.668,
+      "step": 2028
+    },
+    {
+      "epoch": 1.1869674918638242,
+      "grad_norm": 0.05407036614682854,
+      "learning_rate": 3.3563605444201785e-06,
+      "loss": 0.6085,
+      "step": 2029
+    },
+    {
+      "epoch": 1.1875525651808243,
+      "grad_norm": 0.03428910211347692,
+      "learning_rate": 3.35319420409614e-06,
+      "loss": 0.5544,
+      "step": 2030
+    },
+    {
+      "epoch": 1.1881376384978242,
+      "grad_norm": 0.04221499184751478,
+      "learning_rate": 3.3500287481170847e-06,
+      "loss": 0.7486,
+      "step": 2031
+    },
+    {
+      "epoch": 1.1887227118148243,
+      "grad_norm": 0.03921755539646231,
+      "learning_rate": 3.3468641791572003e-06,
+      "loss": 0.6926,
+      "step": 2032
+    },
+    {
+      "epoch": 1.1893077851318243,
+      "grad_norm": 0.0366270608635313,
+      "learning_rate": 3.343700499889926e-06,
+      "loss": 0.6307,
+      "step": 2033
+    },
+    {
+      "epoch": 1.1898928584488244,
+      "grad_norm": 0.036660916253845384,
+      "learning_rate": 3.340537712987948e-06,
+      "loss": 0.5853,
+      "step": 2034
+    },
+    {
+      "epoch": 1.1904779317658245,
+      "grad_norm": 0.03505156447428476,
+      "learning_rate": 3.3373758211231974e-06,
+      "loss": 0.5874,
+      "step": 2035
+    },
+    {
+      "epoch": 1.1910630050828244,
+      "grad_norm": 0.040178249649878345,
+      "learning_rate": 3.3342148269668525e-06,
+      "loss": 0.7006,
+      "step": 2036
+    },
+    {
+      "epoch": 1.1916480783998245,
+      "grad_norm": 0.03533674390624903,
+      "learning_rate": 3.331054733189332e-06,
+      "loss": 0.5886,
+      "step": 2037
+    },
+    {
+      "epoch": 1.1922331517168245,
+      "grad_norm": 0.03692572558788983,
+      "learning_rate": 3.327895542460292e-06,
+      "loss": 0.6155,
+      "step": 2038
+    },
+    {
+      "epoch": 1.1928182250338246,
+      "grad_norm": 0.03938265009175071,
+      "learning_rate": 3.3247372574486302e-06,
+      "loss": 0.6368,
+      "step": 2039
+    },
+    {
+      "epoch": 1.1934032983508245,
+      "grad_norm": 0.04043208798531885,
+      "learning_rate": 3.321579880822473e-06,
+      "loss": 0.541,
+      "step": 2040
+    },
+    {
+      "epoch": 1.1939883716678246,
+      "grad_norm": 0.049286317781066166,
+      "learning_rate": 3.318423415249184e-06,
+      "loss": 0.5622,
+      "step": 2041
+    },
+    {
+      "epoch": 1.1945734449848247,
+      "grad_norm": 0.035803834914162815,
+      "learning_rate": 3.3152678633953554e-06,
+      "loss": 0.6379,
+      "step": 2042
+    },
+    {
+      "epoch": 1.1951585183018247,
+      "grad_norm": 0.03534002984635534,
+      "learning_rate": 3.312113227926807e-06,
+      "loss": 0.5832,
+      "step": 2043
+    },
+    {
+      "epoch": 1.1957435916188248,
+      "grad_norm": 0.03595263497698763,
+      "learning_rate": 3.3089595115085863e-06,
+      "loss": 0.6027,
+      "step": 2044
+    },
+    {
+      "epoch": 1.1963286649358247,
+      "grad_norm": 0.04310890199047902,
+      "learning_rate": 3.305806716804964e-06,
+      "loss": 0.6981,
+      "step": 2045
+    },
+    {
+      "epoch": 1.1969137382528248,
+      "grad_norm": 0.034609593715224284,
+      "learning_rate": 3.302654846479431e-06,
+      "loss": 0.5436,
+      "step": 2046
+    },
+    {
+      "epoch": 1.1974988115698249,
+      "grad_norm": 0.038358840771114136,
+      "learning_rate": 3.2995039031946954e-06,
+      "loss": 0.7055,
+      "step": 2047
+    },
+    {
+      "epoch": 1.198083884886825,
+      "grad_norm": 0.04563367988331309,
+      "learning_rate": 3.2963538896126863e-06,
+      "loss": 0.5472,
+      "step": 2048
+    },
+    {
+      "epoch": 1.1986689582038248,
+      "grad_norm": 0.03704695342032978,
+      "learning_rate": 3.293204808394547e-06,
+      "loss": 0.5715,
+      "step": 2049
+    },
+    {
+      "epoch": 1.1992540315208249,
+      "grad_norm": 0.03741146134749107,
+      "learning_rate": 3.2900566622006284e-06,
+      "loss": 0.6012,
+      "step": 2050
+    },
+    {
+      "epoch": 1.199839104837825,
+      "grad_norm": 0.038195831443333515,
+      "learning_rate": 3.2869094536904997e-06,
+      "loss": 0.5778,
+      "step": 2051
+    },
+    {
+      "epoch": 1.200424178154825,
+      "grad_norm": 0.03838310724213011,
+      "learning_rate": 3.283763185522927e-06,
+      "loss": 0.6203,
+      "step": 2052
+    },
+    {
+      "epoch": 1.2010092514718251,
+      "grad_norm": 0.03679246721866951,
+      "learning_rate": 3.2806178603558923e-06,
+      "loss": 0.5866,
+      "step": 2053
+    },
+    {
+      "epoch": 1.2015943247888252,
+      "grad_norm": 0.037329068284633296,
+      "learning_rate": 3.2774734808465746e-06,
+      "loss": 0.5712,
+      "step": 2054
+    },
+    {
+      "epoch": 1.202179398105825,
+      "grad_norm": 0.03824528624286622,
+      "learning_rate": 3.2743300496513567e-06,
+      "loss": 0.6278,
+      "step": 2055
+    },
+    {
+      "epoch": 1.2027644714228252,
+      "grad_norm": 0.036655970928988306,
+      "learning_rate": 3.2711875694258202e-06,
+      "loss": 0.6234,
+      "step": 2056
+    },
+    {
+      "epoch": 1.2033495447398252,
+      "grad_norm": 0.03552385584165378,
+      "learning_rate": 3.268046042824742e-06,
+      "loss": 0.4884,
+      "step": 2057
+    },
+    {
+      "epoch": 1.2039346180568253,
+      "grad_norm": 0.03841046001130388,
+      "learning_rate": 3.2649054725020947e-06,
+      "loss": 0.664,
+      "step": 2058
+    },
+    {
+      "epoch": 1.2045196913738252,
+      "grad_norm": 0.04118795268419084,
+      "learning_rate": 3.2617658611110405e-06,
+      "loss": 0.6416,
+      "step": 2059
+    },
+    {
+      "epoch": 1.2051047646908253,
+      "grad_norm": 0.03850987838668136,
+      "learning_rate": 3.2586272113039368e-06,
+      "loss": 0.5786,
+      "step": 2060
+    },
+    {
+      "epoch": 1.2056898380078254,
+      "grad_norm": 0.038494984464432946,
+      "learning_rate": 3.255489525732323e-06,
+      "loss": 0.6438,
+      "step": 2061
+    },
+    {
+      "epoch": 1.2062749113248254,
+      "grad_norm": 0.035177849537891506,
+      "learning_rate": 3.252352807046927e-06,
+      "loss": 0.5675,
+      "step": 2062
+    },
+    {
+      "epoch": 1.2068599846418255,
+      "grad_norm": 0.04015731464340186,
+      "learning_rate": 3.249217057897658e-06,
+      "loss": 0.5964,
+      "step": 2063
+    },
+    {
+      "epoch": 1.2074450579588254,
+      "grad_norm": 0.04066687077935284,
+      "learning_rate": 3.246082280933609e-06,
+      "loss": 0.5642,
+      "step": 2064
+    },
+    {
+      "epoch": 1.2080301312758255,
+      "grad_norm": 0.050549965114778064,
+      "learning_rate": 3.242948478803048e-06,
+      "loss": 0.5849,
+      "step": 2065
+    },
+    {
+      "epoch": 1.2086152045928256,
+      "grad_norm": 0.04128169248463715,
+      "learning_rate": 3.239815654153422e-06,
+      "loss": 0.6244,
+      "step": 2066
+    },
+    {
+      "epoch": 1.2092002779098256,
+      "grad_norm": 0.037146084532940146,
+      "learning_rate": 3.236683809631354e-06,
+      "loss": 0.5625,
+      "step": 2067
+    },
+    {
+      "epoch": 1.2097853512268255,
+      "grad_norm": 0.037953855830558866,
+      "learning_rate": 3.2335529478826324e-06,
+      "loss": 0.6,
+      "step": 2068
+    },
+    {
+      "epoch": 1.2103704245438256,
+      "grad_norm": 0.04943862913081636,
+      "learning_rate": 3.230423071552222e-06,
+      "loss": 0.5952,
+      "step": 2069
+    },
+    {
+      "epoch": 1.2109554978608257,
+      "grad_norm": 0.03936477462539683,
+      "learning_rate": 3.227294183284252e-06,
+      "loss": 0.6401,
+      "step": 2070
+    },
+    {
+      "epoch": 1.2115405711778258,
+      "grad_norm": 0.04126544367881334,
+      "learning_rate": 3.2241662857220176e-06,
+      "loss": 0.6159,
+      "step": 2071
+    },
+    {
+      "epoch": 1.2121256444948258,
+      "grad_norm": 0.041020884786539924,
+      "learning_rate": 3.2210393815079773e-06,
+      "loss": 0.635,
+      "step": 2072
+    },
+    {
+      "epoch": 1.2127107178118257,
+      "grad_norm": 0.0358878906382118,
+      "learning_rate": 3.21791347328375e-06,
+      "loss": 0.6414,
+      "step": 2073
+    },
+    {
+      "epoch": 1.2132957911288258,
+      "grad_norm": 0.03516770521480586,
+      "learning_rate": 3.2147885636901136e-06,
+      "loss": 0.6641,
+      "step": 2074
+    },
+    {
+      "epoch": 1.2138808644458259,
+      "grad_norm": 0.03926231427900321,
+      "learning_rate": 3.2116646553669994e-06,
+      "loss": 0.626,
+      "step": 2075
+    },
+    {
+      "epoch": 1.214465937762826,
+      "grad_norm": 0.03869243094148781,
+      "learning_rate": 3.2085417509534994e-06,
+      "loss": 0.6433,
+      "step": 2076
+    },
+    {
+      "epoch": 1.2150510110798258,
+      "grad_norm": 0.03914650880562719,
+      "learning_rate": 3.205419853087851e-06,
+      "loss": 0.6486,
+      "step": 2077
+    },
+    {
+      "epoch": 1.215636084396826,
+      "grad_norm": 0.03719572794194729,
+      "learning_rate": 3.202298964407445e-06,
+      "loss": 0.6049,
+      "step": 2078
+    },
+    {
+      "epoch": 1.216221157713826,
+      "grad_norm": 0.039960704683083076,
+      "learning_rate": 3.199179087548818e-06,
+      "loss": 0.5408,
+      "step": 2079
+    },
+    {
+      "epoch": 1.216806231030826,
+      "grad_norm": 0.038852419722138556,
+      "learning_rate": 3.1960602251476508e-06,
+      "loss": 0.6398,
+      "step": 2080
+    },
+    {
+      "epoch": 1.2173913043478262,
+      "grad_norm": 0.043605000020103986,
+      "learning_rate": 3.1929423798387706e-06,
+      "loss": 0.6421,
+      "step": 2081
+    },
+    {
+      "epoch": 1.217976377664826,
+      "grad_norm": 0.03878520292515712,
+      "learning_rate": 3.189825554256142e-06,
+      "loss": 0.6067,
+      "step": 2082
+    },
+    {
+      "epoch": 1.218561450981826,
+      "grad_norm": 0.041501370594404216,
+      "learning_rate": 3.186709751032871e-06,
+      "loss": 0.6132,
+      "step": 2083
+    },
+    {
+      "epoch": 1.2191465242988262,
+      "grad_norm": 0.03951034345012451,
+      "learning_rate": 3.183594972801199e-06,
+      "loss": 0.6016,
+      "step": 2084
+    },
+    {
+      "epoch": 1.2197315976158263,
+      "grad_norm": 0.04111002191035941,
+      "learning_rate": 3.1804812221924973e-06,
+      "loss": 0.7306,
+      "step": 2085
+    },
+    {
+      "epoch": 1.2203166709328264,
+      "grad_norm": 0.04145625722473185,
+      "learning_rate": 3.177368501837277e-06,
+      "loss": 0.6854,
+      "step": 2086
+    },
+    {
+      "epoch": 1.2209017442498262,
+      "grad_norm": 0.037269331694012336,
+      "learning_rate": 3.1742568143651704e-06,
+      "loss": 0.6833,
+      "step": 2087
+    },
+    {
+      "epoch": 1.2214868175668263,
+      "grad_norm": 0.039255496027705915,
+      "learning_rate": 3.171146162404945e-06,
+      "loss": 0.5482,
+      "step": 2088
+    },
+    {
+      "epoch": 1.2220718908838264,
+      "grad_norm": 0.037070583225945676,
+      "learning_rate": 3.168036548584488e-06,
+      "loss": 0.5844,
+      "step": 2089
+    },
+    {
+      "epoch": 1.2226569642008265,
+      "grad_norm": 0.11493255093128565,
+      "learning_rate": 3.164927975530812e-06,
+      "loss": 0.6224,
+      "step": 2090
+    },
+    {
+      "epoch": 1.2232420375178266,
+      "grad_norm": 0.037232680426261475,
+      "learning_rate": 3.161820445870048e-06,
+      "loss": 0.6469,
+      "step": 2091
+    },
+    {
+      "epoch": 1.2238271108348264,
+      "grad_norm": 0.04032812893153942,
+      "learning_rate": 3.1587139622274482e-06,
+      "loss": 0.6057,
+      "step": 2092
+    },
+    {
+      "epoch": 1.2244121841518265,
+      "grad_norm": 0.03634438517480883,
+      "learning_rate": 3.1556085272273826e-06,
+      "loss": 0.544,
+      "step": 2093
+    },
+    {
+      "epoch": 1.2249972574688266,
+      "grad_norm": 0.04037981016869187,
+      "learning_rate": 3.1525041434933304e-06,
+      "loss": 0.6455,
+      "step": 2094
+    },
+    {
+      "epoch": 1.2255823307858267,
+      "grad_norm": 0.038882617943542315,
+      "learning_rate": 3.149400813647886e-06,
+      "loss": 0.6386,
+      "step": 2095
+    },
+    {
+      "epoch": 1.2261674041028265,
+      "grad_norm": 0.051431593131162474,
+      "learning_rate": 3.1462985403127505e-06,
+      "loss": 0.6475,
+      "step": 2096
+    },
+    {
+      "epoch": 1.2267524774198266,
+      "grad_norm": 0.03733875116363724,
+      "learning_rate": 3.143197326108738e-06,
+      "loss": 0.5444,
+      "step": 2097
+    },
+    {
+      "epoch": 1.2273375507368267,
+      "grad_norm": 0.035252501575400275,
+      "learning_rate": 3.140097173655763e-06,
+      "loss": 0.5779,
+      "step": 2098
+    },
+    {
+      "epoch": 1.2279226240538268,
+      "grad_norm": 0.05240477506652852,
+      "learning_rate": 3.136998085572844e-06,
+      "loss": 0.5887,
+      "step": 2099
+    },
+    {
+      "epoch": 1.2285076973708269,
+      "grad_norm": 0.03805752290907818,
+      "learning_rate": 3.1339000644781022e-06,
+      "loss": 0.5838,
+      "step": 2100
+    },
+    {
+      "epoch": 1.2290927706878267,
+      "grad_norm": 0.04383435394432525,
+      "learning_rate": 3.130803112988755e-06,
+      "loss": 0.6787,
+      "step": 2101
+    },
+    {
+      "epoch": 1.2296778440048268,
+      "grad_norm": 0.03732331672312392,
+      "learning_rate": 3.1277072337211177e-06,
+      "loss": 0.6056,
+      "step": 2102
+    },
+    {
+      "epoch": 1.230262917321827,
+      "grad_norm": 0.036913976358249244,
+      "learning_rate": 3.1246124292905977e-06,
+      "loss": 0.5794,
+      "step": 2103
+    },
+    {
+      "epoch": 1.230847990638827,
+      "grad_norm": 0.045315422806353534,
+      "learning_rate": 3.1215187023116983e-06,
+      "loss": 0.6712,
+      "step": 2104
+    },
+    {
+      "epoch": 1.2314330639558269,
+      "grad_norm": 0.03796939196362949,
+      "learning_rate": 3.11842605539801e-06,
+      "loss": 0.5266,
+      "step": 2105
+    },
+    {
+      "epoch": 1.232018137272827,
+      "grad_norm": 0.26883650914769597,
+      "learning_rate": 3.1153344911622093e-06,
+      "loss": 0.5122,
+      "step": 2106
+    },
+    {
+      "epoch": 1.232603210589827,
+      "grad_norm": 0.046560578154564256,
+      "learning_rate": 3.1122440122160603e-06,
+      "loss": 0.6583,
+      "step": 2107
+    },
+    {
+      "epoch": 1.233188283906827,
+      "grad_norm": 0.039487212735197195,
+      "learning_rate": 3.10915462117041e-06,
+      "loss": 0.566,
+      "step": 2108
+    },
+    {
+      "epoch": 1.2337733572238272,
+      "grad_norm": 0.044996273718641414,
+      "learning_rate": 3.106066320635188e-06,
+      "loss": 0.6701,
+      "step": 2109
+    },
+    {
+      "epoch": 1.234358430540827,
+      "grad_norm": 0.03714598377688584,
+      "learning_rate": 3.1029791132193978e-06,
+      "loss": 0.5602,
+      "step": 2110
+    },
+    {
+      "epoch": 1.2349435038578271,
+      "grad_norm": 0.04302535784595455,
+      "learning_rate": 3.099893001531126e-06,
+      "loss": 0.7246,
+      "step": 2111
+    },
+    {
+      "epoch": 1.2355285771748272,
+      "grad_norm": 0.0377049325672757,
+      "learning_rate": 3.0968079881775257e-06,
+      "loss": 0.6343,
+      "step": 2112
+    },
+    {
+      "epoch": 1.2361136504918273,
+      "grad_norm": 0.04228612856033121,
+      "learning_rate": 3.0937240757648287e-06,
+      "loss": 0.6611,
+      "step": 2113
+    },
+    {
+      "epoch": 1.2366987238088272,
+      "grad_norm": 0.035689681356042134,
+      "learning_rate": 3.090641266898335e-06,
+      "loss": 0.5975,
+      "step": 2114
+    },
+    {
+      "epoch": 1.2372837971258273,
+      "grad_norm": 0.03471425835558048,
+      "learning_rate": 3.0875595641824096e-06,
+      "loss": 0.5574,
+      "step": 2115
+    },
+    {
+      "epoch": 1.2378688704428273,
+      "grad_norm": 0.03692356893026278,
+      "learning_rate": 3.0844789702204886e-06,
+      "loss": 0.595,
+      "step": 2116
+    },
+    {
+      "epoch": 1.2384539437598274,
+      "grad_norm": 0.03922116778653568,
+      "learning_rate": 3.081399487615062e-06,
+      "loss": 0.6189,
+      "step": 2117
+    },
+    {
+      "epoch": 1.2390390170768275,
+      "grad_norm": 0.05993653316174921,
+      "learning_rate": 3.078321118967693e-06,
+      "loss": 0.6765,
+      "step": 2118
+    },
+    {
+      "epoch": 1.2396240903938276,
+      "grad_norm": 0.04167106571338825,
+      "learning_rate": 3.0752438668789927e-06,
+      "loss": 0.6151,
+      "step": 2119
+    },
+    {
+      "epoch": 1.2402091637108275,
+      "grad_norm": 0.03821887002642105,
+      "learning_rate": 3.072167733948638e-06,
+      "loss": 0.607,
+      "step": 2120
+    },
+    {
+      "epoch": 1.2407942370278275,
+      "grad_norm": 0.03807112943742371,
+      "learning_rate": 3.0690927227753537e-06,
+      "loss": 0.6594,
+      "step": 2121
+    },
+    {
+      "epoch": 1.2413793103448276,
+      "grad_norm": 0.040714922346881566,
+      "learning_rate": 3.066018835956921e-06,
+      "loss": 0.5834,
+      "step": 2122
+    },
+    {
+      "epoch": 1.2419643836618277,
+      "grad_norm": 0.039133521242853846,
+      "learning_rate": 3.0629460760901673e-06,
+      "loss": 0.597,
+      "step": 2123
+    },
+    {
+      "epoch": 1.2425494569788276,
+      "grad_norm": 0.0400356957836742,
+      "learning_rate": 3.0598744457709707e-06,
+      "loss": 0.6055,
+      "step": 2124
+    },
+    {
+      "epoch": 1.2431345302958277,
+      "grad_norm": 0.04123802601032258,
+      "learning_rate": 3.0568039475942556e-06,
+      "loss": 0.5654,
+      "step": 2125
+    },
+    {
+      "epoch": 1.2437196036128277,
+      "grad_norm": 0.03639596855605065,
+      "learning_rate": 3.0537345841539895e-06,
+      "loss": 0.6412,
+      "step": 2126
+    },
+    {
+      "epoch": 1.2443046769298278,
+      "grad_norm": 0.04019621315125225,
+      "learning_rate": 3.0506663580431805e-06,
+      "loss": 0.5935,
+      "step": 2127
+    },
+    {
+      "epoch": 1.244889750246828,
+      "grad_norm": 0.04079355192375721,
+      "learning_rate": 3.0475992718538734e-06,
+      "loss": 0.5552,
+      "step": 2128
+    },
+    {
+      "epoch": 1.2454748235638278,
+      "grad_norm": 0.041408451026744786,
+      "learning_rate": 3.044533328177157e-06,
+      "loss": 0.584,
+      "step": 2129
+    },
+    {
+      "epoch": 1.2460598968808279,
+      "grad_norm": 0.05090259880775753,
+      "learning_rate": 3.041468529603147e-06,
+      "loss": 0.6054,
+      "step": 2130
+    },
+    {
+      "epoch": 1.246644970197828,
+      "grad_norm": 0.03868926058783337,
+      "learning_rate": 3.038404878720997e-06,
+      "loss": 0.5618,
+      "step": 2131
+    },
+    {
+      "epoch": 1.247230043514828,
+      "grad_norm": 0.06601222584424662,
+      "learning_rate": 3.0353423781188897e-06,
+      "loss": 0.5544,
+      "step": 2132
+    },
+    {
+      "epoch": 1.2478151168318279,
+      "grad_norm": 0.03702617313486051,
+      "learning_rate": 3.032281030384036e-06,
+      "loss": 0.5989,
+      "step": 2133
+    },
+    {
+      "epoch": 1.248400190148828,
+      "grad_norm": 0.03705800256241544,
+      "learning_rate": 3.0292208381026703e-06,
+      "loss": 0.6298,
+      "step": 2134
+    },
+    {
+      "epoch": 1.248985263465828,
+      "grad_norm": 0.038560461227645317,
+      "learning_rate": 3.0261618038600556e-06,
+      "loss": 0.5769,
+      "step": 2135
+    },
+    {
+      "epoch": 1.2495703367828281,
+      "grad_norm": 0.04189426245677959,
+      "learning_rate": 3.023103930240471e-06,
+      "loss": 0.5847,
+      "step": 2136
+    },
+    {
+      "epoch": 1.2501554100998282,
+      "grad_norm": 0.05575302954020764,
+      "learning_rate": 3.020047219827222e-06,
+      "loss": 0.5339,
+      "step": 2137
+    },
+    {
+      "epoch": 1.2507404834168283,
+      "grad_norm": 0.04057133712962797,
+      "learning_rate": 3.016991675202627e-06,
+      "loss": 0.5519,
+      "step": 2138
+    },
+    {
+      "epoch": 1.2513255567338282,
+      "grad_norm": 0.03597036946672713,
+      "learning_rate": 3.0139372989480167e-06,
+      "loss": 0.5791,
+      "step": 2139
+    },
+    {
+      "epoch": 1.2519106300508283,
+      "grad_norm": 0.04334307004734275,
+      "learning_rate": 3.0108840936437413e-06,
+      "loss": 0.5256,
+      "step": 2140
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3418,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 428,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8878468884332544.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}