diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,40231 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999310276874569,
+  "eval_steps": 400,
+  "global_step": 11415,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005255033336617729,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0,
+      "loss": 9.1222,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010510066673235458,
+      "grad_norm": 38.698402404785156,
+      "learning_rate": 9.999123882950763e-05,
+      "loss": 9.1362,
+      "step": 4
+    },
+    {
+      "epoch": 0.0015765100009853187,
+      "grad_norm": 23.888906478881836,
+      "learning_rate": 9.997371648852288e-05,
+      "loss": 6.4836,
+      "step": 6
+    },
+    {
+      "epoch": 0.0021020133346470915,
+      "grad_norm": 12.829351425170898,
+      "learning_rate": 9.995619414753812e-05,
+      "loss": 5.5976,
+      "step": 8
+    },
+    {
+      "epoch": 0.0026275166683088647,
+      "grad_norm": 13.441329956054688,
+      "learning_rate": 9.993867180655336e-05,
+      "loss": 4.922,
+      "step": 10
+    },
+    {
+      "epoch": 0.0031530200019706375,
+      "grad_norm": 8.487077713012695,
+      "learning_rate": 9.99211494655686e-05,
+      "loss": 4.3391,
+      "step": 12
+    },
+    {
+      "epoch": 0.0036785233356324103,
+      "grad_norm": 7.427611351013184,
+      "learning_rate": 9.990362712458385e-05,
+      "loss": 3.8921,
+      "step": 14
+    },
+    {
+      "epoch": 0.004204026669294183,
+      "grad_norm": 6.296542167663574,
+      "learning_rate": 9.988610478359909e-05,
+      "loss": 3.5532,
+      "step": 16
+    },
+    {
+      "epoch": 0.004729530002955957,
+      "grad_norm": 5.726244926452637,
+      "learning_rate": 9.986858244261433e-05,
+      "loss": 3.2786,
+      "step": 18
+    },
+    {
+      "epoch": 0.0052550333366177295,
+      "grad_norm": 5.131892204284668,
+      "learning_rate": 9.985106010162958e-05,
+      "loss": 3.016,
+      "step": 20
+    },
+    {
+      "epoch": 0.005780536670279502,
+      "grad_norm": 4.895572185516357,
+      "learning_rate": 9.983353776064483e-05,
+      "loss": 2.8689,
+      "step": 22
+    },
+    {
+      "epoch": 0.006306040003941275,
+      "grad_norm": 4.9487714767456055,
+      "learning_rate": 9.981601541966006e-05,
+      "loss": 2.7813,
+      "step": 24
+    },
+    {
+      "epoch": 0.006831543337603048,
+      "grad_norm": 4.282960891723633,
+      "learning_rate": 9.979849307867531e-05,
+      "loss": 2.6648,
+      "step": 26
+    },
+    {
+      "epoch": 0.0073570466712648205,
+      "grad_norm": 3.610912561416626,
+      "learning_rate": 9.978097073769056e-05,
+      "loss": 2.6083,
+      "step": 28
+    },
+    {
+      "epoch": 0.007882550004926594,
+      "grad_norm": 3.4365551471710205,
+      "learning_rate": 9.97634483967058e-05,
+      "loss": 2.5791,
+      "step": 30
+    },
+    {
+      "epoch": 0.008408053338588366,
+      "grad_norm": 3.2651638984680176,
+      "learning_rate": 9.974592605572105e-05,
+      "loss": 2.5128,
+      "step": 32
+    },
+    {
+      "epoch": 0.00893355667225014,
+      "grad_norm": 3.259965658187866,
+      "learning_rate": 9.97284037147363e-05,
+      "loss": 2.4589,
+      "step": 34
+    },
+    {
+      "epoch": 0.009459060005911913,
+      "grad_norm": 3.229142904281616,
+      "learning_rate": 9.971088137375154e-05,
+      "loss": 2.4857,
+      "step": 36
+    },
+    {
+      "epoch": 0.009984563339573685,
+      "grad_norm": 3.1428916454315186,
+      "learning_rate": 9.969335903276678e-05,
+      "loss": 2.4305,
+      "step": 38
+    },
+    {
+      "epoch": 0.010510066673235459,
+      "grad_norm": 2.565325975418091,
+      "learning_rate": 9.967583669178203e-05,
+      "loss": 2.4047,
+      "step": 40
+    },
+    {
+      "epoch": 0.01103557000689723,
+      "grad_norm": 2.9086647033691406,
+      "learning_rate": 9.965831435079726e-05,
+      "loss": 2.3909,
+      "step": 42
+    },
+    {
+      "epoch": 0.011561073340559004,
+      "grad_norm": 2.6053147315979004,
+      "learning_rate": 9.964079200981251e-05,
+      "loss": 2.3981,
+      "step": 44
+    },
+    {
+      "epoch": 0.012086576674220776,
+      "grad_norm": 2.3066649436950684,
+      "learning_rate": 9.962326966882776e-05,
+      "loss": 2.3786,
+      "step": 46
+    },
+    {
+      "epoch": 0.01261208000788255,
+      "grad_norm": 2.5991337299346924,
+      "learning_rate": 9.960574732784301e-05,
+      "loss": 2.3309,
+      "step": 48
+    },
+    {
+      "epoch": 0.013137583341544324,
+      "grad_norm": 2.6027159690856934,
+      "learning_rate": 9.958822498685824e-05,
+      "loss": 2.3522,
+      "step": 50
+    },
+    {
+      "epoch": 0.013663086675206096,
+      "grad_norm": 2.2059361934661865,
+      "learning_rate": 9.957070264587349e-05,
+      "loss": 2.3221,
+      "step": 52
+    },
+    {
+      "epoch": 0.01418859000886787,
+      "grad_norm": 2.3481605052948,
+      "learning_rate": 9.955318030488874e-05,
+      "loss": 2.3223,
+      "step": 54
+    },
+    {
+      "epoch": 0.014714093342529641,
+      "grad_norm": 2.489518880844116,
+      "learning_rate": 9.953565796390398e-05,
+      "loss": 2.299,
+      "step": 56
+    },
+    {
+      "epoch": 0.015239596676191415,
+      "grad_norm": 1.992858648300171,
+      "learning_rate": 9.951813562291923e-05,
+      "loss": 2.2695,
+      "step": 58
+    },
+    {
+      "epoch": 0.01576510000985319,
+      "grad_norm": 2.3498594760894775,
+      "learning_rate": 9.950061328193448e-05,
+      "loss": 2.305,
+      "step": 60
+    },
+    {
+      "epoch": 0.01629060334351496,
+      "grad_norm": 2.3116061687469482,
+      "learning_rate": 9.948309094094971e-05,
+      "loss": 2.28,
+      "step": 62
+    },
+    {
+      "epoch": 0.016816106677176732,
+      "grad_norm": 1.8108229637145996,
+      "learning_rate": 9.946556859996496e-05,
+      "loss": 2.2844,
+      "step": 64
+    },
+    {
+      "epoch": 0.017341610010838508,
+      "grad_norm": 1.7650134563446045,
+      "learning_rate": 9.944804625898021e-05,
+      "loss": 2.2585,
+      "step": 66
+    },
+    {
+      "epoch": 0.01786711334450028,
+      "grad_norm": 2.000683546066284,
+      "learning_rate": 9.943052391799544e-05,
+      "loss": 2.2866,
+      "step": 68
+    },
+    {
+      "epoch": 0.01839261667816205,
+      "grad_norm": 1.7432600259780884,
+      "learning_rate": 9.941300157701069e-05,
+      "loss": 2.2551,
+      "step": 70
+    },
+    {
+      "epoch": 0.018918120011823827,
+      "grad_norm": 1.8153115510940552,
+      "learning_rate": 9.939547923602594e-05,
+      "loss": 2.2829,
+      "step": 72
+    },
+    {
+      "epoch": 0.0194436233454856,
+      "grad_norm": 1.7774672508239746,
+      "learning_rate": 9.937795689504118e-05,
+      "loss": 2.2656,
+      "step": 74
+    },
+    {
+      "epoch": 0.01996912667914737,
+      "grad_norm": 1.8277621269226074,
+      "learning_rate": 9.936043455405642e-05,
+      "loss": 2.2602,
+      "step": 76
+    },
+    {
+      "epoch": 0.020494630012809142,
+      "grad_norm": 1.8183931112289429,
+      "learning_rate": 9.934291221307167e-05,
+      "loss": 2.2581,
+      "step": 78
+    },
+    {
+      "epoch": 0.021020133346470918,
+      "grad_norm": 1.5881011486053467,
+      "learning_rate": 9.932538987208691e-05,
+      "loss": 2.2305,
+      "step": 80
+    },
+    {
+      "epoch": 0.02154563668013269,
+      "grad_norm": 1.653342843055725,
+      "learning_rate": 9.930786753110216e-05,
+      "loss": 2.2092,
+      "step": 82
+    },
+    {
+      "epoch": 0.02207114001379446,
+      "grad_norm": 1.7351162433624268,
+      "learning_rate": 9.929034519011741e-05,
+      "loss": 2.2031,
+      "step": 84
+    },
+    {
+      "epoch": 0.022596643347456237,
+      "grad_norm": 1.679905891418457,
+      "learning_rate": 9.927282284913266e-05,
+      "loss": 2.2373,
+      "step": 86
+    },
+    {
+      "epoch": 0.02312214668111801,
+      "grad_norm": 1.742342472076416,
+      "learning_rate": 9.925530050814789e-05,
+      "loss": 2.1937,
+      "step": 88
+    },
+    {
+      "epoch": 0.02364765001477978,
+      "grad_norm": 1.5579359531402588,
+      "learning_rate": 9.923777816716314e-05,
+      "loss": 2.2137,
+      "step": 90
+    },
+    {
+      "epoch": 0.024173153348441553,
+      "grad_norm": 1.548086404800415,
+      "learning_rate": 9.922025582617839e-05,
+      "loss": 2.2123,
+      "step": 92
+    },
+    {
+      "epoch": 0.024698656682103328,
+      "grad_norm": 1.6375842094421387,
+      "learning_rate": 9.920273348519362e-05,
+      "loss": 2.182,
+      "step": 94
+    },
+    {
+      "epoch": 0.0252241600157651,
+      "grad_norm": 1.6106681823730469,
+      "learning_rate": 9.918521114420887e-05,
+      "loss": 2.1649,
+      "step": 96
+    },
+    {
+      "epoch": 0.025749663349426872,
+      "grad_norm": 1.6409622430801392,
+      "learning_rate": 9.916768880322411e-05,
+      "loss": 2.1702,
+      "step": 98
+    },
+    {
+      "epoch": 0.026275166683088647,
+      "grad_norm": 2.1687161922454834,
+      "learning_rate": 9.915016646223936e-05,
+      "loss": 2.1519,
+      "step": 100
+    },
+    {
+      "epoch": 0.02680067001675042,
+      "grad_norm": 1.595245122909546,
+      "learning_rate": 9.91326441212546e-05,
+      "loss": 2.1314,
+      "step": 102
+    },
+    {
+      "epoch": 0.02732617335041219,
+      "grad_norm": 1.6276413202285767,
+      "learning_rate": 9.911512178026984e-05,
+      "loss": 2.1189,
+      "step": 104
+    },
+    {
+      "epoch": 0.027851676684073963,
+      "grad_norm": 1.7238609790802002,
+      "learning_rate": 9.909759943928509e-05,
+      "loss": 2.1617,
+      "step": 106
+    },
+    {
+      "epoch": 0.02837718001773574,
+      "grad_norm": 1.5262254476547241,
+      "learning_rate": 9.908007709830034e-05,
+      "loss": 2.1105,
+      "step": 108
+    },
+    {
+      "epoch": 0.02890268335139751,
+      "grad_norm": 1.6144453287124634,
+      "learning_rate": 9.906255475731559e-05,
+      "loss": 2.1287,
+      "step": 110
+    },
+    {
+      "epoch": 0.029428186685059282,
+      "grad_norm": 1.528970718383789,
+      "learning_rate": 9.904503241633083e-05,
+      "loss": 2.0744,
+      "step": 112
+    },
+    {
+      "epoch": 0.029953690018721058,
+      "grad_norm": 1.703235149383545,
+      "learning_rate": 9.902751007534607e-05,
+      "loss": 2.1285,
+      "step": 114
+    },
+    {
+      "epoch": 0.03047919335238283,
+      "grad_norm": 1.567522406578064,
+      "learning_rate": 9.900998773436132e-05,
+      "loss": 2.0736,
+      "step": 116
+    },
+    {
+      "epoch": 0.0310046966860446,
+      "grad_norm": 1.4350054264068604,
+      "learning_rate": 9.899246539337655e-05,
+      "loss": 2.0745,
+      "step": 118
+    },
+    {
+      "epoch": 0.03153020001970638,
+      "grad_norm": 1.438694953918457,
+      "learning_rate": 9.89749430523918e-05,
+      "loss": 2.0607,
+      "step": 120
+    },
+    {
+      "epoch": 0.032055703353368145,
+      "grad_norm": 1.3507471084594727,
+      "learning_rate": 9.895742071140705e-05,
+      "loss": 2.0601,
+      "step": 122
+    },
+    {
+      "epoch": 0.03258120668702992,
+      "grad_norm": 1.2798346281051636,
+      "learning_rate": 9.893989837042229e-05,
+      "loss": 2.0925,
+      "step": 124
+    },
+    {
+      "epoch": 0.033106710020691696,
+      "grad_norm": 1.1719093322753906,
+      "learning_rate": 9.892237602943754e-05,
+      "loss": 2.069,
+      "step": 126
+    },
+    {
+      "epoch": 0.033632213354353464,
+      "grad_norm": 1.3536409139633179,
+      "learning_rate": 9.890485368845279e-05,
+      "loss": 2.0716,
+      "step": 128
+    },
+    {
+      "epoch": 0.03415771668801524,
+      "grad_norm": 1.3446725606918335,
+      "learning_rate": 9.888733134746802e-05,
+      "loss": 2.0164,
+      "step": 130
+    },
+    {
+      "epoch": 0.034683220021677015,
+      "grad_norm": 1.322689414024353,
+      "learning_rate": 9.886980900648327e-05,
+      "loss": 2.0752,
+      "step": 132
+    },
+    {
+      "epoch": 0.035208723355338784,
+      "grad_norm": 1.3145771026611328,
+      "learning_rate": 9.885228666549852e-05,
+      "loss": 2.0852,
+      "step": 134
+    },
+    {
+      "epoch": 0.03573422668900056,
+      "grad_norm": 1.326093316078186,
+      "learning_rate": 9.883476432451376e-05,
+      "loss": 2.0414,
+      "step": 136
+    },
+    {
+      "epoch": 0.036259730022662334,
+      "grad_norm": 1.234911561012268,
+      "learning_rate": 9.881724198352901e-05,
+      "loss": 2.0627,
+      "step": 138
+    },
+    {
+      "epoch": 0.0367852333563241,
+      "grad_norm": 1.2666422128677368,
+      "learning_rate": 9.879971964254426e-05,
+      "loss": 2.039,
+      "step": 140
+    },
+    {
+      "epoch": 0.03731073668998588,
+      "grad_norm": 1.285933494567871,
+      "learning_rate": 9.87821973015595e-05,
+      "loss": 2.0428,
+      "step": 142
+    },
+    {
+      "epoch": 0.03783624002364765,
+      "grad_norm": 1.330095648765564,
+      "learning_rate": 9.876467496057473e-05,
+      "loss": 2.0458,
+      "step": 144
+    },
+    {
+      "epoch": 0.03836174335730942,
+      "grad_norm": 1.338354229927063,
+      "learning_rate": 9.874715261958998e-05,
+      "loss": 2.0588,
+      "step": 146
+    },
+    {
+      "epoch": 0.0388872466909712,
+      "grad_norm": 1.2980471849441528,
+      "learning_rate": 9.872963027860522e-05,
+      "loss": 2.092,
+      "step": 148
+    },
+    {
+      "epoch": 0.039412750024632966,
+      "grad_norm": 1.1613810062408447,
+      "learning_rate": 9.871210793762047e-05,
+      "loss": 2.0558,
+      "step": 150
+    },
+    {
+      "epoch": 0.03993825335829474,
+      "grad_norm": 1.2073849439620972,
+      "learning_rate": 9.869458559663572e-05,
+      "loss": 2.0317,
+      "step": 152
+    },
+    {
+      "epoch": 0.040463756691956516,
+      "grad_norm": 1.2727243900299072,
+      "learning_rate": 9.867706325565097e-05,
+      "loss": 2.0637,
+      "step": 154
+    },
+    {
+      "epoch": 0.040989260025618285,
+      "grad_norm": 1.1971285343170166,
+      "learning_rate": 9.86595409146662e-05,
+      "loss": 2.0245,
+      "step": 156
+    },
+    {
+      "epoch": 0.04151476335928006,
+      "grad_norm": 1.2773100137710571,
+      "learning_rate": 9.864201857368145e-05,
+      "loss": 1.9971,
+      "step": 158
+    },
+    {
+      "epoch": 0.042040266692941836,
+      "grad_norm": 1.1450934410095215,
+      "learning_rate": 9.86244962326967e-05,
+      "loss": 2.0367,
+      "step": 160
+    },
+    {
+      "epoch": 0.042565770026603604,
+      "grad_norm": 1.0777571201324463,
+      "learning_rate": 9.860697389171194e-05,
+      "loss": 2.0667,
+      "step": 162
+    },
+    {
+      "epoch": 0.04309127336026538,
+      "grad_norm": 1.1527281999588013,
+      "learning_rate": 9.858945155072719e-05,
+      "loss": 2.0556,
+      "step": 164
+    },
+    {
+      "epoch": 0.043616776693927155,
+      "grad_norm": 1.2145709991455078,
+      "learning_rate": 9.857192920974244e-05,
+      "loss": 2.0638,
+      "step": 166
+    },
+    {
+      "epoch": 0.04414228002758892,
+      "grad_norm": 1.2327994108200073,
+      "learning_rate": 9.855440686875767e-05,
+      "loss": 2.0521,
+      "step": 168
+    },
+    {
+      "epoch": 0.0446677833612507,
+      "grad_norm": 1.1262874603271484,
+      "learning_rate": 9.85368845277729e-05,
+      "loss": 2.0456,
+      "step": 170
+    },
+    {
+      "epoch": 0.045193286694912474,
+      "grad_norm": 1.261991262435913,
+      "learning_rate": 9.851936218678815e-05,
+      "loss": 2.0438,
+      "step": 172
+    },
+    {
+      "epoch": 0.04571879002857424,
+      "grad_norm": 1.1301138401031494,
+      "learning_rate": 9.85018398458034e-05,
+      "loss": 2.0385,
+      "step": 174
+    },
+    {
+      "epoch": 0.04624429336223602,
+      "grad_norm": 1.1710069179534912,
+      "learning_rate": 9.848431750481865e-05,
+      "loss": 1.9769,
+      "step": 176
+    },
+    {
+      "epoch": 0.046769796695897786,
+      "grad_norm": 1.417973279953003,
+      "learning_rate": 9.84667951638339e-05,
+      "loss": 2.0309,
+      "step": 178
+    },
+    {
+      "epoch": 0.04729530002955956,
+      "grad_norm": 1.1423699855804443,
+      "learning_rate": 9.844927282284914e-05,
+      "loss": 2.0254,
+      "step": 180
+    },
+    {
+      "epoch": 0.04782080336322134,
+      "grad_norm": 1.036130428314209,
+      "learning_rate": 9.843175048186438e-05,
+      "loss": 2.0361,
+      "step": 182
+    },
+    {
+      "epoch": 0.048346306696883105,
+      "grad_norm": 1.1310175657272339,
+      "learning_rate": 9.841422814087962e-05,
+      "loss": 2.0172,
+      "step": 184
+    },
+    {
+      "epoch": 0.04887181003054488,
+      "grad_norm": 1.1112390756607056,
+      "learning_rate": 9.839670579989487e-05,
+      "loss": 2.021,
+      "step": 186
+    },
+    {
+      "epoch": 0.049397313364206656,
+      "grad_norm": 1.1196744441986084,
+      "learning_rate": 9.837918345891012e-05,
+      "loss": 2.0398,
+      "step": 188
+    },
+    {
+      "epoch": 0.049922816697868425,
+      "grad_norm": 1.1641274690628052,
+      "learning_rate": 9.836166111792537e-05,
+      "loss": 2.0211,
+      "step": 190
+    },
+    {
+      "epoch": 0.0504483200315302,
+      "grad_norm": 1.231698989868164,
+      "learning_rate": 9.834413877694061e-05,
+      "loss": 1.9938,
+      "step": 192
+    },
+    {
+      "epoch": 0.050973823365191975,
+      "grad_norm": 1.116532325744629,
+      "learning_rate": 9.832661643595585e-05,
+      "loss": 1.9916,
+      "step": 194
+    },
+    {
+      "epoch": 0.051499326698853744,
+      "grad_norm": 1.1912415027618408,
+      "learning_rate": 9.830909409497108e-05,
+      "loss": 2.0206,
+      "step": 196
+    },
+    {
+      "epoch": 0.05202483003251552,
+      "grad_norm": 1.0664820671081543,
+      "learning_rate": 9.829157175398633e-05,
+      "loss": 2.0244,
+      "step": 198
+    },
+    {
+      "epoch": 0.052550333366177295,
+      "grad_norm": 1.1066045761108398,
+      "learning_rate": 9.827404941300158e-05,
+      "loss": 2.0005,
+      "step": 200
+    },
+    {
+      "epoch": 0.05307583669983906,
+      "grad_norm": 1.1267125606536865,
+      "learning_rate": 9.825652707201683e-05,
+      "loss": 2.0185,
+      "step": 202
+    },
+    {
+      "epoch": 0.05360134003350084,
+      "grad_norm": 1.14983069896698,
+      "learning_rate": 9.823900473103207e-05,
+      "loss": 1.9726,
+      "step": 204
+    },
+    {
+      "epoch": 0.054126843367162614,
+      "grad_norm": 1.2234214544296265,
+      "learning_rate": 9.822148239004732e-05,
+      "loss": 2.0497,
+      "step": 206
+    },
+    {
+      "epoch": 0.05465234670082438,
+      "grad_norm": 1.1713297367095947,
+      "learning_rate": 9.820396004906255e-05,
+      "loss": 2.0167,
+      "step": 208
+    },
+    {
+      "epoch": 0.05517785003448616,
+      "grad_norm": 1.2977114915847778,
+      "learning_rate": 9.81864377080778e-05,
+      "loss": 1.9795,
+      "step": 210
+    },
+    {
+      "epoch": 0.055703353368147926,
+      "grad_norm": 1.1145280599594116,
+      "learning_rate": 9.816891536709305e-05,
+      "loss": 2.0099,
+      "step": 212
+    },
+    {
+      "epoch": 0.0562288567018097,
+      "grad_norm": 1.126206874847412,
+      "learning_rate": 9.81513930261083e-05,
+      "loss": 2.0053,
+      "step": 214
+    },
+    {
+      "epoch": 0.05675436003547148,
+      "grad_norm": 1.1073702573776245,
+      "learning_rate": 9.813387068512355e-05,
+      "loss": 1.985,
+      "step": 216
+    },
+    {
+      "epoch": 0.057279863369133245,
+      "grad_norm": 1.2039167881011963,
+      "learning_rate": 9.811634834413879e-05,
+      "loss": 1.9898,
+      "step": 218
+    },
+    {
+      "epoch": 0.05780536670279502,
+      "grad_norm": 1.2644699811935425,
+      "learning_rate": 9.809882600315403e-05,
+      "loss": 1.9904,
+      "step": 220
+    },
+    {
+      "epoch": 0.058330870036456796,
+      "grad_norm": 0.9593138694763184,
+      "learning_rate": 9.808130366216926e-05,
+      "loss": 2.0065,
+      "step": 222
+    },
+    {
+      "epoch": 0.058856373370118564,
+      "grad_norm": 0.9157779216766357,
+      "learning_rate": 9.806378132118451e-05,
+      "loss": 1.9853,
+      "step": 224
+    },
+    {
+      "epoch": 0.05938187670378034,
+      "grad_norm": 1.0334917306900024,
+      "learning_rate": 9.804625898019976e-05,
+      "loss": 2.0129,
+      "step": 226
+    },
+    {
+      "epoch": 0.059907380037442115,
+      "grad_norm": 1.0476865768432617,
+      "learning_rate": 9.8028736639215e-05,
+      "loss": 1.9816,
+      "step": 228
+    },
+    {
+      "epoch": 0.060432883371103883,
+      "grad_norm": 1.1626946926116943,
+      "learning_rate": 9.801121429823025e-05,
+      "loss": 2.0278,
+      "step": 230
+    },
+    {
+      "epoch": 0.06095838670476566,
+      "grad_norm": 1.3142833709716797,
+      "learning_rate": 9.79936919572455e-05,
+      "loss": 1.998,
+      "step": 232
+    },
+    {
+      "epoch": 0.061483890038427434,
+      "grad_norm": 1.195894718170166,
+      "learning_rate": 9.797616961626073e-05,
+      "loss": 1.9813,
+      "step": 234
+    },
+    {
+      "epoch": 0.0620093933720892,
+      "grad_norm": 1.0905667543411255,
+      "learning_rate": 9.795864727527598e-05,
+      "loss": 1.9725,
+      "step": 236
+    },
+    {
+      "epoch": 0.06253489670575098,
+      "grad_norm": 1.0762920379638672,
+      "learning_rate": 9.794112493429123e-05,
+      "loss": 2.0036,
+      "step": 238
+    },
+    {
+      "epoch": 0.06306040003941275,
+      "grad_norm": 0.9937522411346436,
+      "learning_rate": 9.792360259330648e-05,
+      "loss": 2.0208,
+      "step": 240
+    },
+    {
+      "epoch": 0.06358590337307453,
+      "grad_norm": 1.3950233459472656,
+      "learning_rate": 9.790608025232172e-05,
+      "loss": 1.9667,
+      "step": 242
+    },
+    {
+      "epoch": 0.06411140670673629,
+      "grad_norm": 1.0673043727874756,
+      "learning_rate": 9.788855791133697e-05,
+      "loss": 1.984,
+      "step": 244
+    },
+    {
+      "epoch": 0.06463691004039807,
+      "grad_norm": 1.0824543237686157,
+      "learning_rate": 9.78710355703522e-05,
+      "loss": 2.0271,
+      "step": 246
+    },
+    {
+      "epoch": 0.06516241337405984,
+      "grad_norm": 1.2005363702774048,
+      "learning_rate": 9.785351322936744e-05,
+      "loss": 1.9912,
+      "step": 248
+    },
+    {
+      "epoch": 0.06568791670772162,
+      "grad_norm": 1.0987967252731323,
+      "learning_rate": 9.783599088838269e-05,
+      "loss": 1.9886,
+      "step": 250
+    },
+    {
+      "epoch": 0.06621342004138339,
+      "grad_norm": 1.0441079139709473,
+      "learning_rate": 9.781846854739793e-05,
+      "loss": 1.9888,
+      "step": 252
+    },
+    {
+      "epoch": 0.06673892337504517,
+      "grad_norm": 1.3533669710159302,
+      "learning_rate": 9.780094620641318e-05,
+      "loss": 2.0232,
+      "step": 254
+    },
+    {
+      "epoch": 0.06726442670870693,
+      "grad_norm": 0.9276557564735413,
+      "learning_rate": 9.778342386542843e-05,
+      "loss": 1.969,
+      "step": 256
+    },
+    {
+      "epoch": 0.0677899300423687,
+      "grad_norm": 1.143988847732544,
+      "learning_rate": 9.776590152444368e-05,
+      "loss": 1.9643,
+      "step": 258
+    },
+    {
+      "epoch": 0.06831543337603048,
+      "grad_norm": 0.9717089533805847,
+      "learning_rate": 9.774837918345891e-05,
+      "loss": 2.0169,
+      "step": 260
+    },
+    {
+      "epoch": 0.06884093670969225,
+      "grad_norm": 1.328565001487732,
+      "learning_rate": 9.773085684247416e-05,
+      "loss": 1.9733,
+      "step": 262
+    },
+    {
+      "epoch": 0.06936644004335403,
+      "grad_norm": 1.0031944513320923,
+      "learning_rate": 9.77133345014894e-05,
+      "loss": 2.0026,
+      "step": 264
+    },
+    {
+      "epoch": 0.06989194337701579,
+      "grad_norm": 0.9560984969139099,
+      "learning_rate": 9.769581216050465e-05,
+      "loss": 1.9893,
+      "step": 266
+    },
+    {
+      "epoch": 0.07041744671067757,
+      "grad_norm": 1.1324944496154785,
+      "learning_rate": 9.76782898195199e-05,
+      "loss": 1.9499,
+      "step": 268
+    },
+    {
+      "epoch": 0.07094295004433934,
+      "grad_norm": 1.1018491983413696,
+      "learning_rate": 9.766076747853515e-05,
+      "loss": 1.9915,
+      "step": 270
+    },
+    {
+      "epoch": 0.07146845337800112,
+      "grad_norm": 1.1043239831924438,
+      "learning_rate": 9.764324513755038e-05,
+      "loss": 2.0021,
+      "step": 272
+    },
+    {
+      "epoch": 0.0719939567116629,
+      "grad_norm": 1.1171072721481323,
+      "learning_rate": 9.762572279656562e-05,
+      "loss": 1.9665,
+      "step": 274
+    },
+    {
+      "epoch": 0.07251946004532467,
+      "grad_norm": 1.040802001953125,
+      "learning_rate": 9.760820045558086e-05,
+      "loss": 2.0249,
+      "step": 276
+    },
+    {
+      "epoch": 0.07304496337898643,
+      "grad_norm": 1.0344669818878174,
+      "learning_rate": 9.759067811459611e-05,
+      "loss": 2.0102,
+      "step": 278
+    },
+    {
+      "epoch": 0.0735704667126482,
+      "grad_norm": 0.9199603199958801,
+      "learning_rate": 9.757315577361136e-05,
+      "loss": 1.9836,
+      "step": 280
+    },
+    {
+      "epoch": 0.07409597004630998,
+      "grad_norm": 1.129347562789917,
+      "learning_rate": 9.75556334326266e-05,
+      "loss": 1.9141,
+      "step": 282
+    },
+    {
+      "epoch": 0.07462147337997176,
+      "grad_norm": 0.9745813012123108,
+      "learning_rate": 9.753811109164185e-05,
+      "loss": 1.9807,
+      "step": 284
+    },
+    {
+      "epoch": 0.07514697671363353,
+      "grad_norm": 1.2225230932235718,
+      "learning_rate": 9.752058875065709e-05,
+      "loss": 1.9786,
+      "step": 286
+    },
+    {
+      "epoch": 0.0756724800472953,
+      "grad_norm": 1.7652974128723145,
+      "learning_rate": 9.750306640967234e-05,
+      "loss": 1.9575,
+      "step": 288
+    },
+    {
+      "epoch": 0.07619798338095707,
+      "grad_norm": 1.075140118598938,
+      "learning_rate": 9.748554406868758e-05,
+      "loss": 2.0034,
+      "step": 290
+    },
+    {
+      "epoch": 0.07672348671461884,
+      "grad_norm": 1.0169627666473389,
+      "learning_rate": 9.746802172770283e-05,
+      "loss": 2.0072,
+      "step": 292
+    },
+    {
+      "epoch": 0.07724899004828062,
+      "grad_norm": 1.0737385749816895,
+      "learning_rate": 9.745049938671808e-05,
+      "loss": 1.9446,
+      "step": 294
+    },
+    {
+      "epoch": 0.0777744933819424,
+      "grad_norm": 1.2098944187164307,
+      "learning_rate": 9.743297704573331e-05,
+      "loss": 1.9875,
+      "step": 296
+    },
+    {
+      "epoch": 0.07829999671560417,
+      "grad_norm": 1.030888557434082,
+      "learning_rate": 9.741545470474856e-05,
+      "loss": 2.0323,
+      "step": 298
+    },
+    {
+      "epoch": 0.07882550004926593,
+      "grad_norm": 1.1690922975540161,
+      "learning_rate": 9.73979323637638e-05,
+      "loss": 1.9544,
+      "step": 300
+    },
+    {
+      "epoch": 0.0793510033829277,
+      "grad_norm": 0.9702273607254028,
+      "learning_rate": 9.738041002277904e-05,
+      "loss": 1.954,
+      "step": 302
+    },
+    {
+      "epoch": 0.07987650671658948,
+      "grad_norm": 1.1210618019104004,
+      "learning_rate": 9.736288768179429e-05,
+      "loss": 1.9786,
+      "step": 304
+    },
+    {
+      "epoch": 0.08040201005025126,
+      "grad_norm": 1.0246062278747559,
+      "learning_rate": 9.734536534080954e-05,
+      "loss": 1.9777,
+      "step": 306
+    },
+    {
+      "epoch": 0.08092751338391303,
+      "grad_norm": 0.9073227047920227,
+      "learning_rate": 9.732784299982478e-05,
+      "loss": 1.9737,
+      "step": 308
+    },
+    {
+      "epoch": 0.08145301671757481,
+      "grad_norm": 0.9827994704246521,
+      "learning_rate": 9.731032065884003e-05,
+      "loss": 1.9794,
+      "step": 310
+    },
+    {
+      "epoch": 0.08197852005123657,
+      "grad_norm": 0.9744251370429993,
+      "learning_rate": 9.729279831785527e-05,
+      "loss": 1.9771,
+      "step": 312
+    },
+    {
+      "epoch": 0.08250402338489835,
+      "grad_norm": 1.0426684617996216,
+      "learning_rate": 9.727527597687051e-05,
+      "loss": 1.9522,
+      "step": 314
+    },
+    {
+      "epoch": 0.08302952671856012,
+      "grad_norm": 1.0296825170516968,
+      "learning_rate": 9.725775363588576e-05,
+      "loss": 1.9727,
+      "step": 316
+    },
+    {
+      "epoch": 0.0835550300522219,
+      "grad_norm": 1.2643887996673584,
+      "learning_rate": 9.724023129490101e-05,
+      "loss": 1.9414,
+      "step": 318
+    },
+    {
+      "epoch": 0.08408053338588367,
+      "grad_norm": 1.0346665382385254,
+      "learning_rate": 9.722270895391626e-05,
+      "loss": 1.9964,
+      "step": 320
+    },
+    {
+      "epoch": 0.08460603671954543,
+      "grad_norm": 0.9016798734664917,
+      "learning_rate": 9.720518661293149e-05,
+      "loss": 1.9762,
+      "step": 322
+    },
+    {
+      "epoch": 0.08513154005320721,
+      "grad_norm": 1.0269689559936523,
+      "learning_rate": 9.718766427194674e-05,
+      "loss": 1.9969,
+      "step": 324
+    },
+    {
+      "epoch": 0.08565704338686898,
+      "grad_norm": 1.0594022274017334,
+      "learning_rate": 9.717014193096197e-05,
+      "loss": 1.9624,
+      "step": 326
+    },
+    {
+      "epoch": 0.08618254672053076,
+      "grad_norm": 1.1567680835723877,
+      "learning_rate": 9.715261958997722e-05,
+      "loss": 1.959,
+      "step": 328
+    },
+    {
+      "epoch": 0.08670805005419253,
+      "grad_norm": 0.870560884475708,
+      "learning_rate": 9.713509724899247e-05,
+      "loss": 1.9569,
+      "step": 330
+    },
+    {
+      "epoch": 0.08723355338785431,
+      "grad_norm": 1.0716711282730103,
+      "learning_rate": 9.711757490800771e-05,
+      "loss": 1.951,
+      "step": 332
+    },
+    {
+      "epoch": 0.08775905672151607,
+      "grad_norm": 1.1096898317337036,
+      "learning_rate": 9.710005256702296e-05,
+      "loss": 1.9796,
+      "step": 334
+    },
+    {
+      "epoch": 0.08828456005517785,
+      "grad_norm": 0.8860819339752197,
+      "learning_rate": 9.708253022603821e-05,
+      "loss": 1.9591,
+      "step": 336
+    },
+    {
+      "epoch": 0.08881006338883962,
+      "grad_norm": 0.9616803526878357,
+      "learning_rate": 9.706500788505344e-05,
+      "loss": 1.9743,
+      "step": 338
+    },
+    {
+      "epoch": 0.0893355667225014,
+      "grad_norm": 0.8491392135620117,
+      "learning_rate": 9.704748554406869e-05,
+      "loss": 1.959,
+      "step": 340
+    },
+    {
+      "epoch": 0.08986107005616317,
+      "grad_norm": 1.068489670753479,
+      "learning_rate": 9.702996320308394e-05,
+      "loss": 1.971,
+      "step": 342
+    },
+    {
+      "epoch": 0.09038657338982495,
+      "grad_norm": 0.8963595032691956,
+      "learning_rate": 9.701244086209919e-05,
+      "loss": 1.9239,
+      "step": 344
+    },
+    {
+      "epoch": 0.09091207672348671,
+      "grad_norm": 1.232599139213562,
+      "learning_rate": 9.699491852111443e-05,
+      "loss": 1.9282,
+      "step": 346
+    },
+    {
+      "epoch": 0.09143758005714848,
+      "grad_norm": 1.0067182779312134,
+      "learning_rate": 9.697739618012967e-05,
+      "loss": 1.9539,
+      "step": 348
+    },
+    {
+      "epoch": 0.09196308339081026,
+      "grad_norm": 0.927861750125885,
+      "learning_rate": 9.695987383914491e-05,
+      "loss": 1.9518,
+      "step": 350
+    },
+    {
+      "epoch": 0.09248858672447204,
+      "grad_norm": 1.1443390846252441,
+      "learning_rate": 9.694235149816015e-05,
+      "loss": 1.9527,
+      "step": 352
+    },
+    {
+      "epoch": 0.09301409005813381,
+      "grad_norm": 1.3024427890777588,
+      "learning_rate": 9.69248291571754e-05,
+      "loss": 1.9276,
+      "step": 354
+    },
+    {
+      "epoch": 0.09353959339179557,
+      "grad_norm": 1.0598125457763672,
+      "learning_rate": 9.690730681619064e-05,
+      "loss": 1.9728,
+      "step": 356
+    },
+    {
+      "epoch": 0.09406509672545735,
+      "grad_norm": 1.0384907722473145,
+      "learning_rate": 9.688978447520589e-05,
+      "loss": 1.9543,
+      "step": 358
+    },
+    {
+      "epoch": 0.09459060005911912,
+      "grad_norm": 1.4060821533203125,
+      "learning_rate": 9.687226213422114e-05,
+      "loss": 1.9005,
+      "step": 360
+    },
+    {
+      "epoch": 0.0951161033927809,
+      "grad_norm": 0.9630605578422546,
+      "learning_rate": 9.685473979323639e-05,
+      "loss": 1.9306,
+      "step": 362
+    },
+    {
+      "epoch": 0.09564160672644267,
+      "grad_norm": 0.9025930166244507,
+      "learning_rate": 9.683721745225162e-05,
+      "loss": 1.9391,
+      "step": 364
+    },
+    {
+      "epoch": 0.09616711006010445,
+      "grad_norm": 1.0383387804031372,
+      "learning_rate": 9.681969511126687e-05,
+      "loss": 1.9506,
+      "step": 366
+    },
+    {
+      "epoch": 0.09669261339376621,
+      "grad_norm": 0.9397739768028259,
+      "learning_rate": 9.680217277028212e-05,
+      "loss": 1.9132,
+      "step": 368
+    },
+    {
+      "epoch": 0.09721811672742799,
+      "grad_norm": 0.9472708106040955,
+      "learning_rate": 9.678465042929736e-05,
+      "loss": 1.9731,
+      "step": 370
+    },
+    {
+      "epoch": 0.09774362006108976,
+      "grad_norm": 0.8391740322113037,
+      "learning_rate": 9.676712808831261e-05,
+      "loss": 1.9956,
+      "step": 372
+    },
+    {
+      "epoch": 0.09826912339475154,
+      "grad_norm": 1.160340666770935,
+      "learning_rate": 9.674960574732785e-05,
+      "loss": 1.9576,
+      "step": 374
+    },
+    {
+      "epoch": 0.09879462672841331,
+      "grad_norm": 1.1132557392120361,
+      "learning_rate": 9.673208340634309e-05,
+      "loss": 1.9407,
+      "step": 376
+    },
+    {
+      "epoch": 0.09932013006207509,
+      "grad_norm": 0.9549061059951782,
+      "learning_rate": 9.671456106535834e-05,
+      "loss": 1.96,
+      "step": 378
+    },
+    {
+      "epoch": 0.09984563339573685,
+      "grad_norm": 1.953350305557251,
+      "learning_rate": 9.669703872437357e-05,
+      "loss": 1.9146,
+      "step": 380
+    },
+    {
+      "epoch": 0.10037113672939862,
+      "grad_norm": 0.8253780603408813,
+      "learning_rate": 9.667951638338882e-05,
+      "loss": 1.9576,
+      "step": 382
+    },
+    {
+      "epoch": 0.1008966400630604,
+      "grad_norm": 1.76423180103302,
+      "learning_rate": 9.666199404240407e-05,
+      "loss": 1.9411,
+      "step": 384
+    },
+    {
+      "epoch": 0.10142214339672218,
+      "grad_norm": 1.2772194147109985,
+      "learning_rate": 9.664447170141932e-05,
+      "loss": 1.939,
+      "step": 386
+    },
+    {
+      "epoch": 0.10194764673038395,
+      "grad_norm": 1.1129158735275269,
+      "learning_rate": 9.662694936043456e-05,
+      "loss": 1.9587,
+      "step": 388
+    },
+    {
+      "epoch": 0.10247315006404571,
+      "grad_norm": 1.5891733169555664,
+      "learning_rate": 9.660942701944981e-05,
+      "loss": 1.9677,
+      "step": 390
+    },
+    {
+      "epoch": 0.10299865339770749,
+      "grad_norm": 0.8642210364341736,
+      "learning_rate": 9.659190467846505e-05,
+      "loss": 1.9356,
+      "step": 392
+    },
+    {
+      "epoch": 0.10352415673136926,
+      "grad_norm": 1.0561408996582031,
+      "learning_rate": 9.65743823374803e-05,
+      "loss": 1.961,
+      "step": 394
+    },
+    {
+      "epoch": 0.10404966006503104,
+      "grad_norm": 0.9040923118591309,
+      "learning_rate": 9.655685999649554e-05,
+      "loss": 1.9217,
+      "step": 396
+    },
+    {
+      "epoch": 0.10457516339869281,
+      "grad_norm": 0.9783763289451599,
+      "learning_rate": 9.653933765551078e-05,
+      "loss": 1.9042,
+      "step": 398
+    },
+    {
+      "epoch": 0.10510066673235459,
+      "grad_norm": 1.0281099081039429,
+      "learning_rate": 9.652181531452602e-05,
+      "loss": 1.8989,
+      "step": 400
+    },
+    {
+      "epoch": 0.10510066673235459,
+      "eval_loss": 1.89004647731781,
+      "eval_runtime": 487.2614,
+      "eval_samples_per_second": 249.946,
+      "eval_steps_per_second": 31.244,
+      "step": 400
+    },
+    {
+      "epoch": 0.10562617006601635,
+      "grad_norm": 0.9645543694496155,
+      "learning_rate": 9.650429297354127e-05,
+      "loss": 1.9487,
+      "step": 402
+    },
+    {
+      "epoch": 0.10615167339967813,
+      "grad_norm": 1.1213475465774536,
+      "learning_rate": 9.648677063255652e-05,
+      "loss": 1.963,
+      "step": 404
+    },
+    {
+      "epoch": 0.1066771767333399,
+      "grad_norm": 0.8866876363754272,
+      "learning_rate": 9.646924829157175e-05,
+      "loss": 1.925,
+      "step": 406
+    },
+    {
+      "epoch": 0.10720268006700168,
+      "grad_norm": 0.9458855390548706,
+      "learning_rate": 9.6451725950587e-05,
+      "loss": 1.9641,
+      "step": 408
+    },
+    {
+      "epoch": 0.10772818340066345,
+      "grad_norm": 0.9262669086456299,
+      "learning_rate": 9.643420360960225e-05,
+      "loss": 1.9257,
+      "step": 410
+    },
+    {
+      "epoch": 0.10825368673432523,
+      "grad_norm": 0.9147223830223083,
+      "learning_rate": 9.64166812686175e-05,
+      "loss": 1.9242,
+      "step": 412
+    },
+    {
+      "epoch": 0.10877919006798699,
+      "grad_norm": 1.1678857803344727,
+      "learning_rate": 9.639915892763274e-05,
+      "loss": 1.9474,
+      "step": 414
+    },
+    {
+      "epoch": 0.10930469340164876,
+      "grad_norm": 1.0557537078857422,
+      "learning_rate": 9.638163658664799e-05,
+      "loss": 1.9527,
+      "step": 416
+    },
+    {
+      "epoch": 0.10983019673531054,
+      "grad_norm": 1.047229528427124,
+      "learning_rate": 9.636411424566322e-05,
+      "loss": 1.9116,
+      "step": 418
+    },
+    {
+      "epoch": 0.11035570006897231,
+      "grad_norm": 0.9046667814254761,
+      "learning_rate": 9.634659190467847e-05,
+      "loss": 1.914,
+      "step": 420
+    },
+    {
+      "epoch": 0.11088120340263409,
+      "grad_norm": 1.019170880317688,
+      "learning_rate": 9.632906956369372e-05,
+      "loss": 1.9667,
+      "step": 422
+    },
+    {
+      "epoch": 0.11140670673629585,
+      "grad_norm": 1.061903953552246,
+      "learning_rate": 9.631154722270895e-05,
+      "loss": 1.8915,
+      "step": 424
+    },
+    {
+      "epoch": 0.11193221006995763,
+      "grad_norm": 1.1627757549285889,
+      "learning_rate": 9.62940248817242e-05,
+      "loss": 1.9272,
+      "step": 426
+    },
+    {
+      "epoch": 0.1124577134036194,
+      "grad_norm": 0.9440149068832397,
+      "learning_rate": 9.627650254073945e-05,
+      "loss": 1.9686,
+      "step": 428
+    },
+    {
+      "epoch": 0.11298321673728118,
+      "grad_norm": 1.0250455141067505,
+      "learning_rate": 9.62589801997547e-05,
+      "loss": 1.937,
+      "step": 430
+    },
+    {
+      "epoch": 0.11350872007094295,
+      "grad_norm": 0.8273470401763916,
+      "learning_rate": 9.624145785876993e-05,
+      "loss": 1.9411,
+      "step": 432
+    },
+    {
+      "epoch": 0.11403422340460473,
+      "grad_norm": 1.4610145092010498,
+      "learning_rate": 9.622393551778518e-05,
+      "loss": 1.954,
+      "step": 434
+    },
+    {
+      "epoch": 0.11455972673826649,
+      "grad_norm": 0.9790822267532349,
+      "learning_rate": 9.620641317680042e-05,
+      "loss": 1.9255,
+      "step": 436
+    },
+    {
+      "epoch": 0.11508523007192827,
+      "grad_norm": 1.1833688020706177,
+      "learning_rate": 9.618889083581567e-05,
+      "loss": 1.9031,
+      "step": 438
+    },
+    {
+      "epoch": 0.11561073340559004,
+      "grad_norm": 0.879008948802948,
+      "learning_rate": 9.617136849483092e-05,
+      "loss": 1.937,
+      "step": 440
+    },
+    {
+      "epoch": 0.11613623673925182,
+      "grad_norm": 1.2231090068817139,
+      "learning_rate": 9.615384615384617e-05,
+      "loss": 1.9649,
+      "step": 442
+    },
+    {
+      "epoch": 0.11666174007291359,
+      "grad_norm": 1.1259514093399048,
+      "learning_rate": 9.61363238128614e-05,
+      "loss": 1.8921,
+      "step": 444
+    },
+    {
+      "epoch": 0.11718724340657537,
+      "grad_norm": 1.294339656829834,
+      "learning_rate": 9.611880147187665e-05,
+      "loss": 1.9084,
+      "step": 446
+    },
+    {
+      "epoch": 0.11771274674023713,
+      "grad_norm": 0.9451860785484314,
+      "learning_rate": 9.61012791308919e-05,
+      "loss": 1.9441,
+      "step": 448
+    },
+    {
+      "epoch": 0.1182382500738989,
+      "grad_norm": 0.9239658117294312,
+      "learning_rate": 9.608375678990713e-05,
+      "loss": 1.9291,
+      "step": 450
+    },
+    {
+      "epoch": 0.11876375340756068,
+      "grad_norm": 1.0765795707702637,
+      "learning_rate": 9.606623444892238e-05,
+      "loss": 1.9249,
+      "step": 452
+    },
+    {
+      "epoch": 0.11928925674122245,
+      "grad_norm": 0.919032633304596,
+      "learning_rate": 9.604871210793763e-05,
+      "loss": 1.9399,
+      "step": 454
+    },
+    {
+      "epoch": 0.11981476007488423,
+      "grad_norm": 0.8893537521362305,
+      "learning_rate": 9.603118976695287e-05,
+      "loss": 1.8985,
+      "step": 456
+    },
+    {
+      "epoch": 0.12034026340854599,
+      "grad_norm": 1.4801400899887085,
+      "learning_rate": 9.601366742596811e-05,
+      "loss": 1.8984,
+      "step": 458
+    },
+    {
+      "epoch": 0.12086576674220777,
+      "grad_norm": 0.8458815217018127,
+      "learning_rate": 9.599614508498335e-05,
+      "loss": 1.9365,
+      "step": 460
+    },
+    {
+      "epoch": 0.12139127007586954,
+      "grad_norm": 1.32188081741333,
+      "learning_rate": 9.59786227439986e-05,
+      "loss": 1.914,
+      "step": 462
+    },
+    {
+      "epoch": 0.12191677340953132,
+      "grad_norm": 0.8433484435081482,
+      "learning_rate": 9.596110040301385e-05,
+      "loss": 1.9746,
+      "step": 464
+    },
+    {
+      "epoch": 0.1224422767431931,
+      "grad_norm": 1.1092686653137207,
+      "learning_rate": 9.59435780620291e-05,
+      "loss": 1.9672,
+      "step": 466
+    },
+    {
+      "epoch": 0.12296778007685487,
+      "grad_norm": 0.9566219449043274,
+      "learning_rate": 9.592605572104434e-05,
+      "loss": 1.9225,
+      "step": 468
+    },
+    {
+      "epoch": 0.12349328341051663,
+      "grad_norm": 1.2604984045028687,
+      "learning_rate": 9.590853338005958e-05,
+      "loss": 1.9309,
+      "step": 470
+    },
+    {
+      "epoch": 0.1240187867441784,
+      "grad_norm": 1.2344852685928345,
+      "learning_rate": 9.589101103907483e-05,
+      "loss": 1.9236,
+      "step": 472
+    },
+    {
+      "epoch": 0.12454429007784018,
+      "grad_norm": 1.2368342876434326,
+      "learning_rate": 9.587348869809007e-05,
+      "loss": 1.9058,
+      "step": 474
+    },
+    {
+      "epoch": 0.12506979341150196,
+      "grad_norm": 1.3377318382263184,
+      "learning_rate": 9.585596635710531e-05,
+      "loss": 1.9345,
+      "step": 476
+    },
+    {
+      "epoch": 0.12559529674516373,
+      "grad_norm": 0.981948733329773,
+      "learning_rate": 9.583844401612056e-05,
+      "loss": 1.9253,
+      "step": 478
+    },
+    {
+      "epoch": 0.1261208000788255,
+      "grad_norm": 1.3136436939239502,
+      "learning_rate": 9.58209216751358e-05,
+      "loss": 1.933,
+      "step": 480
+    },
+    {
+      "epoch": 0.12664630341248728,
+      "grad_norm": 0.9691250324249268,
+      "learning_rate": 9.580339933415105e-05,
+      "loss": 1.9274,
+      "step": 482
+    },
+    {
+      "epoch": 0.12717180674614906,
+      "grad_norm": 1.4272805452346802,
+      "learning_rate": 9.578587699316628e-05,
+      "loss": 1.9242,
+      "step": 484
+    },
+    {
+      "epoch": 0.1276973100798108,
+      "grad_norm": 1.0407310724258423,
+      "learning_rate": 9.576835465218153e-05,
+      "loss": 1.9399,
+      "step": 486
+    },
+    {
+      "epoch": 0.12822281341347258,
+      "grad_norm": 0.8317910432815552,
+      "learning_rate": 9.575083231119678e-05,
+      "loss": 1.8733,
+      "step": 488
+    },
+    {
+      "epoch": 0.12874831674713436,
+      "grad_norm": 1.1145474910736084,
+      "learning_rate": 9.573330997021203e-05,
+      "loss": 1.9443,
+      "step": 490
+    },
+    {
+      "epoch": 0.12927382008079613,
+      "grad_norm": 1.0935791730880737,
+      "learning_rate": 9.571578762922728e-05,
+      "loss": 1.9324,
+      "step": 492
+    },
+    {
+      "epoch": 0.1297993234144579,
+      "grad_norm": 1.340178370475769,
+      "learning_rate": 9.569826528824252e-05,
+      "loss": 1.9046,
+      "step": 494
+    },
+    {
+      "epoch": 0.13032482674811968,
+      "grad_norm": 0.8448821902275085,
+      "learning_rate": 9.568074294725776e-05,
+      "loss": 1.9172,
+      "step": 496
+    },
+    {
+      "epoch": 0.13085033008178146,
+      "grad_norm": 1.1193829774856567,
+      "learning_rate": 9.5663220606273e-05,
+      "loss": 1.9471,
+      "step": 498
+    },
+    {
+      "epoch": 0.13137583341544323,
+      "grad_norm": 1.0313880443572998,
+      "learning_rate": 9.564569826528824e-05,
+      "loss": 1.9027,
+      "step": 500
+    },
+    {
+      "epoch": 0.131901336749105,
+      "grad_norm": 1.037778615951538,
+      "learning_rate": 9.562817592430349e-05,
+      "loss": 1.9375,
+      "step": 502
+    },
+    {
+      "epoch": 0.13242684008276678,
+      "grad_norm": 1.1615263223648071,
+      "learning_rate": 9.561065358331873e-05,
+      "loss": 1.9226,
+      "step": 504
+    },
+    {
+      "epoch": 0.13295234341642856,
+      "grad_norm": 1.5699961185455322,
+      "learning_rate": 9.559313124233398e-05,
+      "loss": 1.9097,
+      "step": 506
+    },
+    {
+      "epoch": 0.13347784675009033,
+      "grad_norm": 2.0286099910736084,
+      "learning_rate": 9.557560890134923e-05,
+      "loss": 1.9207,
+      "step": 508
+    },
+    {
+      "epoch": 0.13400335008375208,
+      "grad_norm": 1.0447641611099243,
+      "learning_rate": 9.555808656036446e-05,
+      "loss": 1.9415,
+      "step": 510
+    },
+    {
+      "epoch": 0.13452885341741386,
+      "grad_norm": 1.357127070426941,
+      "learning_rate": 9.554056421937971e-05,
+      "loss": 1.9165,
+      "step": 512
+    },
+    {
+      "epoch": 0.13505435675107563,
+      "grad_norm": 1.3550409078598022,
+      "learning_rate": 9.552304187839496e-05,
+      "loss": 1.9043,
+      "step": 514
+    },
+    {
+      "epoch": 0.1355798600847374,
+      "grad_norm": 1.1390573978424072,
+      "learning_rate": 9.55055195374102e-05,
+      "loss": 1.8865,
+      "step": 516
+    },
+    {
+      "epoch": 0.13610536341839918,
+      "grad_norm": 1.2006975412368774,
+      "learning_rate": 9.548799719642545e-05,
+      "loss": 1.9478,
+      "step": 518
+    },
+    {
+      "epoch": 0.13663086675206096,
+      "grad_norm": 0.9007882475852966,
+      "learning_rate": 9.54704748554407e-05,
+      "loss": 1.8973,
+      "step": 520
+    },
+    {
+      "epoch": 0.13715637008572273,
+      "grad_norm": 1.0151760578155518,
+      "learning_rate": 9.545295251445593e-05,
+      "loss": 1.9178,
+      "step": 522
+    },
+    {
+      "epoch": 0.1376818734193845,
+      "grad_norm": 1.0484004020690918,
+      "learning_rate": 9.543543017347118e-05,
+      "loss": 1.9464,
+      "step": 524
+    },
+    {
+      "epoch": 0.13820737675304628,
+      "grad_norm": 0.9942461848258972,
+      "learning_rate": 9.541790783248642e-05,
+      "loss": 1.8783,
+      "step": 526
+    },
+    {
+      "epoch": 0.13873288008670806,
+      "grad_norm": 1.2193222045898438,
+      "learning_rate": 9.540038549150166e-05,
+      "loss": 1.911,
+      "step": 528
+    },
+    {
+      "epoch": 0.13925838342036984,
+      "grad_norm": 0.9065030813217163,
+      "learning_rate": 9.538286315051691e-05,
+      "loss": 1.9214,
+      "step": 530
+    },
+    {
+      "epoch": 0.13978388675403158,
+      "grad_norm": 1.1993494033813477,
+      "learning_rate": 9.536534080953216e-05,
+      "loss": 1.9159,
+      "step": 532
+    },
+    {
+      "epoch": 0.14030939008769336,
+      "grad_norm": 1.1094871759414673,
+      "learning_rate": 9.53478184685474e-05,
+      "loss": 1.9393,
+      "step": 534
+    },
+    {
+      "epoch": 0.14083489342135513,
+      "grad_norm": 1.1761337518692017,
+      "learning_rate": 9.533029612756264e-05,
+      "loss": 1.9315,
+      "step": 536
+    },
+    {
+      "epoch": 0.1413603967550169,
+      "grad_norm": 0.8991339802742004,
+      "learning_rate": 9.531277378657789e-05,
+      "loss": 1.9138,
+      "step": 538
+    },
+    {
+      "epoch": 0.14188590008867868,
+      "grad_norm": 1.0923110246658325,
+      "learning_rate": 9.529525144559314e-05,
+      "loss": 1.9254,
+      "step": 540
+    },
+    {
+      "epoch": 0.14241140342234046,
+      "grad_norm": 0.8643242716789246,
+      "learning_rate": 9.527772910460838e-05,
+      "loss": 1.912,
+      "step": 542
+    },
+    {
+      "epoch": 0.14293690675600224,
+      "grad_norm": 0.8578360080718994,
+      "learning_rate": 9.526020676362363e-05,
+      "loss": 1.9403,
+      "step": 544
+    },
+    {
+      "epoch": 0.143462410089664,
+      "grad_norm": 0.854049563407898,
+      "learning_rate": 9.524268442263888e-05,
+      "loss": 1.9251,
+      "step": 546
+    },
+    {
+      "epoch": 0.1439879134233258,
+      "grad_norm": 1.1244609355926514,
+      "learning_rate": 9.522516208165411e-05,
+      "loss": 1.928,
+      "step": 548
+    },
+    {
+      "epoch": 0.14451341675698756,
+      "grad_norm": 1.0357218980789185,
+      "learning_rate": 9.520763974066936e-05,
+      "loss": 1.9137,
+      "step": 550
+    },
+    {
+      "epoch": 0.14503892009064934,
+      "grad_norm": 1.2136287689208984,
+      "learning_rate": 9.51901173996846e-05,
+      "loss": 1.9009,
+      "step": 552
+    },
+    {
+      "epoch": 0.14556442342431108,
+      "grad_norm": 1.5104690790176392,
+      "learning_rate": 9.517259505869984e-05,
+      "loss": 1.8873,
+      "step": 554
+    },
+    {
+      "epoch": 0.14608992675797286,
+      "grad_norm": 0.9920956492424011,
+      "learning_rate": 9.515507271771509e-05,
+      "loss": 1.9168,
+      "step": 556
+    },
+    {
+      "epoch": 0.14661543009163464,
+      "grad_norm": 1.0754443407058716,
+      "learning_rate": 9.513755037673034e-05,
+      "loss": 1.9258,
+      "step": 558
+    },
+    {
+      "epoch": 0.1471409334252964,
+      "grad_norm": 0.9329949617385864,
+      "learning_rate": 9.512002803574558e-05,
+      "loss": 1.8999,
+      "step": 560
+    },
+    {
+      "epoch": 0.14766643675895819,
+      "grad_norm": 0.9859107136726379,
+      "learning_rate": 9.510250569476082e-05,
+      "loss": 1.8973,
+      "step": 562
+    },
+    {
+      "epoch": 0.14819194009261996,
+      "grad_norm": 0.8728905916213989,
+      "learning_rate": 9.508498335377607e-05,
+      "loss": 1.8981,
+      "step": 564
+    },
+    {
+      "epoch": 0.14871744342628174,
+      "grad_norm": 0.9786973595619202,
+      "learning_rate": 9.506746101279131e-05,
+      "loss": 1.9167,
+      "step": 566
+    },
+    {
+      "epoch": 0.1492429467599435,
+      "grad_norm": 0.9139885902404785,
+      "learning_rate": 9.504993867180656e-05,
+      "loss": 1.8988,
+      "step": 568
+    },
+    {
+      "epoch": 0.1497684500936053,
+      "grad_norm": 0.9002274870872498,
+      "learning_rate": 9.503241633082181e-05,
+      "loss": 1.8964,
+      "step": 570
+    },
+    {
+      "epoch": 0.15029395342726706,
+      "grad_norm": 0.9470378756523132,
+      "learning_rate": 9.501489398983706e-05,
+      "loss": 1.8877,
+      "step": 572
+    },
+    {
+      "epoch": 0.15081945676092884,
+      "grad_norm": 1.2751479148864746,
+      "learning_rate": 9.499737164885229e-05,
+      "loss": 1.9328,
+      "step": 574
+    },
+    {
+      "epoch": 0.1513449600945906,
+      "grad_norm": 1.0107779502868652,
+      "learning_rate": 9.497984930786754e-05,
+      "loss": 1.9271,
+      "step": 576
+    },
+    {
+      "epoch": 0.15187046342825236,
+      "grad_norm": 1.3902311325073242,
+      "learning_rate": 9.496232696688277e-05,
+      "loss": 1.8903,
+      "step": 578
+    },
+    {
+      "epoch": 0.15239596676191414,
+      "grad_norm": 0.8747096657752991,
+      "learning_rate": 9.494480462589802e-05,
+      "loss": 1.9058,
+      "step": 580
+    },
+    {
+      "epoch": 0.1529214700955759,
+      "grad_norm": 0.9474460482597351,
+      "learning_rate": 9.492728228491327e-05,
+      "loss": 1.9357,
+      "step": 582
+    },
+    {
+      "epoch": 0.1534469734292377,
+      "grad_norm": 0.9786020517349243,
+      "learning_rate": 9.490975994392851e-05,
+      "loss": 1.9267,
+      "step": 584
+    },
+    {
+      "epoch": 0.15397247676289946,
+      "grad_norm": 0.9129196405410767,
+      "learning_rate": 9.489223760294376e-05,
+      "loss": 1.9129,
+      "step": 586
+    },
+    {
+      "epoch": 0.15449798009656124,
+      "grad_norm": 1.2065305709838867,
+      "learning_rate": 9.4874715261959e-05,
+      "loss": 1.8809,
+      "step": 588
+    },
+    {
+      "epoch": 0.155023483430223,
+      "grad_norm": 1.0768754482269287,
+      "learning_rate": 9.485719292097424e-05,
+      "loss": 1.8827,
+      "step": 590
+    },
+    {
+      "epoch": 0.1555489867638848,
+      "grad_norm": 1.3688833713531494,
+      "learning_rate": 9.483967057998949e-05,
+      "loss": 1.9325,
+      "step": 592
+    },
+    {
+      "epoch": 0.15607449009754656,
+      "grad_norm": 1.0570260286331177,
+      "learning_rate": 9.482214823900474e-05,
+      "loss": 1.8599,
+      "step": 594
+    },
+    {
+      "epoch": 0.15659999343120834,
+      "grad_norm": 1.0146839618682861,
+      "learning_rate": 9.480462589801999e-05,
+      "loss": 1.9205,
+      "step": 596
+    },
+    {
+      "epoch": 0.15712549676487012,
+      "grad_norm": 1.6670814752578735,
+      "learning_rate": 9.478710355703523e-05,
+      "loss": 1.8699,
+      "step": 598
+    },
+    {
+      "epoch": 0.15765100009853186,
+      "grad_norm": 1.178289532661438,
+      "learning_rate": 9.476958121605047e-05,
+      "loss": 1.9022,
+      "step": 600
+    },
+    {
+      "epoch": 0.15817650343219364,
+      "grad_norm": 1.6808775663375854,
+      "learning_rate": 9.47520588750657e-05,
+      "loss": 1.9264,
+      "step": 602
+    },
+    {
+      "epoch": 0.1587020067658554,
+      "grad_norm": 0.9852617383003235,
+      "learning_rate": 9.473453653408095e-05,
+      "loss": 1.9114,
+      "step": 604
+    },
+    {
+      "epoch": 0.1592275100995172,
+      "grad_norm": 0.8299278616905212,
+      "learning_rate": 9.47170141930962e-05,
+      "loss": 1.9012,
+      "step": 606
+    },
+    {
+      "epoch": 0.15975301343317896,
+      "grad_norm": 0.9899255037307739,
+      "learning_rate": 9.469949185211144e-05,
+      "loss": 1.9187,
+      "step": 608
+    },
+    {
+      "epoch": 0.16027851676684074,
+      "grad_norm": 0.9521119594573975,
+      "learning_rate": 9.468196951112669e-05,
+      "loss": 1.8972,
+      "step": 610
+    },
+    {
+      "epoch": 0.16080402010050251,
+      "grad_norm": 1.2446365356445312,
+      "learning_rate": 9.466444717014194e-05,
+      "loss": 1.8962,
+      "step": 612
+    },
+    {
+      "epoch": 0.1613295234341643,
+      "grad_norm": 1.1197859048843384,
+      "learning_rate": 9.464692482915717e-05,
+      "loss": 1.8753,
+      "step": 614
+    },
+    {
+      "epoch": 0.16185502676782607,
+      "grad_norm": 0.9934601783752441,
+      "learning_rate": 9.462940248817242e-05,
+      "loss": 1.8974,
+      "step": 616
+    },
+    {
+      "epoch": 0.16238053010148784,
+      "grad_norm": 0.9261951446533203,
+      "learning_rate": 9.461188014718767e-05,
+      "loss": 1.9052,
+      "step": 618
+    },
+    {
+      "epoch": 0.16290603343514962,
+      "grad_norm": 1.2140916585922241,
+      "learning_rate": 9.459435780620292e-05,
+      "loss": 1.8779,
+      "step": 620
+    },
+    {
+      "epoch": 0.16343153676881136,
+      "grad_norm": 1.067958116531372,
+      "learning_rate": 9.457683546521816e-05,
+      "loss": 1.8771,
+      "step": 622
+    },
+    {
+      "epoch": 0.16395704010247314,
+      "grad_norm": 0.9392744302749634,
+      "learning_rate": 9.455931312423341e-05,
+      "loss": 1.892,
+      "step": 624
+    },
+    {
+      "epoch": 0.16448254343613491,
+      "grad_norm": 1.6888792514801025,
+      "learning_rate": 9.454179078324864e-05,
+      "loss": 1.8829,
+      "step": 626
+    },
+    {
+      "epoch": 0.1650080467697967,
+      "grad_norm": 0.9804869890213013,
+      "learning_rate": 9.452426844226389e-05,
+      "loss": 1.9113,
+      "step": 628
+    },
+    {
+      "epoch": 0.16553355010345847,
+      "grad_norm": 1.0176154375076294,
+      "learning_rate": 9.450674610127913e-05,
+      "loss": 1.9063,
+      "step": 630
+    },
+    {
+      "epoch": 0.16605905343712024,
+      "grad_norm": 0.9681494235992432,
+      "learning_rate": 9.448922376029437e-05,
+      "loss": 1.9095,
+      "step": 632
+    },
+    {
+      "epoch": 0.16658455677078202,
+      "grad_norm": 0.9195823669433594,
+      "learning_rate": 9.447170141930962e-05,
+      "loss": 1.8881,
+      "step": 634
+    },
+    {
+      "epoch": 0.1671100601044438,
+      "grad_norm": 0.9165635108947754,
+      "learning_rate": 9.445417907832487e-05,
+      "loss": 1.8788,
+      "step": 636
+    },
+    {
+      "epoch": 0.16763556343810557,
+      "grad_norm": 1.2198373079299927,
+      "learning_rate": 9.443665673734012e-05,
+      "loss": 1.8717,
+      "step": 638
+    },
+    {
+      "epoch": 0.16816106677176734,
+      "grad_norm": 0.8923284411430359,
+      "learning_rate": 9.441913439635536e-05,
+      "loss": 1.918,
+      "step": 640
+    },
+    {
+      "epoch": 0.16868657010542912,
+      "grad_norm": 1.1938835382461548,
+      "learning_rate": 9.44016120553706e-05,
+      "loss": 1.9118,
+      "step": 642
+    },
+    {
+      "epoch": 0.16921207343909087,
+      "grad_norm": 0.8569881319999695,
+      "learning_rate": 9.438408971438585e-05,
+      "loss": 1.8894,
+      "step": 644
+    },
+    {
+      "epoch": 0.16973757677275264,
+      "grad_norm": 0.79718017578125,
+      "learning_rate": 9.43665673734011e-05,
+      "loss": 1.8815,
+      "step": 646
+    },
+    {
+      "epoch": 0.17026308010641442,
+      "grad_norm": 0.9728055000305176,
+      "learning_rate": 9.434904503241634e-05,
+      "loss": 1.8942,
+      "step": 648
+    },
+    {
+      "epoch": 0.1707885834400762,
+      "grad_norm": 1.2369781732559204,
+      "learning_rate": 9.433152269143159e-05,
+      "loss": 1.8663,
+      "step": 650
+    },
+    {
+      "epoch": 0.17131408677373797,
+      "grad_norm": 1.0035600662231445,
+      "learning_rate": 9.431400035044684e-05,
+      "loss": 1.9365,
+      "step": 652
+    },
+    {
+      "epoch": 0.17183959010739974,
+      "grad_norm": 0.8765824437141418,
+      "learning_rate": 9.429647800946207e-05,
+      "loss": 1.8818,
+      "step": 654
+    },
+    {
+      "epoch": 0.17236509344106152,
+      "grad_norm": 1.0349746942520142,
+      "learning_rate": 9.42789556684773e-05,
+      "loss": 1.8732,
+      "step": 656
+    },
+    {
+      "epoch": 0.1728905967747233,
+      "grad_norm": 0.9079018235206604,
+      "learning_rate": 9.426143332749255e-05,
+      "loss": 1.8841,
+      "step": 658
+    },
+    {
+      "epoch": 0.17341610010838507,
+      "grad_norm": 0.9596872925758362,
+      "learning_rate": 9.42439109865078e-05,
+      "loss": 1.8824,
+      "step": 660
+    },
+    {
+      "epoch": 0.17394160344204684,
+      "grad_norm": 0.8027825355529785,
+      "learning_rate": 9.422638864552305e-05,
+      "loss": 1.9056,
+      "step": 662
+    },
+    {
+      "epoch": 0.17446710677570862,
+      "grad_norm": 1.0182610750198364,
+      "learning_rate": 9.42088663045383e-05,
+      "loss": 1.8637,
+      "step": 664
+    },
+    {
+      "epoch": 0.1749926101093704,
+      "grad_norm": 1.0941540002822876,
+      "learning_rate": 9.419134396355354e-05,
+      "loss": 1.9068,
+      "step": 666
+    },
+    {
+      "epoch": 0.17551811344303214,
+      "grad_norm": 0.8376652598381042,
+      "learning_rate": 9.417382162256878e-05,
+      "loss": 1.88,
+      "step": 668
+    },
+    {
+      "epoch": 0.17604361677669392,
+      "grad_norm": 1.1947648525238037,
+      "learning_rate": 9.415629928158402e-05,
+      "loss": 1.8974,
+      "step": 670
+    },
+    {
+      "epoch": 0.1765691201103557,
+      "grad_norm": 1.1815409660339355,
+      "learning_rate": 9.413877694059927e-05,
+      "loss": 1.9003,
+      "step": 672
+    },
+    {
+      "epoch": 0.17709462344401747,
+      "grad_norm": 1.298322319984436,
+      "learning_rate": 9.412125459961452e-05,
+      "loss": 1.9069,
+      "step": 674
+    },
+    {
+      "epoch": 0.17762012677767924,
+      "grad_norm": 0.9089726209640503,
+      "learning_rate": 9.410373225862977e-05,
+      "loss": 1.8631,
+      "step": 676
+    },
+    {
+      "epoch": 0.17814563011134102,
+      "grad_norm": 2.489449977874756,
+      "learning_rate": 9.408620991764501e-05,
+      "loss": 1.8517,
+      "step": 678
+    },
+    {
+      "epoch": 0.1786711334450028,
+      "grad_norm": 1.456616759300232,
+      "learning_rate": 9.406868757666025e-05,
+      "loss": 1.9026,
+      "step": 680
+    },
+    {
+      "epoch": 0.17919663677866457,
+      "grad_norm": 1.3760594129562378,
+      "learning_rate": 9.405116523567548e-05,
+      "loss": 1.9106,
+      "step": 682
+    },
+    {
+      "epoch": 0.17972214011232635,
+      "grad_norm": 1.0166149139404297,
+      "learning_rate": 9.403364289469073e-05,
+      "loss": 1.8985,
+      "step": 684
+    },
+    {
+      "epoch": 0.18024764344598812,
+      "grad_norm": 0.9207403659820557,
+      "learning_rate": 9.401612055370598e-05,
+      "loss": 1.9231,
+      "step": 686
+    },
+    {
+      "epoch": 0.1807731467796499,
+      "grad_norm": 1.1127294301986694,
+      "learning_rate": 9.399859821272122e-05,
+      "loss": 1.9001,
+      "step": 688
+    },
+    {
+      "epoch": 0.18129865011331164,
+      "grad_norm": 1.1024377346038818,
+      "learning_rate": 9.398107587173647e-05,
+      "loss": 1.8673,
+      "step": 690
+    },
+    {
+      "epoch": 0.18182415344697342,
+      "grad_norm": 1.2524852752685547,
+      "learning_rate": 9.396355353075172e-05,
+      "loss": 1.8481,
+      "step": 692
+    },
+    {
+      "epoch": 0.1823496567806352,
+      "grad_norm": 1.0531530380249023,
+      "learning_rate": 9.394603118976695e-05,
+      "loss": 1.8818,
+      "step": 694
+    },
+    {
+      "epoch": 0.18287516011429697,
+      "grad_norm": 1.0276319980621338,
+      "learning_rate": 9.39285088487822e-05,
+      "loss": 1.8516,
+      "step": 696
+    },
+    {
+      "epoch": 0.18340066344795874,
+      "grad_norm": 0.9953783750534058,
+      "learning_rate": 9.391098650779745e-05,
+      "loss": 1.8906,
+      "step": 698
+    },
+    {
+      "epoch": 0.18392616678162052,
+      "grad_norm": 1.245374321937561,
+      "learning_rate": 9.38934641668127e-05,
+      "loss": 1.8797,
+      "step": 700
+    },
+    {
+      "epoch": 0.1844516701152823,
+      "grad_norm": 0.7994580864906311,
+      "learning_rate": 9.387594182582794e-05,
+      "loss": 1.8902,
+      "step": 702
+    },
+    {
+      "epoch": 0.18497717344894407,
+      "grad_norm": 0.941091775894165,
+      "learning_rate": 9.385841948484318e-05,
+      "loss": 1.8964,
+      "step": 704
+    },
+    {
+      "epoch": 0.18550267678260585,
+      "grad_norm": 0.8709573149681091,
+      "learning_rate": 9.384089714385843e-05,
+      "loss": 1.894,
+      "step": 706
+    },
+    {
+      "epoch": 0.18602818011626762,
+      "grad_norm": 1.1883244514465332,
+      "learning_rate": 9.382337480287366e-05,
+      "loss": 1.9171,
+      "step": 708
+    },
+    {
+      "epoch": 0.1865536834499294,
+      "grad_norm": 2.1829898357391357,
+      "learning_rate": 9.380585246188891e-05,
+      "loss": 1.9089,
+      "step": 710
+    },
+    {
+      "epoch": 0.18707918678359114,
+      "grad_norm": 0.9389178156852722,
+      "learning_rate": 9.378833012090415e-05,
+      "loss": 1.9028,
+      "step": 712
+    },
+    {
+      "epoch": 0.18760469011725292,
+      "grad_norm": 0.8789083957672119,
+      "learning_rate": 9.37708077799194e-05,
+      "loss": 1.8881,
+      "step": 714
+    },
+    {
+      "epoch": 0.1881301934509147,
+      "grad_norm": 1.3515336513519287,
+      "learning_rate": 9.375328543893465e-05,
+      "loss": 1.897,
+      "step": 716
+    },
+    {
+      "epoch": 0.18865569678457647,
+      "grad_norm": 1.123223066329956,
+      "learning_rate": 9.37357630979499e-05,
+      "loss": 1.9337,
+      "step": 718
+    },
+    {
+      "epoch": 0.18918120011823825,
+      "grad_norm": 1.16353440284729,
+      "learning_rate": 9.371824075696513e-05,
+      "loss": 1.9257,
+      "step": 720
+    },
+    {
+      "epoch": 0.18970670345190002,
+      "grad_norm": 0.8678929209709167,
+      "learning_rate": 9.370071841598038e-05,
+      "loss": 1.8686,
+      "step": 722
+    },
+    {
+      "epoch": 0.1902322067855618,
+      "grad_norm": 0.8892553448677063,
+      "learning_rate": 9.368319607499563e-05,
+      "loss": 1.8724,
+      "step": 724
+    },
+    {
+      "epoch": 0.19075771011922357,
+      "grad_norm": 1.2203401327133179,
+      "learning_rate": 9.366567373401087e-05,
+      "loss": 1.8663,
+      "step": 726
+    },
+    {
+      "epoch": 0.19128321345288535,
+      "grad_norm": 0.8035688400268555,
+      "learning_rate": 9.364815139302612e-05,
+      "loss": 1.8748,
+      "step": 728
+    },
+    {
+      "epoch": 0.19180871678654712,
+      "grad_norm": 0.8692450523376465,
+      "learning_rate": 9.363062905204136e-05,
+      "loss": 1.9093,
+      "step": 730
+    },
+    {
+      "epoch": 0.1923342201202089,
+      "grad_norm": 1.1297273635864258,
+      "learning_rate": 9.36131067110566e-05,
+      "loss": 1.8618,
+      "step": 732
+    },
+    {
+      "epoch": 0.19285972345387067,
+      "grad_norm": 0.853670597076416,
+      "learning_rate": 9.359558437007184e-05,
+      "loss": 1.8859,
+      "step": 734
+    },
+    {
+      "epoch": 0.19338522678753242,
+      "grad_norm": 1.115268588066101,
+      "learning_rate": 9.357806202908708e-05,
+      "loss": 1.8823,
+      "step": 736
+    },
+    {
+      "epoch": 0.1939107301211942,
+      "grad_norm": 1.002034068107605,
+      "learning_rate": 9.356053968810233e-05,
+      "loss": 1.8743,
+      "step": 738
+    },
+    {
+      "epoch": 0.19443623345485597,
+      "grad_norm": 1.3825932741165161,
+      "learning_rate": 9.354301734711758e-05,
+      "loss": 1.8897,
+      "step": 740
+    },
+    {
+      "epoch": 0.19496173678851775,
+      "grad_norm": 1.4974256753921509,
+      "learning_rate": 9.352549500613283e-05,
+      "loss": 1.899,
+      "step": 742
+    },
+    {
+      "epoch": 0.19548724012217952,
+      "grad_norm": 1.2865314483642578,
+      "learning_rate": 9.350797266514808e-05,
+      "loss": 1.9042,
+      "step": 744
+    },
+    {
+      "epoch": 0.1960127434558413,
+      "grad_norm": 1.036019206047058,
+      "learning_rate": 9.349045032416331e-05,
+      "loss": 1.9169,
+      "step": 746
+    },
+    {
+      "epoch": 0.19653824678950307,
+      "grad_norm": 1.1401413679122925,
+      "learning_rate": 9.347292798317856e-05,
+      "loss": 1.8908,
+      "step": 748
+    },
+    {
+      "epoch": 0.19706375012316485,
+      "grad_norm": 1.3356719017028809,
+      "learning_rate": 9.34554056421938e-05,
+      "loss": 1.8642,
+      "step": 750
+    },
+    {
+      "epoch": 0.19758925345682662,
+      "grad_norm": 1.025230884552002,
+      "learning_rate": 9.343788330120905e-05,
+      "loss": 1.8733,
+      "step": 752
+    },
+    {
+      "epoch": 0.1981147567904884,
+      "grad_norm": 0.981281578540802,
+      "learning_rate": 9.34203609602243e-05,
+      "loss": 1.8894,
+      "step": 754
+    },
+    {
+      "epoch": 0.19864026012415018,
+      "grad_norm": 1.0571757555007935,
+      "learning_rate": 9.340283861923953e-05,
+      "loss": 1.8753,
+      "step": 756
+    },
+    {
+      "epoch": 0.19916576345781192,
+      "grad_norm": 0.9619776606559753,
+      "learning_rate": 9.338531627825478e-05,
+      "loss": 1.8791,
+      "step": 758
+    },
+    {
+      "epoch": 0.1996912667914737,
+      "grad_norm": 0.8430715203285217,
+      "learning_rate": 9.336779393727001e-05,
+      "loss": 1.9133,
+      "step": 760
+    },
+    {
+      "epoch": 0.20021677012513547,
+      "grad_norm": 1.8572043180465698,
+      "learning_rate": 9.335027159628526e-05,
+      "loss": 1.9123,
+      "step": 762
+    },
+    {
+      "epoch": 0.20074227345879725,
+      "grad_norm": 1.0308524370193481,
+      "learning_rate": 9.333274925530051e-05,
+      "loss": 1.8696,
+      "step": 764
+    },
+    {
+      "epoch": 0.20126777679245902,
+      "grad_norm": 1.3068006038665771,
+      "learning_rate": 9.331522691431576e-05,
+      "loss": 1.8932,
+      "step": 766
+    },
+    {
+      "epoch": 0.2017932801261208,
+      "grad_norm": 0.9089499711990356,
+      "learning_rate": 9.3297704573331e-05,
+      "loss": 1.8717,
+      "step": 768
+    },
+    {
+      "epoch": 0.20231878345978258,
+      "grad_norm": 1.1663358211517334,
+      "learning_rate": 9.328018223234625e-05,
+      "loss": 1.8356,
+      "step": 770
+    },
+    {
+      "epoch": 0.20284428679344435,
+      "grad_norm": 1.1666746139526367,
+      "learning_rate": 9.326265989136149e-05,
+      "loss": 1.8785,
+      "step": 772
+    },
+    {
+      "epoch": 0.20336979012710613,
+      "grad_norm": 0.8820154666900635,
+      "learning_rate": 9.324513755037673e-05,
+      "loss": 1.8353,
+      "step": 774
+    },
+    {
+      "epoch": 0.2038952934607679,
+      "grad_norm": 1.0544767379760742,
+      "learning_rate": 9.322761520939198e-05,
+      "loss": 1.9217,
+      "step": 776
+    },
+    {
+      "epoch": 0.20442079679442968,
+      "grad_norm": 1.0300171375274658,
+      "learning_rate": 9.321009286840723e-05,
+      "loss": 1.8982,
+      "step": 778
+    },
+    {
+      "epoch": 0.20494630012809142,
+      "grad_norm": 0.884556770324707,
+      "learning_rate": 9.319257052742248e-05,
+      "loss": 1.8783,
+      "step": 780
+    },
+    {
+      "epoch": 0.2054718034617532,
+      "grad_norm": 1.150823712348938,
+      "learning_rate": 9.317504818643771e-05,
+      "loss": 1.9116,
+      "step": 782
+    },
+    {
+      "epoch": 0.20599730679541498,
+      "grad_norm": 1.1606664657592773,
+      "learning_rate": 9.315752584545296e-05,
+      "loss": 1.8594,
+      "step": 784
+    },
+    {
+      "epoch": 0.20652281012907675,
+      "grad_norm": 0.9920266270637512,
+      "learning_rate": 9.314000350446819e-05,
+      "loss": 1.8659,
+      "step": 786
+    },
+    {
+      "epoch": 0.20704831346273853,
+      "grad_norm": 1.0931354761123657,
+      "learning_rate": 9.312248116348344e-05,
+      "loss": 1.8876,
+      "step": 788
+    },
+    {
+      "epoch": 0.2075738167964003,
+      "grad_norm": 1.097580075263977,
+      "learning_rate": 9.310495882249869e-05,
+      "loss": 1.8961,
+      "step": 790
+    },
+    {
+      "epoch": 0.20809932013006208,
+      "grad_norm": 1.160636067390442,
+      "learning_rate": 9.308743648151394e-05,
+      "loss": 1.8925,
+      "step": 792
+    },
+    {
+      "epoch": 0.20862482346372385,
+      "grad_norm": 0.9510796666145325,
+      "learning_rate": 9.306991414052918e-05,
+      "loss": 1.8449,
+      "step": 794
+    },
+    {
+      "epoch": 0.20915032679738563,
+      "grad_norm": 1.0860143899917603,
+      "learning_rate": 9.305239179954443e-05,
+      "loss": 1.8647,
+      "step": 796
+    },
+    {
+      "epoch": 0.2096758301310474,
+      "grad_norm": 1.1009169816970825,
+      "learning_rate": 9.303486945855966e-05,
+      "loss": 1.8992,
+      "step": 798
+    },
+    {
+      "epoch": 0.21020133346470918,
+      "grad_norm": 1.0895287990570068,
+      "learning_rate": 9.301734711757491e-05,
+      "loss": 1.8563,
+      "step": 800
+    },
+    {
+      "epoch": 0.21020133346470918,
+      "eval_loss": 1.8377995491027832,
+      "eval_runtime": 487.1876,
+      "eval_samples_per_second": 249.984,
+      "eval_steps_per_second": 31.249,
+      "step": 800
+    },
+    {
+      "epoch": 0.21072683679837095,
+      "grad_norm": 1.2973535060882568,
+      "learning_rate": 9.299982477659016e-05,
+      "loss": 1.8701,
+      "step": 802
+    },
+    {
+      "epoch": 0.2112523401320327,
+      "grad_norm": 0.9440937042236328,
+      "learning_rate": 9.298230243560541e-05,
+      "loss": 1.8858,
+      "step": 804
+    },
+    {
+      "epoch": 0.21177784346569448,
+      "grad_norm": 0.8491653203964233,
+      "learning_rate": 9.296478009462064e-05,
+      "loss": 1.872,
+      "step": 806
+    },
+    {
+      "epoch": 0.21230334679935625,
+      "grad_norm": 1.0194580554962158,
+      "learning_rate": 9.294725775363589e-05,
+      "loss": 1.8803,
+      "step": 808
+    },
+    {
+      "epoch": 0.21282885013301803,
+      "grad_norm": 0.957872748374939,
+      "learning_rate": 9.292973541265114e-05,
+      "loss": 1.8825,
+      "step": 810
+    },
+    {
+      "epoch": 0.2133543534666798,
+      "grad_norm": 1.1070436239242554,
+      "learning_rate": 9.291221307166637e-05,
+      "loss": 1.8616,
+      "step": 812
+    },
+    {
+      "epoch": 0.21387985680034158,
+      "grad_norm": 0.8715662956237793,
+      "learning_rate": 9.289469073068162e-05,
+      "loss": 1.8634,
+      "step": 814
+    },
+    {
+      "epoch": 0.21440536013400335,
+      "grad_norm": 1.1930649280548096,
+      "learning_rate": 9.287716838969687e-05,
+      "loss": 1.842,
+      "step": 816
+    },
+    {
+      "epoch": 0.21493086346766513,
+      "grad_norm": 1.0091701745986938,
+      "learning_rate": 9.285964604871211e-05,
+      "loss": 1.9128,
+      "step": 818
+    },
+    {
+      "epoch": 0.2154563668013269,
+      "grad_norm": 1.1418068408966064,
+      "learning_rate": 9.284212370772736e-05,
+      "loss": 1.8697,
+      "step": 820
+    },
+    {
+      "epoch": 0.21598187013498868,
+      "grad_norm": 1.0154426097869873,
+      "learning_rate": 9.282460136674261e-05,
+      "loss": 1.9324,
+      "step": 822
+    },
+    {
+      "epoch": 0.21650737346865045,
+      "grad_norm": 1.2113468647003174,
+      "learning_rate": 9.280707902575784e-05,
+      "loss": 1.8379,
+      "step": 824
+    },
+    {
+      "epoch": 0.2170328768023122,
+      "grad_norm": 1.0505681037902832,
+      "learning_rate": 9.278955668477309e-05,
+      "loss": 1.866,
+      "step": 826
+    },
+    {
+      "epoch": 0.21755838013597398,
+      "grad_norm": 0.938463568687439,
+      "learning_rate": 9.277203434378834e-05,
+      "loss": 1.8504,
+      "step": 828
+    },
+    {
+      "epoch": 0.21808388346963575,
+      "grad_norm": 0.7944304943084717,
+      "learning_rate": 9.275451200280358e-05,
+      "loss": 1.8919,
+      "step": 830
+    },
+    {
+      "epoch": 0.21860938680329753,
+      "grad_norm": 0.9631441235542297,
+      "learning_rate": 9.273698966181882e-05,
+      "loss": 1.8451,
+      "step": 832
+    },
+    {
+      "epoch": 0.2191348901369593,
+      "grad_norm": 0.9219480156898499,
+      "learning_rate": 9.271946732083407e-05,
+      "loss": 1.8546,
+      "step": 834
+    },
+    {
+      "epoch": 0.21966039347062108,
+      "grad_norm": 0.8851411938667297,
+      "learning_rate": 9.270194497984931e-05,
+      "loss": 1.8685,
+      "step": 836
+    },
+    {
+      "epoch": 0.22018589680428285,
+      "grad_norm": 0.9707255959510803,
+      "learning_rate": 9.268442263886455e-05,
+      "loss": 1.887,
+      "step": 838
+    },
+    {
+      "epoch": 0.22071140013794463,
+      "grad_norm": 0.9294309616088867,
+      "learning_rate": 9.26669002978798e-05,
+      "loss": 1.8967,
+      "step": 840
+    },
+    {
+      "epoch": 0.2212369034716064,
+      "grad_norm": 0.9156199097633362,
+      "learning_rate": 9.264937795689504e-05,
+      "loss": 1.8756,
+      "step": 842
+    },
+    {
+      "epoch": 0.22176240680526818,
+      "grad_norm": 0.8118696808815002,
+      "learning_rate": 9.263185561591029e-05,
+      "loss": 1.8766,
+      "step": 844
+    },
+    {
+      "epoch": 0.22228791013892996,
+      "grad_norm": 0.9590555429458618,
+      "learning_rate": 9.261433327492554e-05,
+      "loss": 1.8693,
+      "step": 846
+    },
+    {
+      "epoch": 0.2228134134725917,
+      "grad_norm": 1.385361671447754,
+      "learning_rate": 9.259681093394079e-05,
+      "loss": 1.8824,
+      "step": 848
+    },
+    {
+      "epoch": 0.22333891680625348,
+      "grad_norm": 0.9501360654830933,
+      "learning_rate": 9.257928859295602e-05,
+      "loss": 1.8414,
+      "step": 850
+    },
+    {
+      "epoch": 0.22386442013991525,
+      "grad_norm": 1.0095267295837402,
+      "learning_rate": 9.256176625197127e-05,
+      "loss": 1.8499,
+      "step": 852
+    },
+    {
+      "epoch": 0.22438992347357703,
+      "grad_norm": 0.8820069432258606,
+      "learning_rate": 9.254424391098651e-05,
+      "loss": 1.8513,
+      "step": 854
+    },
+    {
+      "epoch": 0.2249154268072388,
+      "grad_norm": 0.7832709550857544,
+      "learning_rate": 9.252672157000176e-05,
+      "loss": 1.8585,
+      "step": 856
+    },
+    {
+      "epoch": 0.22544093014090058,
+      "grad_norm": 1.4282846450805664,
+      "learning_rate": 9.2509199229017e-05,
+      "loss": 1.8776,
+      "step": 858
+    },
+    {
+      "epoch": 0.22596643347456236,
+      "grad_norm": 0.8925049901008606,
+      "learning_rate": 9.249167688803224e-05,
+      "loss": 1.8626,
+      "step": 860
+    },
+    {
+      "epoch": 0.22649193680822413,
+      "grad_norm": 1.4945679903030396,
+      "learning_rate": 9.247415454704749e-05,
+      "loss": 1.874,
+      "step": 862
+    },
+    {
+      "epoch": 0.2270174401418859,
+      "grad_norm": 0.9479649066925049,
+      "learning_rate": 9.245663220606273e-05,
+      "loss": 1.8789,
+      "step": 864
+    },
+    {
+      "epoch": 0.22754294347554768,
+      "grad_norm": 1.023941159248352,
+      "learning_rate": 9.243910986507797e-05,
+      "loss": 1.8945,
+      "step": 866
+    },
+    {
+      "epoch": 0.22806844680920946,
+      "grad_norm": 1.0005570650100708,
+      "learning_rate": 9.242158752409322e-05,
+      "loss": 1.8736,
+      "step": 868
+    },
+    {
+      "epoch": 0.2285939501428712,
+      "grad_norm": 0.9343464374542236,
+      "learning_rate": 9.240406518310847e-05,
+      "loss": 1.844,
+      "step": 870
+    },
+    {
+      "epoch": 0.22911945347653298,
+      "grad_norm": 0.8942081332206726,
+      "learning_rate": 9.238654284212372e-05,
+      "loss": 1.8737,
+      "step": 872
+    },
+    {
+      "epoch": 0.22964495681019476,
+      "grad_norm": 1.0856554508209229,
+      "learning_rate": 9.236902050113896e-05,
+      "loss": 1.8492,
+      "step": 874
+    },
+    {
+      "epoch": 0.23017046014385653,
+      "grad_norm": 0.8268112540245056,
+      "learning_rate": 9.23514981601542e-05,
+      "loss": 1.8804,
+      "step": 876
+    },
+    {
+      "epoch": 0.2306959634775183,
+      "grad_norm": 0.9946966171264648,
+      "learning_rate": 9.233397581916944e-05,
+      "loss": 1.8592,
+      "step": 878
+    },
+    {
+      "epoch": 0.23122146681118008,
+      "grad_norm": 1.0663763284683228,
+      "learning_rate": 9.231645347818469e-05,
+      "loss": 1.8991,
+      "step": 880
+    },
+    {
+      "epoch": 0.23174697014484186,
+      "grad_norm": 1.2675755023956299,
+      "learning_rate": 9.229893113719993e-05,
+      "loss": 1.8621,
+      "step": 882
+    },
+    {
+      "epoch": 0.23227247347850363,
+      "grad_norm": 1.1921007633209229,
+      "learning_rate": 9.228140879621517e-05,
+      "loss": 1.9014,
+      "step": 884
+    },
+    {
+      "epoch": 0.2327979768121654,
+      "grad_norm": 0.8989017009735107,
+      "learning_rate": 9.226388645523042e-05,
+      "loss": 1.8575,
+      "step": 886
+    },
+    {
+      "epoch": 0.23332348014582718,
+      "grad_norm": 1.2373161315917969,
+      "learning_rate": 9.224636411424567e-05,
+      "loss": 1.8905,
+      "step": 888
+    },
+    {
+      "epoch": 0.23384898347948896,
+      "grad_norm": 0.938845694065094,
+      "learning_rate": 9.222884177326092e-05,
+      "loss": 1.9011,
+      "step": 890
+    },
+    {
+      "epoch": 0.23437448681315073,
+      "grad_norm": 0.8705965876579285,
+      "learning_rate": 9.221131943227615e-05,
+      "loss": 1.831,
+      "step": 892
+    },
+    {
+      "epoch": 0.23489999014681248,
+      "grad_norm": 1.3164907693862915,
+      "learning_rate": 9.21937970912914e-05,
+      "loss": 1.8455,
+      "step": 894
+    },
+    {
+      "epoch": 0.23542549348047426,
+      "grad_norm": 0.9536553025245667,
+      "learning_rate": 9.217627475030665e-05,
+      "loss": 1.8557,
+      "step": 896
+    },
+    {
+      "epoch": 0.23595099681413603,
+      "grad_norm": 1.0079597234725952,
+      "learning_rate": 9.21587524093219e-05,
+      "loss": 1.8758,
+      "step": 898
+    },
+    {
+      "epoch": 0.2364765001477978,
+      "grad_norm": 2.14665150642395,
+      "learning_rate": 9.214123006833714e-05,
+      "loss": 1.9172,
+      "step": 900
+    },
+    {
+      "epoch": 0.23700200348145958,
+      "grad_norm": 1.007122278213501,
+      "learning_rate": 9.212370772735239e-05,
+      "loss": 1.9175,
+      "step": 902
+    },
+    {
+      "epoch": 0.23752750681512136,
+      "grad_norm": 0.9156002998352051,
+      "learning_rate": 9.210618538636762e-05,
+      "loss": 1.8589,
+      "step": 904
+    },
+    {
+      "epoch": 0.23805301014878313,
+      "grad_norm": 1.2697999477386475,
+      "learning_rate": 9.208866304538287e-05,
+      "loss": 1.8996,
+      "step": 906
+    },
+    {
+      "epoch": 0.2385785134824449,
+      "grad_norm": 0.8666015863418579,
+      "learning_rate": 9.20711407043981e-05,
+      "loss": 1.8595,
+      "step": 908
+    },
+    {
+      "epoch": 0.23910401681610668,
+      "grad_norm": 1.5307285785675049,
+      "learning_rate": 9.205361836341335e-05,
+      "loss": 1.8551,
+      "step": 910
+    },
+    {
+      "epoch": 0.23962952014976846,
+      "grad_norm": 1.4660929441452026,
+      "learning_rate": 9.20360960224286e-05,
+      "loss": 1.8653,
+      "step": 912
+    },
+    {
+      "epoch": 0.24015502348343024,
+      "grad_norm": 0.9962916970252991,
+      "learning_rate": 9.201857368144385e-05,
+      "loss": 1.8713,
+      "step": 914
+    },
+    {
+      "epoch": 0.24068052681709198,
+      "grad_norm": 1.3509130477905273,
+      "learning_rate": 9.20010513404591e-05,
+      "loss": 1.8946,
+      "step": 916
+    },
+    {
+      "epoch": 0.24120603015075376,
+      "grad_norm": 0.8232421875,
+      "learning_rate": 9.198352899947433e-05,
+      "loss": 1.8795,
+      "step": 918
+    },
+    {
+      "epoch": 0.24173153348441553,
+      "grad_norm": 1.6278520822525024,
+      "learning_rate": 9.196600665848958e-05,
+      "loss": 1.901,
+      "step": 920
+    },
+    {
+      "epoch": 0.2422570368180773,
+      "grad_norm": 0.7896947860717773,
+      "learning_rate": 9.194848431750482e-05,
+      "loss": 1.8681,
+      "step": 922
+    },
+    {
+      "epoch": 0.24278254015173908,
+      "grad_norm": 1.1732158660888672,
+      "learning_rate": 9.193096197652007e-05,
+      "loss": 1.8632,
+      "step": 924
+    },
+    {
+      "epoch": 0.24330804348540086,
+      "grad_norm": 1.7037454843521118,
+      "learning_rate": 9.191343963553532e-05,
+      "loss": 1.8928,
+      "step": 926
+    },
+    {
+      "epoch": 0.24383354681906264,
+      "grad_norm": 0.9831770062446594,
+      "learning_rate": 9.189591729455057e-05,
+      "loss": 1.8491,
+      "step": 928
+    },
+    {
+      "epoch": 0.2443590501527244,
+      "grad_norm": 1.2588465213775635,
+      "learning_rate": 9.18783949535658e-05,
+      "loss": 1.8743,
+      "step": 930
+    },
+    {
+      "epoch": 0.2448845534863862,
+      "grad_norm": 0.8186757564544678,
+      "learning_rate": 9.186087261258105e-05,
+      "loss": 1.8961,
+      "step": 932
+    },
+    {
+      "epoch": 0.24541005682004796,
+      "grad_norm": 1.0041245222091675,
+      "learning_rate": 9.184335027159628e-05,
+      "loss": 1.8264,
+      "step": 934
+    },
+    {
+      "epoch": 0.24593556015370974,
+      "grad_norm": 1.397493600845337,
+      "learning_rate": 9.182582793061153e-05,
+      "loss": 1.8483,
+      "step": 936
+    },
+    {
+      "epoch": 0.24646106348737148,
+      "grad_norm": 0.9180475473403931,
+      "learning_rate": 9.180830558962678e-05,
+      "loss": 1.8609,
+      "step": 938
+    },
+    {
+      "epoch": 0.24698656682103326,
+      "grad_norm": 0.9013431072235107,
+      "learning_rate": 9.179078324864202e-05,
+      "loss": 1.8537,
+      "step": 940
+    },
+    {
+      "epoch": 0.24751207015469504,
+      "grad_norm": 0.7891268134117126,
+      "learning_rate": 9.177326090765727e-05,
+      "loss": 1.8224,
+      "step": 942
+    },
+    {
+      "epoch": 0.2480375734883568,
+      "grad_norm": 1.082982063293457,
+      "learning_rate": 9.17557385666725e-05,
+      "loss": 1.8799,
+      "step": 944
+    },
+    {
+      "epoch": 0.24856307682201859,
+      "grad_norm": 0.9956438541412354,
+      "learning_rate": 9.173821622568775e-05,
+      "loss": 1.8638,
+      "step": 946
+    },
+    {
+      "epoch": 0.24908858015568036,
+      "grad_norm": 1.1269092559814453,
+      "learning_rate": 9.1720693884703e-05,
+      "loss": 1.8589,
+      "step": 948
+    },
+    {
+      "epoch": 0.24961408348934214,
+      "grad_norm": 0.88730788230896,
+      "learning_rate": 9.170317154371825e-05,
+      "loss": 1.8901,
+      "step": 950
+    },
+    {
+      "epoch": 0.2501395868230039,
+      "grad_norm": 0.9314135313034058,
+      "learning_rate": 9.16856492027335e-05,
+      "loss": 1.8457,
+      "step": 952
+    },
+    {
+      "epoch": 0.25066509015666566,
+      "grad_norm": 1.0120025873184204,
+      "learning_rate": 9.166812686174874e-05,
+      "loss": 1.8666,
+      "step": 954
+    },
+    {
+      "epoch": 0.25119059349032746,
+      "grad_norm": 1.0400328636169434,
+      "learning_rate": 9.165060452076398e-05,
+      "loss": 1.8666,
+      "step": 956
+    },
+    {
+      "epoch": 0.2517160968239892,
+      "grad_norm": 1.084693431854248,
+      "learning_rate": 9.163308217977923e-05,
+      "loss": 1.8318,
+      "step": 958
+    },
+    {
+      "epoch": 0.252241600157651,
+      "grad_norm": 0.911669909954071,
+      "learning_rate": 9.161555983879446e-05,
+      "loss": 1.859,
+      "step": 960
+    },
+    {
+      "epoch": 0.25276710349131276,
+      "grad_norm": 1.0443209409713745,
+      "learning_rate": 9.159803749780971e-05,
+      "loss": 1.8647,
+      "step": 962
+    },
+    {
+      "epoch": 0.25329260682497456,
+      "grad_norm": 0.8681670427322388,
+      "learning_rate": 9.158051515682495e-05,
+      "loss": 1.8975,
+      "step": 964
+    },
+    {
+      "epoch": 0.2538181101586363,
+      "grad_norm": 1.2208961248397827,
+      "learning_rate": 9.15629928158402e-05,
+      "loss": 1.8722,
+      "step": 966
+    },
+    {
+      "epoch": 0.2543436134922981,
+      "grad_norm": 0.8562275767326355,
+      "learning_rate": 9.154547047485545e-05,
+      "loss": 1.8535,
+      "step": 968
+    },
+    {
+      "epoch": 0.25486911682595986,
+      "grad_norm": 0.8852279782295227,
+      "learning_rate": 9.152794813387068e-05,
+      "loss": 1.8713,
+      "step": 970
+    },
+    {
+      "epoch": 0.2553946201596216,
+      "grad_norm": 0.8528086543083191,
+      "learning_rate": 9.151042579288593e-05,
+      "loss": 1.8736,
+      "step": 972
+    },
+    {
+      "epoch": 0.2559201234932834,
+      "grad_norm": 0.886330246925354,
+      "learning_rate": 9.149290345190118e-05,
+      "loss": 1.8915,
+      "step": 974
+    },
+    {
+      "epoch": 0.25644562682694516,
+      "grad_norm": 0.8512532711029053,
+      "learning_rate": 9.147538111091643e-05,
+      "loss": 1.8293,
+      "step": 976
+    },
+    {
+      "epoch": 0.25697113016060696,
+      "grad_norm": 0.9382111430168152,
+      "learning_rate": 9.145785876993167e-05,
+      "loss": 1.8717,
+      "step": 978
+    },
+    {
+      "epoch": 0.2574966334942687,
+      "grad_norm": 0.8720589876174927,
+      "learning_rate": 9.144033642894692e-05,
+      "loss": 1.8839,
+      "step": 980
+    },
+    {
+      "epoch": 0.2580221368279305,
+      "grad_norm": 1.6592185497283936,
+      "learning_rate": 9.142281408796216e-05,
+      "loss": 1.8409,
+      "step": 982
+    },
+    {
+      "epoch": 0.25854764016159226,
+      "grad_norm": 1.2780932188034058,
+      "learning_rate": 9.140529174697739e-05,
+      "loss": 1.8717,
+      "step": 984
+    },
+    {
+      "epoch": 0.25907314349525407,
+      "grad_norm": 0.9220293164253235,
+      "learning_rate": 9.138776940599264e-05,
+      "loss": 1.8539,
+      "step": 986
+    },
+    {
+      "epoch": 0.2595986468289158,
+      "grad_norm": 0.8890568017959595,
+      "learning_rate": 9.137024706500788e-05,
+      "loss": 1.8678,
+      "step": 988
+    },
+    {
+      "epoch": 0.2601241501625776,
+      "grad_norm": 1.0034205913543701,
+      "learning_rate": 9.135272472402313e-05,
+      "loss": 1.8262,
+      "step": 990
+    },
+    {
+      "epoch": 0.26064965349623936,
+      "grad_norm": 1.0338081121444702,
+      "learning_rate": 9.133520238303838e-05,
+      "loss": 1.8566,
+      "step": 992
+    },
+    {
+      "epoch": 0.26117515682990117,
+      "grad_norm": 1.4746791124343872,
+      "learning_rate": 9.131768004205363e-05,
+      "loss": 1.8843,
+      "step": 994
+    },
+    {
+      "epoch": 0.2617006601635629,
+      "grad_norm": 1.0208336114883423,
+      "learning_rate": 9.130015770106886e-05,
+      "loss": 1.8503,
+      "step": 996
+    },
+    {
+      "epoch": 0.26222616349722466,
+      "grad_norm": 0.9133326411247253,
+      "learning_rate": 9.128263536008411e-05,
+      "loss": 1.8805,
+      "step": 998
+    },
+    {
+      "epoch": 0.26275166683088647,
+      "grad_norm": 1.1855682134628296,
+      "learning_rate": 9.126511301909936e-05,
+      "loss": 1.8862,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2632771701645482,
+      "grad_norm": 0.9511350393295288,
+      "learning_rate": 9.12475906781146e-05,
+      "loss": 1.816,
+      "step": 1002
+    },
+    {
+      "epoch": 0.26380267349821,
+      "grad_norm": 1.5805948972702026,
+      "learning_rate": 9.123006833712985e-05,
+      "loss": 1.8539,
+      "step": 1004
+    },
+    {
+      "epoch": 0.26432817683187176,
+      "grad_norm": 1.7137740850448608,
+      "learning_rate": 9.12125459961451e-05,
+      "loss": 1.8516,
+      "step": 1006
+    },
+    {
+      "epoch": 0.26485368016553357,
+      "grad_norm": 1.1085962057113647,
+      "learning_rate": 9.119502365516033e-05,
+      "loss": 1.8548,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2653791834991953,
+      "grad_norm": 0.927699625492096,
+      "learning_rate": 9.117750131417557e-05,
+      "loss": 1.8411,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2659046868328571,
+      "grad_norm": 1.0528203248977661,
+      "learning_rate": 9.115997897319081e-05,
+      "loss": 1.8672,
+      "step": 1012
+    },
+    {
+      "epoch": 0.26643019016651887,
+      "grad_norm": 0.8325463533401489,
+      "learning_rate": 9.114245663220606e-05,
+      "loss": 1.8081,
+      "step": 1014
+    },
+    {
+      "epoch": 0.26695569350018067,
+      "grad_norm": 0.9019527435302734,
+      "learning_rate": 9.112493429122131e-05,
+      "loss": 1.869,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2674811968338424,
+      "grad_norm": 1.3394633531570435,
+      "learning_rate": 9.110741195023656e-05,
+      "loss": 1.8063,
+      "step": 1018
+    },
+    {
+      "epoch": 0.26800670016750416,
+      "grad_norm": 1.0652636289596558,
+      "learning_rate": 9.10898896092518e-05,
+      "loss": 1.8685,
+      "step": 1020
+    },
+    {
+      "epoch": 0.26853220350116597,
+      "grad_norm": 1.0782673358917236,
+      "learning_rate": 9.107236726826704e-05,
+      "loss": 1.8749,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2690577068348277,
+      "grad_norm": 1.4112943410873413,
+      "learning_rate": 9.105484492728229e-05,
+      "loss": 1.8942,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2695832101684895,
+      "grad_norm": 0.9202266931533813,
+      "learning_rate": 9.103732258629753e-05,
+      "loss": 1.8578,
+      "step": 1026
+    },
+    {
+      "epoch": 0.27010871350215127,
+      "grad_norm": 1.4176150560379028,
+      "learning_rate": 9.101980024531278e-05,
+      "loss": 1.8788,
+      "step": 1028
+    },
+    {
+      "epoch": 0.27063421683581307,
+      "grad_norm": 1.0629339218139648,
+      "learning_rate": 9.100227790432803e-05,
+      "loss": 1.8597,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2711597201694748,
+      "grad_norm": 1.4514985084533691,
+      "learning_rate": 9.098475556334328e-05,
+      "loss": 1.8748,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2716852235031366,
+      "grad_norm": 1.7836532592773438,
+      "learning_rate": 9.096723322235851e-05,
+      "loss": 1.8205,
+      "step": 1034
+    },
+    {
+      "epoch": 0.27221072683679837,
+      "grad_norm": 0.9807853698730469,
+      "learning_rate": 9.094971088137374e-05,
+      "loss": 1.8556,
+      "step": 1036
+    },
+    {
+      "epoch": 0.27273623017046017,
+      "grad_norm": 0.9897574782371521,
+      "learning_rate": 9.093218854038899e-05,
+      "loss": 1.8298,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2732617335041219,
+      "grad_norm": 0.8160204887390137,
+      "learning_rate": 9.091466619940424e-05,
+      "loss": 1.8612,
+      "step": 1040
+    },
+    {
+      "epoch": 0.27378723683778367,
+      "grad_norm": 1.4802687168121338,
+      "learning_rate": 9.089714385841949e-05,
+      "loss": 1.8657,
+      "step": 1042
+    },
+    {
+      "epoch": 0.27431274017144547,
+      "grad_norm": 1.0503878593444824,
+      "learning_rate": 9.087962151743474e-05,
+      "loss": 1.8625,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2748382435051072,
+      "grad_norm": 1.03403639793396,
+      "learning_rate": 9.086209917644998e-05,
+      "loss": 1.8339,
+      "step": 1046
+    },
+    {
+      "epoch": 0.275363746838769,
+      "grad_norm": 1.1939598321914673,
+      "learning_rate": 9.084457683546522e-05,
+      "loss": 1.8322,
+      "step": 1048
+    },
+    {
+      "epoch": 0.27588925017243077,
+      "grad_norm": 0.8081462383270264,
+      "learning_rate": 9.082705449448046e-05,
+      "loss": 1.8527,
+      "step": 1050
+    },
+    {
+      "epoch": 0.27641475350609257,
+      "grad_norm": 0.9340723156929016,
+      "learning_rate": 9.080953215349571e-05,
+      "loss": 1.8462,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2769402568397543,
+      "grad_norm": 2.116253137588501,
+      "learning_rate": 9.079200981251096e-05,
+      "loss": 1.8157,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2774657601734161,
+      "grad_norm": 1.5195270776748657,
+      "learning_rate": 9.077448747152621e-05,
+      "loss": 1.8554,
+      "step": 1056
+    },
+    {
+      "epoch": 0.27799126350707787,
+      "grad_norm": 1.1265277862548828,
+      "learning_rate": 9.075696513054145e-05,
+      "loss": 1.8535,
+      "step": 1058
+    },
+    {
+      "epoch": 0.27851676684073967,
+      "grad_norm": 0.8470209240913391,
+      "learning_rate": 9.073944278955669e-05,
+      "loss": 1.8552,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2790422701744014,
+      "grad_norm": 1.014785885810852,
+      "learning_rate": 9.072192044857192e-05,
+      "loss": 1.8314,
+      "step": 1062
+    },
+    {
+      "epoch": 0.27956777350806317,
+      "grad_norm": 0.9315053820610046,
+      "learning_rate": 9.070439810758717e-05,
+      "loss": 1.852,
+      "step": 1064
+    },
+    {
+      "epoch": 0.28009327684172497,
+      "grad_norm": 0.8854875564575195,
+      "learning_rate": 9.068687576660242e-05,
+      "loss": 1.853,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2806187801753867,
+      "grad_norm": 1.0083775520324707,
+      "learning_rate": 9.066935342561767e-05,
+      "loss": 1.8525,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2811442835090485,
+      "grad_norm": 0.8299185633659363,
+      "learning_rate": 9.065183108463291e-05,
+      "loss": 1.858,
+      "step": 1070
+    },
+    {
+      "epoch": 0.28166978684271027,
+      "grad_norm": 0.859104573726654,
+      "learning_rate": 9.063430874364816e-05,
+      "loss": 1.8478,
+      "step": 1072
+    },
+    {
+      "epoch": 0.28219529017637207,
+      "grad_norm": 0.8011692762374878,
+      "learning_rate": 9.06167864026634e-05,
+      "loss": 1.8531,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2827207935100338,
+      "grad_norm": 0.8882426023483276,
+      "learning_rate": 9.059926406167864e-05,
+      "loss": 1.87,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2832462968436956,
+      "grad_norm": 0.9461469650268555,
+      "learning_rate": 9.058174172069389e-05,
+      "loss": 1.8531,
+      "step": 1078
+    },
+    {
+      "epoch": 0.28377180017735737,
+      "grad_norm": 0.92154461145401,
+      "learning_rate": 9.056421937970914e-05,
+      "loss": 1.8378,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2842973035110192,
+      "grad_norm": 0.9303539395332336,
+      "learning_rate": 9.054669703872438e-05,
+      "loss": 1.8897,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2848228068446809,
+      "grad_norm": 0.8764487504959106,
+      "learning_rate": 9.052917469773963e-05,
+      "loss": 1.8908,
+      "step": 1084
+    },
+    {
+      "epoch": 0.28534831017834267,
+      "grad_norm": 1.0205122232437134,
+      "learning_rate": 9.051165235675487e-05,
+      "loss": 1.8712,
+      "step": 1086
+    },
+    {
+      "epoch": 0.28587381351200447,
+      "grad_norm": 1.2372097969055176,
+      "learning_rate": 9.04941300157701e-05,
+      "loss": 1.8288,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2863993168456662,
+      "grad_norm": 0.9842033386230469,
+      "learning_rate": 9.047660767478535e-05,
+      "loss": 1.8767,
+      "step": 1090
+    },
+    {
+      "epoch": 0.286924820179328,
+      "grad_norm": 1.4316095113754272,
+      "learning_rate": 9.04590853338006e-05,
+      "loss": 1.8624,
+      "step": 1092
+    },
+    {
+      "epoch": 0.28745032351298977,
+      "grad_norm": 1.0971202850341797,
+      "learning_rate": 9.044156299281584e-05,
+      "loss": 1.858,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2879758268466516,
+      "grad_norm": 1.3766525983810425,
+      "learning_rate": 9.042404065183109e-05,
+      "loss": 1.8366,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2885013301803133,
+      "grad_norm": 1.5556044578552246,
+      "learning_rate": 9.040651831084634e-05,
+      "loss": 1.827,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2890268335139751,
+      "grad_norm": 0.803501307964325,
+      "learning_rate": 9.038899596986157e-05,
+      "loss": 1.9033,
+      "step": 1100
+    },
+    {
+      "epoch": 0.28955233684763687,
+      "grad_norm": 1.090751051902771,
+      "learning_rate": 9.037147362887682e-05,
+      "loss": 1.8605,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2900778401812987,
+      "grad_norm": 1.6796822547912598,
+      "learning_rate": 9.035395128789207e-05,
+      "loss": 1.8421,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2906033435149604,
+      "grad_norm": 0.8966239094734192,
+      "learning_rate": 9.033642894690731e-05,
+      "loss": 1.8679,
+      "step": 1106
+    },
+    {
+      "epoch": 0.29112884684862217,
+      "grad_norm": 1.390019416809082,
+      "learning_rate": 9.031890660592256e-05,
+      "loss": 1.8322,
+      "step": 1108
+    },
+    {
+      "epoch": 0.291654350182284,
+      "grad_norm": 0.8526667356491089,
+      "learning_rate": 9.030138426493781e-05,
+      "loss": 1.8364,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2921798535159457,
+      "grad_norm": 1.1773560047149658,
+      "learning_rate": 9.028386192395304e-05,
+      "loss": 1.8466,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2927053568496075,
+      "grad_norm": 1.0654343366622925,
+      "learning_rate": 9.026633958296828e-05,
+      "loss": 1.8767,
+      "step": 1114
+    },
+    {
+      "epoch": 0.29323086018326927,
+      "grad_norm": 1.023926854133606,
+      "learning_rate": 9.024881724198353e-05,
+      "loss": 1.8461,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2937563635169311,
+      "grad_norm": 0.94902503490448,
+      "learning_rate": 9.023129490099877e-05,
+      "loss": 1.8412,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2942818668505928,
+      "grad_norm": 1.0602984428405762,
+      "learning_rate": 9.021377256001402e-05,
+      "loss": 1.8704,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2948073701842546,
+      "grad_norm": 0.8396863341331482,
+      "learning_rate": 9.019625021902927e-05,
+      "loss": 1.8557,
+      "step": 1122
+    },
+    {
+      "epoch": 0.29533287351791637,
+      "grad_norm": 1.0940845012664795,
+      "learning_rate": 9.017872787804452e-05,
+      "loss": 1.8773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2958583768515782,
+      "grad_norm": 0.8471454977989197,
+      "learning_rate": 9.016120553705975e-05,
+      "loss": 1.8232,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2963838801852399,
+      "grad_norm": 0.7603086829185486,
+      "learning_rate": 9.0143683196075e-05,
+      "loss": 1.855,
+      "step": 1128
+    },
+    {
+      "epoch": 0.29690938351890167,
+      "grad_norm": 0.8293117880821228,
+      "learning_rate": 9.012616085509024e-05,
+      "loss": 1.8497,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2974348868525635,
+      "grad_norm": 0.8437036275863647,
+      "learning_rate": 9.010863851410549e-05,
+      "loss": 1.8476,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2979603901862252,
+      "grad_norm": 0.9667044878005981,
+      "learning_rate": 9.009111617312074e-05,
+      "loss": 1.7826,
+      "step": 1134
+    },
+    {
+      "epoch": 0.298485893519887,
+      "grad_norm": 0.7626157402992249,
+      "learning_rate": 9.007359383213599e-05,
+      "loss": 1.8588,
+      "step": 1136
+    },
+    {
+      "epoch": 0.29901139685354877,
+      "grad_norm": 0.782361626625061,
+      "learning_rate": 9.005607149115122e-05,
+      "loss": 1.8495,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2995369001872106,
+      "grad_norm": 0.9417736530303955,
+      "learning_rate": 9.003854915016647e-05,
+      "loss": 1.8875,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3000624035208723,
+      "grad_norm": 0.8616002202033997,
+      "learning_rate": 9.00210268091817e-05,
+      "loss": 1.852,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3005879068545341,
+      "grad_norm": 0.8036054372787476,
+      "learning_rate": 9.000350446819695e-05,
+      "loss": 1.8638,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3011134101881959,
+      "grad_norm": 1.0912983417510986,
+      "learning_rate": 8.99859821272122e-05,
+      "loss": 1.8191,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3016389135218577,
+      "grad_norm": 0.9259098172187805,
+      "learning_rate": 8.996845978622745e-05,
+      "loss": 1.8228,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3021644168555194,
+      "grad_norm": 1.3986306190490723,
+      "learning_rate": 8.99509374452427e-05,
+      "loss": 1.8365,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3026899201891812,
+      "grad_norm": 0.8834369778633118,
+      "learning_rate": 8.993341510425794e-05,
+      "loss": 1.857,
+      "step": 1152
+    },
+    {
+      "epoch": 0.303215423522843,
+      "grad_norm": 0.9686596989631653,
+      "learning_rate": 8.991589276327317e-05,
+      "loss": 1.8559,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3037409268565047,
+      "grad_norm": 0.913817822933197,
+      "learning_rate": 8.989837042228842e-05,
+      "loss": 1.8618,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3042664301901665,
+      "grad_norm": 1.0107851028442383,
+      "learning_rate": 8.988084808130367e-05,
+      "loss": 1.8706,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3047919335238283,
+      "grad_norm": 1.2873750925064087,
+      "learning_rate": 8.986332574031892e-05,
+      "loss": 1.8451,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3053174368574901,
+      "grad_norm": 0.9408276677131653,
+      "learning_rate": 8.984580339933417e-05,
+      "loss": 1.8417,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3058429401911518,
+      "grad_norm": 1.078941822052002,
+      "learning_rate": 8.98282810583494e-05,
+      "loss": 1.8526,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3063684435248136,
+      "grad_norm": 0.9041505455970764,
+      "learning_rate": 8.981075871736465e-05,
+      "loss": 1.8779,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3068939468584754,
+      "grad_norm": 0.8624897599220276,
+      "learning_rate": 8.979323637637988e-05,
+      "loss": 1.7867,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3074194501921372,
+      "grad_norm": 0.9410212635993958,
+      "learning_rate": 8.977571403539513e-05,
+      "loss": 1.8498,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3079449535257989,
+      "grad_norm": 0.9149646162986755,
+      "learning_rate": 8.975819169441038e-05,
+      "loss": 1.8229,
+      "step": 1172
+    },
+    {
+      "epoch": 0.30847045685946073,
+      "grad_norm": 0.7817291021347046,
+      "learning_rate": 8.974066935342562e-05,
+      "loss": 1.8886,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3089959601931225,
+      "grad_norm": 1.3264005184173584,
+      "learning_rate": 8.972314701244087e-05,
+      "loss": 1.851,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3095214635267842,
+      "grad_norm": 1.0288749933242798,
+      "learning_rate": 8.970562467145612e-05,
+      "loss": 1.8514,
+      "step": 1178
+    },
+    {
+      "epoch": 0.310046966860446,
+      "grad_norm": 0.9613611698150635,
+      "learning_rate": 8.968810233047135e-05,
+      "loss": 1.8633,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3105724701941078,
+      "grad_norm": 1.0935230255126953,
+      "learning_rate": 8.96705799894866e-05,
+      "loss": 1.8175,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3110979735277696,
+      "grad_norm": 0.821371853351593,
+      "learning_rate": 8.965305764850185e-05,
+      "loss": 1.8275,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3116234768614313,
+      "grad_norm": 1.0035851001739502,
+      "learning_rate": 8.96355353075171e-05,
+      "loss": 1.8379,
+      "step": 1186
+    },
+    {
+      "epoch": 0.31214898019509313,
+      "grad_norm": 1.4299391508102417,
+      "learning_rate": 8.961801296653233e-05,
+      "loss": 1.8456,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3126744835287549,
+      "grad_norm": 1.1284465789794922,
+      "learning_rate": 8.960049062554758e-05,
+      "loss": 1.8317,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3131999868624167,
+      "grad_norm": 0.8965946435928345,
+      "learning_rate": 8.958296828456282e-05,
+      "loss": 1.8265,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3137254901960784,
+      "grad_norm": 0.7729229927062988,
+      "learning_rate": 8.956544594357806e-05,
+      "loss": 1.8479,
+      "step": 1194
+    },
+    {
+      "epoch": 0.31425099352974023,
+      "grad_norm": 1.0302493572235107,
+      "learning_rate": 8.95479236025933e-05,
+      "loss": 1.8563,
+      "step": 1196
+    },
+    {
+      "epoch": 0.314776496863402,
+      "grad_norm": 1.2506144046783447,
+      "learning_rate": 8.953040126160855e-05,
+      "loss": 1.8672,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3153020001970637,
+      "grad_norm": 1.190073847770691,
+      "learning_rate": 8.95128789206238e-05,
+      "loss": 1.8476,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3153020001970637,
+      "eval_loss": 1.7993388175964355,
+      "eval_runtime": 487.2115,
+      "eval_samples_per_second": 249.972,
+      "eval_steps_per_second": 31.247,
+      "step": 1200
+    },
+    {
+      "epoch": 0.31582750353072553,
+      "grad_norm": 1.772186279296875,
+      "learning_rate": 8.949535657963905e-05,
+      "loss": 1.8635,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3163530068643873,
+      "grad_norm": 0.8721492290496826,
+      "learning_rate": 8.94778342386543e-05,
+      "loss": 1.8243,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3168785101980491,
+      "grad_norm": 0.9871326684951782,
+      "learning_rate": 8.946031189766953e-05,
+      "loss": 1.8491,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3174040135317108,
+      "grad_norm": 0.8752848505973816,
+      "learning_rate": 8.944278955668478e-05,
+      "loss": 1.8638,
+      "step": 1208
+    },
+    {
+      "epoch": 0.31792951686537263,
+      "grad_norm": 0.9636614918708801,
+      "learning_rate": 8.942526721570003e-05,
+      "loss": 1.8779,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3184550201990344,
+      "grad_norm": 1.027335524559021,
+      "learning_rate": 8.940774487471527e-05,
+      "loss": 1.857,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3189805235326962,
+      "grad_norm": 0.8861249089241028,
+      "learning_rate": 8.939022253373051e-05,
+      "loss": 1.875,
+      "step": 1214
+    },
+    {
+      "epoch": 0.31950602686635793,
+      "grad_norm": 0.8626223802566528,
+      "learning_rate": 8.937270019274575e-05,
+      "loss": 1.8096,
+      "step": 1216
+    },
+    {
+      "epoch": 0.32003153020001973,
+      "grad_norm": 0.8333232402801514,
+      "learning_rate": 8.9355177851761e-05,
+      "loss": 1.8275,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3205570335336815,
+      "grad_norm": 0.7367919087409973,
+      "learning_rate": 8.933765551077624e-05,
+      "loss": 1.8653,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3210825368673432,
+      "grad_norm": 1.1696327924728394,
+      "learning_rate": 8.932013316979148e-05,
+      "loss": 1.8558,
+      "step": 1222
+    },
+    {
+      "epoch": 0.32160804020100503,
+      "grad_norm": 0.9526751637458801,
+      "learning_rate": 8.930261082880673e-05,
+      "loss": 1.8429,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3221335435346668,
+      "grad_norm": 1.0763540267944336,
+      "learning_rate": 8.928508848782198e-05,
+      "loss": 1.8365,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3226590468683286,
+      "grad_norm": 0.8945477604866028,
+      "learning_rate": 8.926756614683723e-05,
+      "loss": 1.8303,
+      "step": 1228
+    },
+    {
+      "epoch": 0.32318455020199033,
+      "grad_norm": 0.8092701435089111,
+      "learning_rate": 8.925004380585247e-05,
+      "loss": 1.8454,
+      "step": 1230
+    },
+    {
+      "epoch": 0.32371005353565213,
+      "grad_norm": 1.0602396726608276,
+      "learning_rate": 8.923252146486771e-05,
+      "loss": 1.8256,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3242355568693139,
+      "grad_norm": 0.8251585960388184,
+      "learning_rate": 8.921499912388296e-05,
+      "loss": 1.8057,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3247610602029757,
+      "grad_norm": 1.2272833585739136,
+      "learning_rate": 8.91974767828982e-05,
+      "loss": 1.8549,
+      "step": 1236
+    },
+    {
+      "epoch": 0.32528656353663743,
+      "grad_norm": 0.9338359832763672,
+      "learning_rate": 8.917995444191345e-05,
+      "loss": 1.8618,
+      "step": 1238
+    },
+    {
+      "epoch": 0.32581206687029923,
+      "grad_norm": 0.8794339299201965,
+      "learning_rate": 8.916243210092868e-05,
+      "loss": 1.846,
+      "step": 1240
+    },
+    {
+      "epoch": 0.326337570203961,
+      "grad_norm": 0.9014391899108887,
+      "learning_rate": 8.914490975994393e-05,
+      "loss": 1.8469,
+      "step": 1242
+    },
+    {
+      "epoch": 0.32686307353762273,
+      "grad_norm": 0.9050634503364563,
+      "learning_rate": 8.912738741895918e-05,
+      "loss": 1.798,
+      "step": 1244
+    },
+    {
+      "epoch": 0.32738857687128453,
+      "grad_norm": 0.9596816897392273,
+      "learning_rate": 8.910986507797441e-05,
+      "loss": 1.8617,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3279140802049463,
+      "grad_norm": 0.8555053472518921,
+      "learning_rate": 8.909234273698966e-05,
+      "loss": 1.8262,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3284395835386081,
+      "grad_norm": 0.7877684831619263,
+      "learning_rate": 8.907482039600491e-05,
+      "loss": 1.8536,
+      "step": 1250
+    },
+    {
+      "epoch": 0.32896508687226983,
+      "grad_norm": 0.9233472347259521,
+      "learning_rate": 8.905729805502016e-05,
+      "loss": 1.8681,
+      "step": 1252
+    },
+    {
+      "epoch": 0.32949059020593163,
+      "grad_norm": 0.8724532127380371,
+      "learning_rate": 8.90397757140354e-05,
+      "loss": 1.8169,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3300160935395934,
+      "grad_norm": 0.8664790987968445,
+      "learning_rate": 8.902225337305065e-05,
+      "loss": 1.8327,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3305415968732552,
+      "grad_norm": 1.0527539253234863,
+      "learning_rate": 8.900473103206589e-05,
+      "loss": 1.849,
+      "step": 1258
+    },
+    {
+      "epoch": 0.33106710020691693,
+      "grad_norm": 1.0830262899398804,
+      "learning_rate": 8.898720869108113e-05,
+      "loss": 1.8293,
+      "step": 1260
+    },
+    {
+      "epoch": 0.33159260354057873,
+      "grad_norm": 0.9514210820198059,
+      "learning_rate": 8.896968635009638e-05,
+      "loss": 1.8509,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3321181068742405,
+      "grad_norm": 1.2806978225708008,
+      "learning_rate": 8.895216400911163e-05,
+      "loss": 1.8622,
+      "step": 1264
+    },
+    {
+      "epoch": 0.33264361020790223,
+      "grad_norm": 0.8732459545135498,
+      "learning_rate": 8.893464166812686e-05,
+      "loss": 1.8385,
+      "step": 1266
+    },
+    {
+      "epoch": 0.33316911354156403,
+      "grad_norm": 0.9644619822502136,
+      "learning_rate": 8.891711932714211e-05,
+      "loss": 1.8368,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3336946168752258,
+      "grad_norm": 0.9549365043640137,
+      "learning_rate": 8.889959698615736e-05,
+      "loss": 1.8363,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3342201202088876,
+      "grad_norm": 0.9528286457061768,
+      "learning_rate": 8.888207464517259e-05,
+      "loss": 1.8548,
+      "step": 1272
+    },
+    {
+      "epoch": 0.33474562354254933,
+      "grad_norm": 1.0315399169921875,
+      "learning_rate": 8.886455230418784e-05,
+      "loss": 1.8354,
+      "step": 1274
+    },
+    {
+      "epoch": 0.33527112687621113,
+      "grad_norm": 0.8084585070610046,
+      "learning_rate": 8.884702996320309e-05,
+      "loss": 1.8423,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3357966302098729,
+      "grad_norm": 1.1312843561172485,
+      "learning_rate": 8.882950762221833e-05,
+      "loss": 1.8374,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3363221335435347,
+      "grad_norm": 1.1717549562454224,
+      "learning_rate": 8.881198528123358e-05,
+      "loss": 1.8204,
+      "step": 1280
+    },
+    {
+      "epoch": 0.33684763687719643,
+      "grad_norm": 0.8000698089599609,
+      "learning_rate": 8.879446294024883e-05,
+      "loss": 1.8577,
+      "step": 1282
+    },
+    {
+      "epoch": 0.33737314021085824,
+      "grad_norm": 1.3130030632019043,
+      "learning_rate": 8.877694059926406e-05,
+      "loss": 1.8072,
+      "step": 1284
+    },
+    {
+      "epoch": 0.33789864354452,
+      "grad_norm": 0.8949963450431824,
+      "learning_rate": 8.875941825827931e-05,
+      "loss": 1.8586,
+      "step": 1286
+    },
+    {
+      "epoch": 0.33842414687818173,
+      "grad_norm": 0.8775150775909424,
+      "learning_rate": 8.874189591729456e-05,
+      "loss": 1.831,
+      "step": 1288
+    },
+    {
+      "epoch": 0.33894965021184353,
+      "grad_norm": 0.8946396708488464,
+      "learning_rate": 8.872437357630979e-05,
+      "loss": 1.8243,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3394751535455053,
+      "grad_norm": 1.271799921989441,
+      "learning_rate": 8.870685123532504e-05,
+      "loss": 1.8174,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3400006568791671,
+      "grad_norm": 0.8557697534561157,
+      "learning_rate": 8.868932889434029e-05,
+      "loss": 1.8429,
+      "step": 1294
+    },
+    {
+      "epoch": 0.34052616021282883,
+      "grad_norm": 0.9884776473045349,
+      "learning_rate": 8.867180655335554e-05,
+      "loss": 1.7985,
+      "step": 1296
+    },
+    {
+      "epoch": 0.34105166354649064,
+      "grad_norm": 0.9385315775871277,
+      "learning_rate": 8.865428421237077e-05,
+      "loss": 1.7948,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3415771668801524,
+      "grad_norm": 1.1224939823150635,
+      "learning_rate": 8.863676187138602e-05,
+      "loss": 1.8564,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3421026702138142,
+      "grad_norm": 0.9227058291435242,
+      "learning_rate": 8.861923953040126e-05,
+      "loss": 1.8432,
+      "step": 1302
+    },
+    {
+      "epoch": 0.34262817354747593,
+      "grad_norm": 1.0591615438461304,
+      "learning_rate": 8.860171718941651e-05,
+      "loss": 1.8267,
+      "step": 1304
+    },
+    {
+      "epoch": 0.34315367688113774,
+      "grad_norm": 0.8901565670967102,
+      "learning_rate": 8.858419484843176e-05,
+      "loss": 1.8452,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3436791802147995,
+      "grad_norm": 0.7925954461097717,
+      "learning_rate": 8.856667250744701e-05,
+      "loss": 1.8395,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3442046835484613,
+      "grad_norm": 0.8542584776878357,
+      "learning_rate": 8.854915016646224e-05,
+      "loss": 1.821,
+      "step": 1310
+    },
+    {
+      "epoch": 0.34473018688212304,
+      "grad_norm": 0.9138728380203247,
+      "learning_rate": 8.853162782547749e-05,
+      "loss": 1.8466,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3452556902157848,
+      "grad_norm": 1.0735788345336914,
+      "learning_rate": 8.851410548449274e-05,
+      "loss": 1.8318,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3457811935494466,
+      "grad_norm": 1.3310229778289795,
+      "learning_rate": 8.849658314350797e-05,
+      "loss": 1.8334,
+      "step": 1316
+    },
+    {
+      "epoch": 0.34630669688310833,
+      "grad_norm": 0.7600061297416687,
+      "learning_rate": 8.847906080252322e-05,
+      "loss": 1.8089,
+      "step": 1318
+    },
+    {
+      "epoch": 0.34683220021677014,
+      "grad_norm": 0.818154513835907,
+      "learning_rate": 8.846153846153847e-05,
+      "loss": 1.8577,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3473577035504319,
+      "grad_norm": 1.0234004259109497,
+      "learning_rate": 8.844401612055371e-05,
+      "loss": 1.8364,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3478832068840937,
+      "grad_norm": 0.880425751209259,
+      "learning_rate": 8.842649377956895e-05,
+      "loss": 1.8671,
+      "step": 1324
+    },
+    {
+      "epoch": 0.34840871021775544,
+      "grad_norm": 0.8950909376144409,
+      "learning_rate": 8.84089714385842e-05,
+      "loss": 1.8217,
+      "step": 1326
+    },
+    {
+      "epoch": 0.34893421355141724,
+      "grad_norm": 1.029801607131958,
+      "learning_rate": 8.839144909759944e-05,
+      "loss": 1.8273,
+      "step": 1328
+    },
+    {
+      "epoch": 0.349459716885079,
+      "grad_norm": 1.1284875869750977,
+      "learning_rate": 8.837392675661469e-05,
+      "loss": 1.8523,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3499852202187408,
+      "grad_norm": 1.3472214937210083,
+      "learning_rate": 8.835640441562994e-05,
+      "loss": 1.8878,
+      "step": 1332
+    },
+    {
+      "epoch": 0.35051072355240254,
+      "grad_norm": 0.8898762464523315,
+      "learning_rate": 8.833888207464518e-05,
+      "loss": 1.8225,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3510362268860643,
+      "grad_norm": 1.2737003564834595,
+      "learning_rate": 8.832135973366042e-05,
+      "loss": 1.7853,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3515617302197261,
+      "grad_norm": 0.9682241678237915,
+      "learning_rate": 8.830383739267567e-05,
+      "loss": 1.8502,
+      "step": 1338
+    },
+    {
+      "epoch": 0.35208723355338784,
+      "grad_norm": 0.8494237661361694,
+      "learning_rate": 8.828631505169091e-05,
+      "loss": 1.8515,
+      "step": 1340
+    },
+    {
+      "epoch": 0.35261273688704964,
+      "grad_norm": 0.923283040523529,
+      "learning_rate": 8.826879271070615e-05,
+      "loss": 1.849,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3531382402207114,
+      "grad_norm": 1.0368821620941162,
+      "learning_rate": 8.82512703697214e-05,
+      "loss": 1.8228,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3536637435543732,
+      "grad_norm": 0.76881343126297,
+      "learning_rate": 8.823374802873664e-05,
+      "loss": 1.8156,
+      "step": 1346
+    },
+    {
+      "epoch": 0.35418924688803494,
+      "grad_norm": 0.7315630316734314,
+      "learning_rate": 8.821622568775189e-05,
+      "loss": 1.8068,
+      "step": 1348
+    },
+    {
+      "epoch": 0.35471475022169674,
+      "grad_norm": 1.254550814628601,
+      "learning_rate": 8.819870334676712e-05,
+      "loss": 1.8498,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3552402535553585,
+      "grad_norm": 1.1354317665100098,
+      "learning_rate": 8.818118100578237e-05,
+      "loss": 1.8547,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3557657568890203,
+      "grad_norm": 1.0078952312469482,
+      "learning_rate": 8.816365866479762e-05,
+      "loss": 1.8635,
+      "step": 1354
+    },
+    {
+      "epoch": 0.35629126022268204,
+      "grad_norm": 1.4484366178512573,
+      "learning_rate": 8.814613632381287e-05,
+      "loss": 1.8405,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3568167635563438,
+      "grad_norm": 0.8407228589057922,
+      "learning_rate": 8.812861398282811e-05,
+      "loss": 1.8295,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3573422668900056,
+      "grad_norm": 0.9024233818054199,
+      "learning_rate": 8.811109164184336e-05,
+      "loss": 1.8052,
+      "step": 1360
+    },
+    {
+      "epoch": 0.35786777022366734,
+      "grad_norm": 0.9681188464164734,
+      "learning_rate": 8.80935693008586e-05,
+      "loss": 1.8024,
+      "step": 1362
+    },
+    {
+      "epoch": 0.35839327355732914,
+      "grad_norm": 0.9130085706710815,
+      "learning_rate": 8.807604695987384e-05,
+      "loss": 1.8159,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3589187768909909,
+      "grad_norm": 0.938353419303894,
+      "learning_rate": 8.805852461888909e-05,
+      "loss": 1.8134,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3594442802246527,
+      "grad_norm": 0.8700679540634155,
+      "learning_rate": 8.804100227790433e-05,
+      "loss": 1.8091,
+      "step": 1368
+    },
+    {
+      "epoch": 0.35996978355831444,
+      "grad_norm": 0.8863296508789062,
+      "learning_rate": 8.802347993691957e-05,
+      "loss": 1.7961,
+      "step": 1370
+    },
+    {
+      "epoch": 0.36049528689197624,
+      "grad_norm": 0.9155923128128052,
+      "learning_rate": 8.800595759593482e-05,
+      "loss": 1.8098,
+      "step": 1372
+    },
+    {
+      "epoch": 0.361020790225638,
+      "grad_norm": 1.020551323890686,
+      "learning_rate": 8.798843525495007e-05,
+      "loss": 1.848,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3615462935592998,
+      "grad_norm": 0.9836577773094177,
+      "learning_rate": 8.79709129139653e-05,
+      "loss": 1.7839,
+      "step": 1376
+    },
+    {
+      "epoch": 0.36207179689296154,
+      "grad_norm": 0.9969834089279175,
+      "learning_rate": 8.795339057298055e-05,
+      "loss": 1.8062,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3625973002266233,
+      "grad_norm": 0.8620086312294006,
+      "learning_rate": 8.79358682319958e-05,
+      "loss": 1.7944,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3631228035602851,
+      "grad_norm": 1.2116692066192627,
+      "learning_rate": 8.791834589101104e-05,
+      "loss": 1.8718,
+      "step": 1382
+    },
+    {
+      "epoch": 0.36364830689394684,
+      "grad_norm": 0.8402097225189209,
+      "learning_rate": 8.790082355002629e-05,
+      "loss": 1.8276,
+      "step": 1384
+    },
+    {
+      "epoch": 0.36417381022760864,
+      "grad_norm": 0.9271780848503113,
+      "learning_rate": 8.788330120904154e-05,
+      "loss": 1.8222,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3646993135612704,
+      "grad_norm": 0.8769554495811462,
+      "learning_rate": 8.786577886805677e-05,
+      "loss": 1.8416,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3652248168949322,
+      "grad_norm": 0.9306502938270569,
+      "learning_rate": 8.784825652707202e-05,
+      "loss": 1.8506,
+      "step": 1390
+    },
+    {
+      "epoch": 0.36575032022859394,
+      "grad_norm": 0.8423568606376648,
+      "learning_rate": 8.783073418608726e-05,
+      "loss": 1.8617,
+      "step": 1392
+    },
+    {
+      "epoch": 0.36627582356225574,
+      "grad_norm": 0.9485574960708618,
+      "learning_rate": 8.78132118451025e-05,
+      "loss": 1.8448,
+      "step": 1394
+    },
+    {
+      "epoch": 0.3668013268959175,
+      "grad_norm": 1.1368005275726318,
+      "learning_rate": 8.779568950411775e-05,
+      "loss": 1.8217,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3673268302295793,
+      "grad_norm": 0.9294119477272034,
+      "learning_rate": 8.7778167163133e-05,
+      "loss": 1.8099,
+      "step": 1398
+    },
+    {
+      "epoch": 0.36785233356324104,
+      "grad_norm": 0.8389936685562134,
+      "learning_rate": 8.776064482214825e-05,
+      "loss": 1.8601,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3683778368969028,
+      "grad_norm": 0.7817425727844238,
+      "learning_rate": 8.77431224811635e-05,
+      "loss": 1.8218,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3689033402305646,
+      "grad_norm": 1.1421295404434204,
+      "learning_rate": 8.772560014017873e-05,
+      "loss": 1.8673,
+      "step": 1404
+    },
+    {
+      "epoch": 0.36942884356422634,
+      "grad_norm": 0.8209173083305359,
+      "learning_rate": 8.770807779919397e-05,
+      "loss": 1.8298,
+      "step": 1406
+    },
+    {
+      "epoch": 0.36995434689788814,
+      "grad_norm": 1.0874011516571045,
+      "learning_rate": 8.769055545820922e-05,
+      "loss": 1.8008,
+      "step": 1408
+    },
+    {
+      "epoch": 0.3704798502315499,
+      "grad_norm": 0.839116096496582,
+      "learning_rate": 8.767303311722447e-05,
+      "loss": 1.8511,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3710053535652117,
+      "grad_norm": 0.956777036190033,
+      "learning_rate": 8.765551077623972e-05,
+      "loss": 1.8761,
+      "step": 1412
+    },
+    {
+      "epoch": 0.37153085689887344,
+      "grad_norm": 0.7702937722206116,
+      "learning_rate": 8.763798843525497e-05,
+      "loss": 1.8231,
+      "step": 1414
+    },
+    {
+      "epoch": 0.37205636023253524,
+      "grad_norm": 0.8248230814933777,
+      "learning_rate": 8.76204660942702e-05,
+      "loss": 1.8538,
+      "step": 1416
+    },
+    {
+      "epoch": 0.372581863566197,
+      "grad_norm": 1.0771416425704956,
+      "learning_rate": 8.760294375328543e-05,
+      "loss": 1.8515,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3731073668998588,
+      "grad_norm": 0.8044272661209106,
+      "learning_rate": 8.758542141230068e-05,
+      "loss": 1.8295,
+      "step": 1420
+    },
+    {
+      "epoch": 0.37363287023352054,
+      "grad_norm": 1.0227196216583252,
+      "learning_rate": 8.756789907131593e-05,
+      "loss": 1.8659,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3741583735671823,
+      "grad_norm": 0.8310641646385193,
+      "learning_rate": 8.755037673033118e-05,
+      "loss": 1.8286,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3746838769008441,
+      "grad_norm": 0.9817164540290833,
+      "learning_rate": 8.753285438934642e-05,
+      "loss": 1.8863,
+      "step": 1426
+    },
+    {
+      "epoch": 0.37520938023450584,
+      "grad_norm": 0.872424840927124,
+      "learning_rate": 8.751533204836167e-05,
+      "loss": 1.8489,
+      "step": 1428
+    },
+    {
+      "epoch": 0.37573488356816764,
+      "grad_norm": 0.9776557087898254,
+      "learning_rate": 8.74978097073769e-05,
+      "loss": 1.8536,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3762603869018294,
+      "grad_norm": 1.3488025665283203,
+      "learning_rate": 8.748028736639215e-05,
+      "loss": 1.8432,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3767858902354912,
+      "grad_norm": 0.8517011404037476,
+      "learning_rate": 8.74627650254074e-05,
+      "loss": 1.847,
+      "step": 1434
+    },
+    {
+      "epoch": 0.37731139356915294,
+      "grad_norm": 0.8631575703620911,
+      "learning_rate": 8.744524268442265e-05,
+      "loss": 1.8053,
+      "step": 1436
+    },
+    {
+      "epoch": 0.37783689690281475,
+      "grad_norm": 0.881100058555603,
+      "learning_rate": 8.74277203434379e-05,
+      "loss": 1.8662,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3783624002364765,
+      "grad_norm": 0.8032435178756714,
+      "learning_rate": 8.741019800245314e-05,
+      "loss": 1.8438,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3788879035701383,
+      "grad_norm": 0.7921327948570251,
+      "learning_rate": 8.739267566146838e-05,
+      "loss": 1.8685,
+      "step": 1442
+    },
+    {
+      "epoch": 0.37941340690380004,
+      "grad_norm": 1.060738444328308,
+      "learning_rate": 8.737515332048361e-05,
+      "loss": 1.8365,
+      "step": 1444
+    },
+    {
+      "epoch": 0.37993891023746185,
+      "grad_norm": 1.0198917388916016,
+      "learning_rate": 8.735763097949886e-05,
+      "loss": 1.8278,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3804644135711236,
+      "grad_norm": 0.9688281416893005,
+      "learning_rate": 8.73401086385141e-05,
+      "loss": 1.8513,
+      "step": 1448
+    },
+    {
+      "epoch": 0.38098991690478534,
+      "grad_norm": 1.2723430395126343,
+      "learning_rate": 8.732258629752935e-05,
+      "loss": 1.7984,
+      "step": 1450
+    },
+    {
+      "epoch": 0.38151542023844714,
+      "grad_norm": 0.8690189123153687,
+      "learning_rate": 8.73050639565446e-05,
+      "loss": 1.8288,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3820409235721089,
+      "grad_norm": 0.9124467968940735,
+      "learning_rate": 8.728754161555985e-05,
+      "loss": 1.8684,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3825664269057707,
+      "grad_norm": 1.1950373649597168,
+      "learning_rate": 8.727001927457508e-05,
+      "loss": 1.8431,
+      "step": 1456
+    },
+    {
+      "epoch": 0.38309193023943244,
+      "grad_norm": 0.7846235632896423,
+      "learning_rate": 8.725249693359033e-05,
+      "loss": 1.8409,
+      "step": 1458
+    },
+    {
+      "epoch": 0.38361743357309425,
+      "grad_norm": 1.450654149055481,
+      "learning_rate": 8.723497459260558e-05,
+      "loss": 1.844,
+      "step": 1460
+    },
+    {
+      "epoch": 0.384142936906756,
+      "grad_norm": 1.0545793771743774,
+      "learning_rate": 8.721745225162083e-05,
+      "loss": 1.8332,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3846684402404178,
+      "grad_norm": 1.000705599784851,
+      "learning_rate": 8.719992991063607e-05,
+      "loss": 1.8486,
+      "step": 1464
+    },
+    {
+      "epoch": 0.38519394357407954,
+      "grad_norm": 1.2795532941818237,
+      "learning_rate": 8.718240756965132e-05,
+      "loss": 1.8323,
+      "step": 1466
+    },
+    {
+      "epoch": 0.38571944690774135,
+      "grad_norm": 0.7551513314247131,
+      "learning_rate": 8.716488522866655e-05,
+      "loss": 1.8573,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3862449502414031,
+      "grad_norm": 1.2810308933258057,
+      "learning_rate": 8.714736288768179e-05,
+      "loss": 1.8167,
+      "step": 1470
+    },
+    {
+      "epoch": 0.38677045357506484,
+      "grad_norm": 1.0538434982299805,
+      "learning_rate": 8.712984054669704e-05,
+      "loss": 1.8501,
+      "step": 1472
+    },
+    {
+      "epoch": 0.38729595690872665,
+      "grad_norm": 1.2018911838531494,
+      "learning_rate": 8.711231820571228e-05,
+      "loss": 1.8291,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3878214602423884,
+      "grad_norm": 1.4515736103057861,
+      "learning_rate": 8.709479586472753e-05,
+      "loss": 1.8728,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3883469635760502,
+      "grad_norm": 0.855747640132904,
+      "learning_rate": 8.707727352374278e-05,
+      "loss": 1.8048,
+      "step": 1478
+    },
+    {
+      "epoch": 0.38887246690971194,
+      "grad_norm": 1.3377580642700195,
+      "learning_rate": 8.705975118275803e-05,
+      "loss": 1.8599,
+      "step": 1480
+    },
+    {
+      "epoch": 0.38939797024337375,
+      "grad_norm": 0.9842968583106995,
+      "learning_rate": 8.704222884177326e-05,
+      "loss": 1.8038,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3899234735770355,
+      "grad_norm": 1.4240106344223022,
+      "learning_rate": 8.702470650078851e-05,
+      "loss": 1.8305,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3904489769106973,
+      "grad_norm": 0.7605730295181274,
+      "learning_rate": 8.700718415980376e-05,
+      "loss": 1.8321,
+      "step": 1486
+    },
+    {
+      "epoch": 0.39097448024435905,
+      "grad_norm": 0.9584787487983704,
+      "learning_rate": 8.6989661818819e-05,
+      "loss": 1.799,
+      "step": 1488
+    },
+    {
+      "epoch": 0.39149998357802085,
+      "grad_norm": 0.8087942004203796,
+      "learning_rate": 8.697213947783425e-05,
+      "loss": 1.8011,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3920254869116826,
+      "grad_norm": 0.7870105504989624,
+      "learning_rate": 8.69546171368495e-05,
+      "loss": 1.7972,
+      "step": 1492
+    },
+    {
+      "epoch": 0.39255099024534434,
+      "grad_norm": 1.1304738521575928,
+      "learning_rate": 8.693709479586473e-05,
+      "loss": 1.8088,
+      "step": 1494
+    },
+    {
+      "epoch": 0.39307649357900615,
+      "grad_norm": 0.8902273178100586,
+      "learning_rate": 8.691957245487997e-05,
+      "loss": 1.8101,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3936019969126679,
+      "grad_norm": 1.1424989700317383,
+      "learning_rate": 8.690205011389521e-05,
+      "loss": 1.8018,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3941275002463297,
+      "grad_norm": 0.9772897362709045,
+      "learning_rate": 8.688452777291046e-05,
+      "loss": 1.8191,
+      "step": 1500
+    },
+    {
+      "epoch": 0.39465300357999145,
+      "grad_norm": 0.9879363775253296,
+      "learning_rate": 8.686700543192571e-05,
+      "loss": 1.7934,
+      "step": 1502
+    },
+    {
+      "epoch": 0.39517850691365325,
+      "grad_norm": 0.8215435147285461,
+      "learning_rate": 8.684948309094096e-05,
+      "loss": 1.8125,
+      "step": 1504
+    },
+    {
+      "epoch": 0.395704010247315,
+      "grad_norm": 0.8453714847564697,
+      "learning_rate": 8.68319607499562e-05,
+      "loss": 1.8385,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3962295135809768,
+      "grad_norm": 0.9266200661659241,
+      "learning_rate": 8.681443840897144e-05,
+      "loss": 1.8537,
+      "step": 1508
+    },
+    {
+      "epoch": 0.39675501691463855,
+      "grad_norm": 1.2535603046417236,
+      "learning_rate": 8.679691606798669e-05,
+      "loss": 1.8138,
+      "step": 1510
+    },
+    {
+      "epoch": 0.39728052024830035,
+      "grad_norm": 1.0080575942993164,
+      "learning_rate": 8.677939372700193e-05,
+      "loss": 1.8193,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3978060235819621,
+      "grad_norm": 0.8419904112815857,
+      "learning_rate": 8.676187138601718e-05,
+      "loss": 1.845,
+      "step": 1514
+    },
+    {
+      "epoch": 0.39833152691562385,
+      "grad_norm": 1.2089139223098755,
+      "learning_rate": 8.674434904503243e-05,
+      "loss": 1.8274,
+      "step": 1516
+    },
+    {
+      "epoch": 0.39885703024928565,
+      "grad_norm": 0.9421194791793823,
+      "learning_rate": 8.672682670404768e-05,
+      "loss": 1.8132,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3993825335829474,
+      "grad_norm": 1.0286279916763306,
+      "learning_rate": 8.670930436306291e-05,
+      "loss": 1.8045,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3999080369166092,
+      "grad_norm": 1.1791476011276245,
+      "learning_rate": 8.669178202207814e-05,
+      "loss": 1.8148,
+      "step": 1522
+    },
+    {
+      "epoch": 0.40043354025027095,
+      "grad_norm": 0.9198878407478333,
+      "learning_rate": 8.667425968109339e-05,
+      "loss": 1.8261,
+      "step": 1524
+    },
+    {
+      "epoch": 0.40095904358393275,
+      "grad_norm": 1.0204558372497559,
+      "learning_rate": 8.665673734010864e-05,
+      "loss": 1.8324,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4014845469175945,
+      "grad_norm": 1.1662263870239258,
+      "learning_rate": 8.663921499912389e-05,
+      "loss": 1.7949,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4020100502512563,
+      "grad_norm": 1.2517145872116089,
+      "learning_rate": 8.662169265813913e-05,
+      "loss": 1.7924,
+      "step": 1530
+    },
+    {
+      "epoch": 0.40253555358491805,
+      "grad_norm": 0.7785090208053589,
+      "learning_rate": 8.660417031715438e-05,
+      "loss": 1.8023,
+      "step": 1532
+    },
+    {
+      "epoch": 0.40306105691857985,
+      "grad_norm": 0.8084584474563599,
+      "learning_rate": 8.658664797616962e-05,
+      "loss": 1.8132,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4035865602522416,
+      "grad_norm": 0.8784323930740356,
+      "learning_rate": 8.656912563518486e-05,
+      "loss": 1.8047,
+      "step": 1536
+    },
+    {
+      "epoch": 0.40411206358590335,
+      "grad_norm": 0.8626761436462402,
+      "learning_rate": 8.655160329420011e-05,
+      "loss": 1.7985,
+      "step": 1538
+    },
+    {
+      "epoch": 0.40463756691956515,
+      "grad_norm": 0.8983022570610046,
+      "learning_rate": 8.653408095321536e-05,
+      "loss": 1.8068,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4051630702532269,
+      "grad_norm": 1.2110192775726318,
+      "learning_rate": 8.65165586122306e-05,
+      "loss": 1.815,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4056885735868887,
+      "grad_norm": 0.9025999307632446,
+      "learning_rate": 8.649903627124585e-05,
+      "loss": 1.8305,
+      "step": 1544
+    },
+    {
+      "epoch": 0.40621407692055045,
+      "grad_norm": 0.8550492525100708,
+      "learning_rate": 8.648151393026109e-05,
+      "loss": 1.7915,
+      "step": 1546
+    },
+    {
+      "epoch": 0.40673958025421225,
+      "grad_norm": 0.8646672368049622,
+      "learning_rate": 8.646399158927632e-05,
+      "loss": 1.8338,
+      "step": 1548
+    },
+    {
+      "epoch": 0.407265083587874,
+      "grad_norm": 0.8966619372367859,
+      "learning_rate": 8.644646924829157e-05,
+      "loss": 1.8072,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4077905869215358,
+      "grad_norm": 0.9003387689590454,
+      "learning_rate": 8.642894690730682e-05,
+      "loss": 1.8041,
+      "step": 1552
+    },
+    {
+      "epoch": 0.40831609025519755,
+      "grad_norm": 1.035400152206421,
+      "learning_rate": 8.641142456632206e-05,
+      "loss": 1.8484,
+      "step": 1554
+    },
+    {
+      "epoch": 0.40884159358885935,
+      "grad_norm": 0.8281182050704956,
+      "learning_rate": 8.639390222533731e-05,
+      "loss": 1.7943,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4093670969225211,
+      "grad_norm": 0.8706338405609131,
+      "learning_rate": 8.637637988435256e-05,
+      "loss": 1.848,
+      "step": 1558
+    },
+    {
+      "epoch": 0.40989260025618285,
+      "grad_norm": 0.9510646462440491,
+      "learning_rate": 8.63588575433678e-05,
+      "loss": 1.81,
+      "step": 1560
+    },
+    {
+      "epoch": 0.41041810358984465,
+      "grad_norm": 1.2758607864379883,
+      "learning_rate": 8.634133520238304e-05,
+      "loss": 1.831,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4109436069235064,
+      "grad_norm": 0.8133296966552734,
+      "learning_rate": 8.632381286139829e-05,
+      "loss": 1.833,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4114691102571682,
+      "grad_norm": 0.8663495779037476,
+      "learning_rate": 8.630629052041354e-05,
+      "loss": 1.7999,
+      "step": 1566
+    },
+    {
+      "epoch": 0.41199461359082995,
+      "grad_norm": 0.8473132252693176,
+      "learning_rate": 8.628876817942878e-05,
+      "loss": 1.8148,
+      "step": 1568
+    },
+    {
+      "epoch": 0.41252011692449175,
+      "grad_norm": 0.7791121006011963,
+      "learning_rate": 8.627124583844402e-05,
+      "loss": 1.8107,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4130456202581535,
+      "grad_norm": 0.8510565161705017,
+      "learning_rate": 8.625372349745927e-05,
+      "loss": 1.8434,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4135711235918153,
+      "grad_norm": 0.7872109413146973,
+      "learning_rate": 8.62362011564745e-05,
+      "loss": 1.86,
+      "step": 1574
+    },
+    {
+      "epoch": 0.41409662692547705,
+      "grad_norm": 0.9160329699516296,
+      "learning_rate": 8.621867881548975e-05,
+      "loss": 1.8408,
+      "step": 1576
+    },
+    {
+      "epoch": 0.41462213025913885,
+      "grad_norm": 0.9173769354820251,
+      "learning_rate": 8.6201156474505e-05,
+      "loss": 1.8057,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4151476335928006,
+      "grad_norm": 1.086470603942871,
+      "learning_rate": 8.618363413352024e-05,
+      "loss": 1.8199,
+      "step": 1580
+    },
+    {
+      "epoch": 0.41567313692646235,
+      "grad_norm": 0.8830829858779907,
+      "learning_rate": 8.616611179253549e-05,
+      "loss": 1.8107,
+      "step": 1582
+    },
+    {
+      "epoch": 0.41619864026012415,
+      "grad_norm": 0.8462435007095337,
+      "learning_rate": 8.614858945155074e-05,
+      "loss": 1.8159,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4167241435937859,
+      "grad_norm": 0.9439370632171631,
+      "learning_rate": 8.613106711056597e-05,
+      "loss": 1.8311,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4172496469274477,
+      "grad_norm": 0.933134138584137,
+      "learning_rate": 8.611354476958122e-05,
+      "loss": 1.8203,
+      "step": 1588
+    },
+    {
+      "epoch": 0.41777515026110945,
+      "grad_norm": 0.7939304709434509,
+      "learning_rate": 8.609602242859647e-05,
+      "loss": 1.7991,
+      "step": 1590
+    },
+    {
+      "epoch": 0.41830065359477125,
+      "grad_norm": 0.9367759823799133,
+      "learning_rate": 8.607850008761171e-05,
+      "loss": 1.8391,
+      "step": 1592
+    },
+    {
+      "epoch": 0.418826156928433,
+      "grad_norm": 0.8476933240890503,
+      "learning_rate": 8.606097774662696e-05,
+      "loss": 1.8135,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4193516602620948,
+      "grad_norm": 0.9385167360305786,
+      "learning_rate": 8.60434554056422e-05,
+      "loss": 1.8396,
+      "step": 1596
+    },
+    {
+      "epoch": 0.41987716359575655,
+      "grad_norm": 0.9880960583686829,
+      "learning_rate": 8.602593306465744e-05,
+      "loss": 1.8282,
+      "step": 1598
+    },
+    {
+      "epoch": 0.42040266692941836,
+      "grad_norm": 1.3682297468185425,
+      "learning_rate": 8.600841072367268e-05,
+      "loss": 1.8063,
+      "step": 1600
+    },
+    {
+      "epoch": 0.42040266692941836,
+      "eval_loss": 1.785941243171692,
+      "eval_runtime": 487.1976,
+      "eval_samples_per_second": 249.979,
+      "eval_steps_per_second": 31.248,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4209281702630801,
+      "grad_norm": 0.8974788784980774,
+      "learning_rate": 8.599088838268792e-05,
+      "loss": 1.8264,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4214536735967419,
+      "grad_norm": 1.3051773309707642,
+      "learning_rate": 8.597336604170317e-05,
+      "loss": 1.8204,
+      "step": 1604
+    },
+    {
+      "epoch": 0.42197917693040365,
+      "grad_norm": 0.8313725590705872,
+      "learning_rate": 8.595584370071842e-05,
+      "loss": 1.8276,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4225046802640654,
+      "grad_norm": 0.7489058375358582,
+      "learning_rate": 8.593832135973367e-05,
+      "loss": 1.8039,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4230301835977272,
+      "grad_norm": 1.303904414176941,
+      "learning_rate": 8.592079901874891e-05,
+      "loss": 1.8117,
+      "step": 1610
+    },
+    {
+      "epoch": 0.42355568693138895,
+      "grad_norm": 0.7930120229721069,
+      "learning_rate": 8.590327667776415e-05,
+      "loss": 1.816,
+      "step": 1612
+    },
+    {
+      "epoch": 0.42408119026505076,
+      "grad_norm": 1.1683326959609985,
+      "learning_rate": 8.58857543367794e-05,
+      "loss": 1.8485,
+      "step": 1614
+    },
+    {
+      "epoch": 0.4246066935987125,
+      "grad_norm": 1.129786491394043,
+      "learning_rate": 8.586823199579464e-05,
+      "loss": 1.8091,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4251321969323743,
+      "grad_norm": 1.006664752960205,
+      "learning_rate": 8.585070965480989e-05,
+      "loss": 1.801,
+      "step": 1618
+    },
+    {
+      "epoch": 0.42565770026603605,
+      "grad_norm": 1.2593824863433838,
+      "learning_rate": 8.583318731382514e-05,
+      "loss": 1.874,
+      "step": 1620
+    },
+    {
+      "epoch": 0.42618320359969786,
+      "grad_norm": 0.7356145977973938,
+      "learning_rate": 8.581566497284037e-05,
+      "loss": 1.8156,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4267087069333596,
+      "grad_norm": 1.3224732875823975,
+      "learning_rate": 8.579814263185562e-05,
+      "loss": 1.8485,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4272342102670214,
+      "grad_norm": 1.0026780366897583,
+      "learning_rate": 8.578062029087085e-05,
+      "loss": 1.8051,
+      "step": 1626
+    },
+    {
+      "epoch": 0.42775971360068316,
+      "grad_norm": 1.235370397567749,
+      "learning_rate": 8.57630979498861e-05,
+      "loss": 1.8267,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4282852169343449,
+      "grad_norm": 1.0629823207855225,
+      "learning_rate": 8.574557560890135e-05,
+      "loss": 1.7937,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4288107202680067,
+      "grad_norm": 0.9405169486999512,
+      "learning_rate": 8.57280532679166e-05,
+      "loss": 1.8023,
+      "step": 1632
+    },
+    {
+      "epoch": 0.42933622360166845,
+      "grad_norm": 1.5205514430999756,
+      "learning_rate": 8.571053092693184e-05,
+      "loss": 1.8112,
+      "step": 1634
+    },
+    {
+      "epoch": 0.42986172693533026,
+      "grad_norm": 1.038159728050232,
+      "learning_rate": 8.569300858594709e-05,
+      "loss": 1.7833,
+      "step": 1636
+    },
+    {
+      "epoch": 0.430387230268992,
+      "grad_norm": 0.8901309370994568,
+      "learning_rate": 8.567548624496233e-05,
+      "loss": 1.8241,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4309127336026538,
+      "grad_norm": 1.166390061378479,
+      "learning_rate": 8.565796390397757e-05,
+      "loss": 1.7899,
+      "step": 1640
+    },
+    {
+      "epoch": 0.43143823693631556,
+      "grad_norm": 1.0582796335220337,
+      "learning_rate": 8.564044156299282e-05,
+      "loss": 1.8208,
+      "step": 1642
+    },
+    {
+      "epoch": 0.43196374026997736,
+      "grad_norm": 0.8496580123901367,
+      "learning_rate": 8.562291922200807e-05,
+      "loss": 1.8206,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4324892436036391,
+      "grad_norm": 0.761249303817749,
+      "learning_rate": 8.560539688102332e-05,
+      "loss": 1.7858,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4330147469373009,
+      "grad_norm": 0.8980756402015686,
+      "learning_rate": 8.558787454003855e-05,
+      "loss": 1.8053,
+      "step": 1648
+    },
+    {
+      "epoch": 0.43354025027096266,
+      "grad_norm": 0.9203025698661804,
+      "learning_rate": 8.55703521990538e-05,
+      "loss": 1.8379,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4340657536046244,
+      "grad_norm": 0.9592378735542297,
+      "learning_rate": 8.555282985806905e-05,
+      "loss": 1.861,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4345912569382862,
+      "grad_norm": 1.0187515020370483,
+      "learning_rate": 8.553530751708428e-05,
+      "loss": 1.7764,
+      "step": 1654
+    },
+    {
+      "epoch": 0.43511676027194796,
+      "grad_norm": 0.8016685247421265,
+      "learning_rate": 8.551778517609953e-05,
+      "loss": 1.8176,
+      "step": 1656
+    },
+    {
+      "epoch": 0.43564226360560976,
+      "grad_norm": 0.7380330562591553,
+      "learning_rate": 8.550026283511477e-05,
+      "loss": 1.8218,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4361677669392715,
+      "grad_norm": 0.6815687417984009,
+      "learning_rate": 8.548274049413002e-05,
+      "loss": 1.8127,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4366932702729333,
+      "grad_norm": 0.6906920075416565,
+      "learning_rate": 8.546521815314527e-05,
+      "loss": 1.773,
+      "step": 1662
+    },
+    {
+      "epoch": 0.43721877360659506,
+      "grad_norm": 0.9510621428489685,
+      "learning_rate": 8.544769581216052e-05,
+      "loss": 1.8305,
+      "step": 1664
+    },
+    {
+      "epoch": 0.43774427694025686,
+      "grad_norm": 0.727105438709259,
+      "learning_rate": 8.543017347117575e-05,
+      "loss": 1.8192,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4382697802739186,
+      "grad_norm": 0.7399454712867737,
+      "learning_rate": 8.5412651130191e-05,
+      "loss": 1.8133,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4387952836075804,
+      "grad_norm": 0.8177588582038879,
+      "learning_rate": 8.539512878920625e-05,
+      "loss": 1.7933,
+      "step": 1670
+    },
+    {
+      "epoch": 0.43932078694124216,
+      "grad_norm": 0.7681954503059387,
+      "learning_rate": 8.537760644822148e-05,
+      "loss": 1.837,
+      "step": 1672
+    },
+    {
+      "epoch": 0.4398462902749039,
+      "grad_norm": 1.5765389204025269,
+      "learning_rate": 8.536008410723673e-05,
+      "loss": 1.8291,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4403717936085657,
+      "grad_norm": 0.7724891304969788,
+      "learning_rate": 8.534256176625198e-05,
+      "loss": 1.8124,
+      "step": 1676
+    },
+    {
+      "epoch": 0.44089729694222746,
+      "grad_norm": 0.8893011808395386,
+      "learning_rate": 8.532503942526722e-05,
+      "loss": 1.8405,
+      "step": 1678
+    },
+    {
+      "epoch": 0.44142280027588926,
+      "grad_norm": 0.878136932849884,
+      "learning_rate": 8.530751708428246e-05,
+      "loss": 1.8106,
+      "step": 1680
+    },
+    {
+      "epoch": 0.441948303609551,
+      "grad_norm": 0.9325633645057678,
+      "learning_rate": 8.52899947432977e-05,
+      "loss": 1.7963,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4424738069432128,
+      "grad_norm": 1.0837180614471436,
+      "learning_rate": 8.527247240231295e-05,
+      "loss": 1.8144,
+      "step": 1684
+    },
+    {
+      "epoch": 0.44299931027687456,
+      "grad_norm": 0.8428369164466858,
+      "learning_rate": 8.52549500613282e-05,
+      "loss": 1.8001,
+      "step": 1686
+    },
+    {
+      "epoch": 0.44352481361053636,
+      "grad_norm": 0.930844783782959,
+      "learning_rate": 8.523742772034345e-05,
+      "loss": 1.8038,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4440503169441981,
+      "grad_norm": 0.7409669756889343,
+      "learning_rate": 8.52199053793587e-05,
+      "loss": 1.7926,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4445758202778599,
+      "grad_norm": 1.0107098817825317,
+      "learning_rate": 8.520238303837393e-05,
+      "loss": 1.7864,
+      "step": 1692
+    },
+    {
+      "epoch": 0.44510132361152166,
+      "grad_norm": 0.760370671749115,
+      "learning_rate": 8.518486069738918e-05,
+      "loss": 1.7881,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4456268269451834,
+      "grad_norm": 1.2437944412231445,
+      "learning_rate": 8.516733835640442e-05,
+      "loss": 1.8152,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4461523302788452,
+      "grad_norm": 0.8944531679153442,
+      "learning_rate": 8.514981601541966e-05,
+      "loss": 1.8272,
+      "step": 1698
+    },
+    {
+      "epoch": 0.44667783361250696,
+      "grad_norm": 0.9550314545631409,
+      "learning_rate": 8.51322936744349e-05,
+      "loss": 1.7936,
+      "step": 1700
+    },
+    {
+      "epoch": 0.44720333694616876,
+      "grad_norm": 0.7015916109085083,
+      "learning_rate": 8.511477133345015e-05,
+      "loss": 1.8737,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4477288402798305,
+      "grad_norm": 1.0813145637512207,
+      "learning_rate": 8.50972489924654e-05,
+      "loss": 1.8118,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4482543436134923,
+      "grad_norm": 0.8479138016700745,
+      "learning_rate": 8.507972665148064e-05,
+      "loss": 1.8074,
+      "step": 1706
+    },
+    {
+      "epoch": 0.44877984694715406,
+      "grad_norm": 0.9790395498275757,
+      "learning_rate": 8.506220431049588e-05,
+      "loss": 1.7785,
+      "step": 1708
+    },
+    {
+      "epoch": 0.44930535028081586,
+      "grad_norm": 0.8719221353530884,
+      "learning_rate": 8.504468196951113e-05,
+      "loss": 1.8162,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4498308536144776,
+      "grad_norm": 1.064282774925232,
+      "learning_rate": 8.502715962852638e-05,
+      "loss": 1.8148,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4503563569481394,
+      "grad_norm": 0.8482780456542969,
+      "learning_rate": 8.500963728754163e-05,
+      "loss": 1.7768,
+      "step": 1714
+    },
+    {
+      "epoch": 0.45088186028180116,
+      "grad_norm": 0.901155412197113,
+      "learning_rate": 8.499211494655687e-05,
+      "loss": 1.8182,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4514073636154629,
+      "grad_norm": 1.0124598741531372,
+      "learning_rate": 8.497459260557211e-05,
+      "loss": 1.8142,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4519328669491247,
+      "grad_norm": 0.8708586692810059,
+      "learning_rate": 8.495707026458735e-05,
+      "loss": 1.8247,
+      "step": 1720
+    },
+    {
+      "epoch": 0.45245837028278646,
+      "grad_norm": 0.9597557187080383,
+      "learning_rate": 8.49395479236026e-05,
+      "loss": 1.7786,
+      "step": 1722
+    },
+    {
+      "epoch": 0.45298387361644826,
+      "grad_norm": 1.1022772789001465,
+      "learning_rate": 8.492202558261784e-05,
+      "loss": 1.8139,
+      "step": 1724
+    },
+    {
+      "epoch": 0.45350937695011,
+      "grad_norm": 1.0891538858413696,
+      "learning_rate": 8.490450324163308e-05,
+      "loss": 1.7986,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4540348802837718,
+      "grad_norm": 0.8127626776695251,
+      "learning_rate": 8.488698090064833e-05,
+      "loss": 1.8026,
+      "step": 1728
+    },
+    {
+      "epoch": 0.45456038361743356,
+      "grad_norm": 0.9313668608665466,
+      "learning_rate": 8.486945855966358e-05,
+      "loss": 1.788,
+      "step": 1730
+    },
+    {
+      "epoch": 0.45508588695109536,
+      "grad_norm": 0.8581985235214233,
+      "learning_rate": 8.485193621867881e-05,
+      "loss": 1.8097,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4556113902847571,
+      "grad_norm": 0.7745251059532166,
+      "learning_rate": 8.483441387769406e-05,
+      "loss": 1.8095,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4561368936184189,
+      "grad_norm": 0.7251246571540833,
+      "learning_rate": 8.481689153670931e-05,
+      "loss": 1.8195,
+      "step": 1736
+    },
+    {
+      "epoch": 0.45666239695208066,
+      "grad_norm": 0.8949863314628601,
+      "learning_rate": 8.479936919572456e-05,
+      "loss": 1.8665,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4571879002857424,
+      "grad_norm": 0.8396829962730408,
+      "learning_rate": 8.47818468547398e-05,
+      "loss": 1.808,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4577134036194042,
+      "grad_norm": 0.8307755589485168,
+      "learning_rate": 8.476432451375505e-05,
+      "loss": 1.8332,
+      "step": 1742
+    },
+    {
+      "epoch": 0.45823890695306596,
+      "grad_norm": 0.8172729015350342,
+      "learning_rate": 8.474680217277028e-05,
+      "loss": 1.7829,
+      "step": 1744
+    },
+    {
+      "epoch": 0.45876441028672776,
+      "grad_norm": 0.7859178185462952,
+      "learning_rate": 8.472927983178553e-05,
+      "loss": 1.8111,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4592899136203895,
+      "grad_norm": 0.8412219882011414,
+      "learning_rate": 8.471175749080078e-05,
+      "loss": 1.8106,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4598154169540513,
+      "grad_norm": 1.1843855381011963,
+      "learning_rate": 8.469423514981601e-05,
+      "loss": 1.8064,
+      "step": 1750
+    },
+    {
+      "epoch": 0.46034092028771306,
+      "grad_norm": 0.7547696232795715,
+      "learning_rate": 8.467671280883126e-05,
+      "loss": 1.7785,
+      "step": 1752
+    },
+    {
+      "epoch": 0.46086642362137487,
+      "grad_norm": 0.8816393613815308,
+      "learning_rate": 8.465919046784651e-05,
+      "loss": 1.7851,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4613919269550366,
+      "grad_norm": 0.985379695892334,
+      "learning_rate": 8.464166812686176e-05,
+      "loss": 1.7841,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4619174302886984,
+      "grad_norm": 0.7585499286651611,
+      "learning_rate": 8.462414578587699e-05,
+      "loss": 1.7928,
+      "step": 1758
+    },
+    {
+      "epoch": 0.46244293362236016,
+      "grad_norm": 0.8072088956832886,
+      "learning_rate": 8.460662344489224e-05,
+      "loss": 1.7867,
+      "step": 1760
+    },
+    {
+      "epoch": 0.46296843695602197,
+      "grad_norm": 0.8818538784980774,
+      "learning_rate": 8.458910110390749e-05,
+      "loss": 1.8085,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4634939402896837,
+      "grad_norm": 0.7545977830886841,
+      "learning_rate": 8.457157876292273e-05,
+      "loss": 1.8134,
+      "step": 1764
+    },
+    {
+      "epoch": 0.46401944362334546,
+      "grad_norm": 0.7525529265403748,
+      "learning_rate": 8.455405642193798e-05,
+      "loss": 1.8235,
+      "step": 1766
+    },
+    {
+      "epoch": 0.46454494695700727,
+      "grad_norm": 1.0930671691894531,
+      "learning_rate": 8.453653408095323e-05,
+      "loss": 1.7796,
+      "step": 1768
+    },
+    {
+      "epoch": 0.465070450290669,
+      "grad_norm": 0.8903842568397522,
+      "learning_rate": 8.451901173996846e-05,
+      "loss": 1.8413,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4655959536243308,
+      "grad_norm": 0.9519714117050171,
+      "learning_rate": 8.450148939898371e-05,
+      "loss": 1.8189,
+      "step": 1772
+    },
+    {
+      "epoch": 0.46612145695799256,
+      "grad_norm": 0.816856861114502,
+      "learning_rate": 8.448396705799894e-05,
+      "loss": 1.8338,
+      "step": 1774
+    },
+    {
+      "epoch": 0.46664696029165437,
+      "grad_norm": 0.8597956299781799,
+      "learning_rate": 8.446644471701419e-05,
+      "loss": 1.7963,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4671724636253161,
+      "grad_norm": 0.8783820271492004,
+      "learning_rate": 8.444892237602944e-05,
+      "loss": 1.7985,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4676979669589779,
+      "grad_norm": 0.9130274653434753,
+      "learning_rate": 8.443140003504469e-05,
+      "loss": 1.791,
+      "step": 1780
+    },
+    {
+      "epoch": 0.46822347029263967,
+      "grad_norm": 1.001056432723999,
+      "learning_rate": 8.441387769405993e-05,
+      "loss": 1.8115,
+      "step": 1782
+    },
+    {
+      "epoch": 0.46874897362630147,
+      "grad_norm": 0.8054397702217102,
+      "learning_rate": 8.439635535307517e-05,
+      "loss": 1.8411,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4692744769599632,
+      "grad_norm": 0.8605888485908508,
+      "learning_rate": 8.437883301209042e-05,
+      "loss": 1.7763,
+      "step": 1786
+    },
+    {
+      "epoch": 0.46979998029362496,
+      "grad_norm": 1.1046404838562012,
+      "learning_rate": 8.436131067110566e-05,
+      "loss": 1.8331,
+      "step": 1788
+    },
+    {
+      "epoch": 0.47032548362728677,
+      "grad_norm": 0.8004000186920166,
+      "learning_rate": 8.434378833012091e-05,
+      "loss": 1.826,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4708509869609485,
+      "grad_norm": 1.05780827999115,
+      "learning_rate": 8.432626598913616e-05,
+      "loss": 1.8101,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4713764902946103,
+      "grad_norm": 0.9407793879508972,
+      "learning_rate": 8.43087436481514e-05,
+      "loss": 1.7962,
+      "step": 1794
+    },
+    {
+      "epoch": 0.47190199362827207,
+      "grad_norm": 0.7705127000808716,
+      "learning_rate": 8.429122130716664e-05,
+      "loss": 1.7978,
+      "step": 1796
+    },
+    {
+      "epoch": 0.47242749696193387,
+      "grad_norm": 0.8127232193946838,
+      "learning_rate": 8.427369896618189e-05,
+      "loss": 1.7736,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4729530002955956,
+      "grad_norm": 1.0332077741622925,
+      "learning_rate": 8.425617662519712e-05,
+      "loss": 1.7882,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4734785036292574,
+      "grad_norm": 0.8787586092948914,
+      "learning_rate": 8.423865428421237e-05,
+      "loss": 1.7942,
+      "step": 1802
+    },
+    {
+      "epoch": 0.47400400696291917,
+      "grad_norm": 1.0014612674713135,
+      "learning_rate": 8.422113194322762e-05,
+      "loss": 1.8046,
+      "step": 1804
+    },
+    {
+      "epoch": 0.47452951029658097,
+      "grad_norm": 0.7657825946807861,
+      "learning_rate": 8.420360960224286e-05,
+      "loss": 1.7933,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4750550136302427,
+      "grad_norm": 0.8316423296928406,
+      "learning_rate": 8.418608726125811e-05,
+      "loss": 1.7988,
+      "step": 1808
+    },
+    {
+      "epoch": 0.47558051696390446,
+      "grad_norm": 0.7967455387115479,
+      "learning_rate": 8.416856492027335e-05,
+      "loss": 1.8043,
+      "step": 1810
+    },
+    {
+      "epoch": 0.47610602029756627,
+      "grad_norm": 0.7839574217796326,
+      "learning_rate": 8.41510425792886e-05,
+      "loss": 1.8401,
+      "step": 1812
+    },
+    {
+      "epoch": 0.476631523631228,
+      "grad_norm": 0.8147667646408081,
+      "learning_rate": 8.413352023830384e-05,
+      "loss": 1.8019,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4771570269648898,
+      "grad_norm": 1.0701135396957397,
+      "learning_rate": 8.411599789731909e-05,
+      "loss": 1.8301,
+      "step": 1816
+    },
+    {
+      "epoch": 0.47768253029855157,
+      "grad_norm": 0.8631109595298767,
+      "learning_rate": 8.409847555633434e-05,
+      "loss": 1.8335,
+      "step": 1818
+    },
+    {
+      "epoch": 0.47820803363221337,
+      "grad_norm": 0.9797492623329163,
+      "learning_rate": 8.408095321534958e-05,
+      "loss": 1.8189,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4787335369658751,
+      "grad_norm": 0.9087586998939514,
+      "learning_rate": 8.406343087436482e-05,
+      "loss": 1.8405,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4792590402995369,
+      "grad_norm": 0.7700434923171997,
+      "learning_rate": 8.404590853338007e-05,
+      "loss": 1.7943,
+      "step": 1824
+    },
+    {
+      "epoch": 0.47978454363319867,
+      "grad_norm": 0.9100522398948669,
+      "learning_rate": 8.40283861923953e-05,
+      "loss": 1.8021,
+      "step": 1826
+    },
+    {
+      "epoch": 0.48031004696686047,
+      "grad_norm": 1.0668331384658813,
+      "learning_rate": 8.401086385141055e-05,
+      "loss": 1.7966,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4808355503005222,
+      "grad_norm": 0.9680224061012268,
+      "learning_rate": 8.39933415104258e-05,
+      "loss": 1.7945,
+      "step": 1830
+    },
+    {
+      "epoch": 0.48136105363418397,
+      "grad_norm": 1.0217275619506836,
+      "learning_rate": 8.397581916944104e-05,
+      "loss": 1.821,
+      "step": 1832
+    },
+    {
+      "epoch": 0.48188655696784577,
+      "grad_norm": 0.854264497756958,
+      "learning_rate": 8.395829682845629e-05,
+      "loss": 1.809,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4824120603015075,
+      "grad_norm": 0.9226179718971252,
+      "learning_rate": 8.394077448747152e-05,
+      "loss": 1.8021,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4829375636351693,
+      "grad_norm": 1.205917239189148,
+      "learning_rate": 8.392325214648677e-05,
+      "loss": 1.7912,
+      "step": 1838
+    },
+    {
+      "epoch": 0.48346306696883107,
+      "grad_norm": 1.0180691480636597,
+      "learning_rate": 8.390572980550202e-05,
+      "loss": 1.8032,
+      "step": 1840
+    },
+    {
+      "epoch": 0.48398857030249287,
+      "grad_norm": 0.9707418084144592,
+      "learning_rate": 8.388820746451727e-05,
+      "loss": 1.8038,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4845140736361546,
+      "grad_norm": 1.21326744556427,
+      "learning_rate": 8.387068512353251e-05,
+      "loss": 1.8264,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4850395769698164,
+      "grad_norm": 0.8171116709709167,
+      "learning_rate": 8.385316278254776e-05,
+      "loss": 1.8037,
+      "step": 1846
+    },
+    {
+      "epoch": 0.48556508030347817,
+      "grad_norm": 0.9211159348487854,
+      "learning_rate": 8.3835640441563e-05,
+      "loss": 1.8085,
+      "step": 1848
+    },
+    {
+      "epoch": 0.48609058363714,
+      "grad_norm": 1.2191238403320312,
+      "learning_rate": 8.381811810057824e-05,
+      "loss": 1.7982,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4866160869708017,
+      "grad_norm": 0.8354179263114929,
+      "learning_rate": 8.380059575959348e-05,
+      "loss": 1.7862,
+      "step": 1852
+    },
+    {
+      "epoch": 0.48714159030446347,
+      "grad_norm": 0.7723087072372437,
+      "learning_rate": 8.378307341860872e-05,
+      "loss": 1.7629,
+      "step": 1854
+    },
+    {
+      "epoch": 0.48766709363812527,
+      "grad_norm": 1.1398297548294067,
+      "learning_rate": 8.376555107762397e-05,
+      "loss": 1.8118,
+      "step": 1856
+    },
+    {
+      "epoch": 0.488192596971787,
+      "grad_norm": 0.904444694519043,
+      "learning_rate": 8.374802873663922e-05,
+      "loss": 1.8417,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4887181003054488,
+      "grad_norm": 1.8874095678329468,
+      "learning_rate": 8.373050639565447e-05,
+      "loss": 1.7985,
+      "step": 1860
+    },
+    {
+      "epoch": 0.48924360363911057,
+      "grad_norm": 1.9317784309387207,
+      "learning_rate": 8.37129840546697e-05,
+      "loss": 1.7974,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4897691069727724,
+      "grad_norm": 0.9307408332824707,
+      "learning_rate": 8.369546171368495e-05,
+      "loss": 1.8137,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4902946103064341,
+      "grad_norm": 0.7302669882774353,
+      "learning_rate": 8.36779393727002e-05,
+      "loss": 1.7555,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4908201136400959,
+      "grad_norm": 0.9116623997688293,
+      "learning_rate": 8.366041703171544e-05,
+      "loss": 1.8122,
+      "step": 1868
+    },
+    {
+      "epoch": 0.49134561697375767,
+      "grad_norm": 0.8545958995819092,
+      "learning_rate": 8.364289469073069e-05,
+      "loss": 1.8192,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4918711203074195,
+      "grad_norm": 0.8263342380523682,
+      "learning_rate": 8.362537234974594e-05,
+      "loss": 1.8059,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4923966236410812,
+      "grad_norm": 1.1116405725479126,
+      "learning_rate": 8.360785000876117e-05,
+      "loss": 1.819,
+      "step": 1874
+    },
+    {
+      "epoch": 0.49292212697474297,
+      "grad_norm": 0.8676914572715759,
+      "learning_rate": 8.359032766777641e-05,
+      "loss": 1.801,
+      "step": 1876
+    },
+    {
+      "epoch": 0.49344763030840477,
+      "grad_norm": 0.7704113125801086,
+      "learning_rate": 8.357280532679165e-05,
+      "loss": 1.7834,
+      "step": 1878
+    },
+    {
+      "epoch": 0.4939731336420665,
+      "grad_norm": 1.1784600019454956,
+      "learning_rate": 8.35552829858069e-05,
+      "loss": 1.8283,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4944986369757283,
+      "grad_norm": 1.0245088338851929,
+      "learning_rate": 8.353776064482215e-05,
+      "loss": 1.8172,
+      "step": 1882
+    },
+    {
+      "epoch": 0.49502414030939007,
+      "grad_norm": 0.9373153448104858,
+      "learning_rate": 8.35202383038374e-05,
+      "loss": 1.786,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4955496436430519,
+      "grad_norm": 1.362306833267212,
+      "learning_rate": 8.350271596285264e-05,
+      "loss": 1.8157,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4960751469767136,
+      "grad_norm": 0.9665769934654236,
+      "learning_rate": 8.348519362186788e-05,
+      "loss": 1.7878,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4966006503103754,
+      "grad_norm": 0.8806145191192627,
+      "learning_rate": 8.346767128088313e-05,
+      "loss": 1.8323,
+      "step": 1890
+    },
+    {
+      "epoch": 0.49712615364403717,
+      "grad_norm": 1.2314025163650513,
+      "learning_rate": 8.345014893989837e-05,
+      "loss": 1.8569,
+      "step": 1892
+    },
+    {
+      "epoch": 0.497651656977699,
+      "grad_norm": 1.8855247497558594,
+      "learning_rate": 8.343262659891362e-05,
+      "loss": 1.8152,
+      "step": 1894
+    },
+    {
+      "epoch": 0.4981771603113607,
+      "grad_norm": 1.1224102973937988,
+      "learning_rate": 8.341510425792887e-05,
+      "loss": 1.8022,
+      "step": 1896
+    },
+    {
+      "epoch": 0.49870266364502247,
+      "grad_norm": 0.7415306568145752,
+      "learning_rate": 8.339758191694412e-05,
+      "loss": 1.7825,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4992281669786843,
+      "grad_norm": 1.4165035486221313,
+      "learning_rate": 8.338005957595935e-05,
+      "loss": 1.8148,
+      "step": 1900
+    },
+    {
+      "epoch": 0.499753670312346,
+      "grad_norm": 0.7428869605064392,
+      "learning_rate": 8.33625372349746e-05,
+      "loss": 1.8034,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5002791736460078,
+      "grad_norm": 0.6850984692573547,
+      "learning_rate": 8.334501489398983e-05,
+      "loss": 1.7891,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5008046769796696,
+      "grad_norm": 0.7525566816329956,
+      "learning_rate": 8.332749255300508e-05,
+      "loss": 1.8046,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5013301803133313,
+      "grad_norm": 0.8629323840141296,
+      "learning_rate": 8.330997021202033e-05,
+      "loss": 1.82,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5018556836469932,
+      "grad_norm": 0.885529637336731,
+      "learning_rate": 8.329244787103557e-05,
+      "loss": 1.7576,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5023811869806549,
+      "grad_norm": 0.7622796893119812,
+      "learning_rate": 8.327492553005082e-05,
+      "loss": 1.8287,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5029066903143167,
+      "grad_norm": 0.8797925710678101,
+      "learning_rate": 8.325740318906607e-05,
+      "loss": 1.8291,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5034321936479784,
+      "grad_norm": 0.8444175124168396,
+      "learning_rate": 8.32398808480813e-05,
+      "loss": 1.8036,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5039576969816403,
+      "grad_norm": 0.9204092025756836,
+      "learning_rate": 8.322235850709655e-05,
+      "loss": 1.8346,
+      "step": 1918
+    },
+    {
+      "epoch": 0.504483200315302,
+      "grad_norm": 0.7571083307266235,
+      "learning_rate": 8.32048361661118e-05,
+      "loss": 1.7909,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5050087036489638,
+      "grad_norm": 0.861875593662262,
+      "learning_rate": 8.318731382512705e-05,
+      "loss": 1.8077,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5055342069826255,
+      "grad_norm": 0.8385462164878845,
+      "learning_rate": 8.31697914841423e-05,
+      "loss": 1.7738,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5060597103162873,
+      "grad_norm": 0.9641711115837097,
+      "learning_rate": 8.315226914315754e-05,
+      "loss": 1.7992,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5065852136499491,
+      "grad_norm": 1.5050584077835083,
+      "learning_rate": 8.313474680217278e-05,
+      "loss": 1.7867,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5071107169836109,
+      "grad_norm": 1.0225639343261719,
+      "learning_rate": 8.311722446118801e-05,
+      "loss": 1.7787,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5076362203172726,
+      "grad_norm": 0.9370511174201965,
+      "learning_rate": 8.309970212020326e-05,
+      "loss": 1.7945,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5081617236509344,
+      "grad_norm": 1.1602392196655273,
+      "learning_rate": 8.30821797792185e-05,
+      "loss": 1.8071,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5086872269845962,
+      "grad_norm": 1.095885157585144,
+      "learning_rate": 8.306465743823375e-05,
+      "loss": 1.8048,
+      "step": 1936
+    },
+    {
+      "epoch": 0.509212730318258,
+      "grad_norm": 1.124812364578247,
+      "learning_rate": 8.3047135097249e-05,
+      "loss": 1.8092,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5097382336519197,
+      "grad_norm": 0.7736151218414307,
+      "learning_rate": 8.302961275626425e-05,
+      "loss": 1.8173,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5102637369855815,
+      "grad_norm": 1.199781894683838,
+      "learning_rate": 8.301209041527948e-05,
+      "loss": 1.7725,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5107892403192432,
+      "grad_norm": 0.772127091884613,
+      "learning_rate": 8.299456807429473e-05,
+      "loss": 1.8158,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5113147436529051,
+      "grad_norm": 0.8498915433883667,
+      "learning_rate": 8.297704573330998e-05,
+      "loss": 1.7641,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5118402469865668,
+      "grad_norm": 1.0101765394210815,
+      "learning_rate": 8.295952339232522e-05,
+      "loss": 1.8214,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5123657503202286,
+      "grad_norm": 0.8255197405815125,
+      "learning_rate": 8.294200105134047e-05,
+      "loss": 1.7684,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5128912536538903,
+      "grad_norm": 0.884064257144928,
+      "learning_rate": 8.292447871035572e-05,
+      "loss": 1.7953,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5134167569875522,
+      "grad_norm": 1.1106665134429932,
+      "learning_rate": 8.290695636937095e-05,
+      "loss": 1.8086,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5139422603212139,
+      "grad_norm": 0.8464235663414001,
+      "learning_rate": 8.288943402838619e-05,
+      "loss": 1.8182,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5144677636548757,
+      "grad_norm": 0.8477456569671631,
+      "learning_rate": 8.287191168740144e-05,
+      "loss": 1.7845,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5149932669885374,
+      "grad_norm": 0.8464748859405518,
+      "learning_rate": 8.285438934641668e-05,
+      "loss": 1.7325,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5155187703221993,
+      "grad_norm": 0.8044800758361816,
+      "learning_rate": 8.283686700543193e-05,
+      "loss": 1.7715,
+      "step": 1962
+    },
+    {
+      "epoch": 0.516044273655861,
+      "grad_norm": 0.8983359336853027,
+      "learning_rate": 8.281934466444718e-05,
+      "loss": 1.8041,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5165697769895228,
+      "grad_norm": 1.2751051187515259,
+      "learning_rate": 8.280182232346243e-05,
+      "loss": 1.8038,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5170952803231845,
+      "grad_norm": 0.9036842584609985,
+      "learning_rate": 8.278429998247766e-05,
+      "loss": 1.7891,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5176207836568463,
+      "grad_norm": 0.8021928668022156,
+      "learning_rate": 8.276677764149291e-05,
+      "loss": 1.8139,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5181462869905081,
+      "grad_norm": 0.8150444030761719,
+      "learning_rate": 8.274925530050815e-05,
+      "loss": 1.7707,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5186717903241699,
+      "grad_norm": 0.7655881643295288,
+      "learning_rate": 8.27317329595234e-05,
+      "loss": 1.822,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5191972936578316,
+      "grad_norm": 1.1329301595687866,
+      "learning_rate": 8.271421061853865e-05,
+      "loss": 1.7675,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5197227969914934,
+      "grad_norm": 1.0726947784423828,
+      "learning_rate": 8.269668827755388e-05,
+      "loss": 1.8203,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5202483003251552,
+      "grad_norm": 0.7917013168334961,
+      "learning_rate": 8.267916593656913e-05,
+      "loss": 1.8326,
+      "step": 1980
+    },
+    {
+      "epoch": 0.520773803658817,
+      "grad_norm": 0.955825686454773,
+      "learning_rate": 8.266164359558437e-05,
+      "loss": 1.7993,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5212993069924787,
+      "grad_norm": 1.0151026248931885,
+      "learning_rate": 8.264412125459961e-05,
+      "loss": 1.7764,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5218248103261405,
+      "grad_norm": 1.11459219455719,
+      "learning_rate": 8.262659891361486e-05,
+      "loss": 1.8025,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5223503136598023,
+      "grad_norm": 0.9148701429367065,
+      "learning_rate": 8.260907657263011e-05,
+      "loss": 1.8218,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5228758169934641,
+      "grad_norm": 0.7594835758209229,
+      "learning_rate": 8.259155423164536e-05,
+      "loss": 1.8266,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5234013203271258,
+      "grad_norm": 1.0429881811141968,
+      "learning_rate": 8.25740318906606e-05,
+      "loss": 1.8223,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5239268236607876,
+      "grad_norm": 0.9307808876037598,
+      "learning_rate": 8.255650954967584e-05,
+      "loss": 1.8322,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5244523269944493,
+      "grad_norm": 0.7135612964630127,
+      "learning_rate": 8.253898720869108e-05,
+      "loss": 1.8104,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5249778303281112,
+      "grad_norm": 0.8590657711029053,
+      "learning_rate": 8.252146486770633e-05,
+      "loss": 1.7808,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5255033336617729,
+      "grad_norm": 0.9131346940994263,
+      "learning_rate": 8.250394252672158e-05,
+      "loss": 1.7846,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5255033336617729,
+      "eval_loss": 1.7627384662628174,
+      "eval_runtime": 487.1945,
+      "eval_samples_per_second": 249.98,
+      "eval_steps_per_second": 31.248,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5260288369954347,
+      "grad_norm": 0.9146105647087097,
+      "learning_rate": 8.248642018573683e-05,
+      "loss": 1.8062,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5265543403290964,
+      "grad_norm": 0.8270906209945679,
+      "learning_rate": 8.246889784475206e-05,
+      "loss": 1.7992,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5270798436627583,
+      "grad_norm": 0.7374973297119141,
+      "learning_rate": 8.245137550376731e-05,
+      "loss": 1.8103,
+      "step": 2006
+    },
+    {
+      "epoch": 0.52760534699642,
+      "grad_norm": 0.748988926410675,
+      "learning_rate": 8.243385316278254e-05,
+      "loss": 1.8284,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5281308503300818,
+      "grad_norm": 0.7624616622924805,
+      "learning_rate": 8.241633082179779e-05,
+      "loss": 1.7947,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5286563536637435,
+      "grad_norm": 0.8599483966827393,
+      "learning_rate": 8.239880848081304e-05,
+      "loss": 1.7732,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5291818569974053,
+      "grad_norm": 0.8122212886810303,
+      "learning_rate": 8.238128613982829e-05,
+      "loss": 1.8128,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5297073603310671,
+      "grad_norm": 0.8008537292480469,
+      "learning_rate": 8.236376379884353e-05,
+      "loss": 1.8028,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5302328636647289,
+      "grad_norm": 0.8155772686004639,
+      "learning_rate": 8.234624145785878e-05,
+      "loss": 1.8124,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5307583669983906,
+      "grad_norm": 0.9014889001846313,
+      "learning_rate": 8.232871911687401e-05,
+      "loss": 1.8056,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5312838703320524,
+      "grad_norm": 0.7495489716529846,
+      "learning_rate": 8.231119677588926e-05,
+      "loss": 1.7891,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5318093736657142,
+      "grad_norm": 0.6256895065307617,
+      "learning_rate": 8.229367443490451e-05,
+      "loss": 1.7671,
+      "step": 2024
+    },
+    {
+      "epoch": 0.532334876999376,
+      "grad_norm": 0.9852064847946167,
+      "learning_rate": 8.227615209391976e-05,
+      "loss": 1.7945,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5328603803330377,
+      "grad_norm": 0.9755546450614929,
+      "learning_rate": 8.2258629752935e-05,
+      "loss": 1.7999,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5333858836666995,
+      "grad_norm": 0.7656162977218628,
+      "learning_rate": 8.224110741195024e-05,
+      "loss": 1.7869,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5339113870003613,
+      "grad_norm": 0.9118050336837769,
+      "learning_rate": 8.222358507096549e-05,
+      "loss": 1.8278,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5344368903340231,
+      "grad_norm": 0.6935544013977051,
+      "learning_rate": 8.220606272998072e-05,
+      "loss": 1.8328,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5349623936676848,
+      "grad_norm": 0.9424465298652649,
+      "learning_rate": 8.218854038899597e-05,
+      "loss": 1.7714,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5354878970013466,
+      "grad_norm": 0.7186869978904724,
+      "learning_rate": 8.217101804801122e-05,
+      "loss": 1.8039,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5360134003350083,
+      "grad_norm": 0.6855552792549133,
+      "learning_rate": 8.215349570702646e-05,
+      "loss": 1.7952,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5365389036686702,
+      "grad_norm": 0.6616165637969971,
+      "learning_rate": 8.213597336604171e-05,
+      "loss": 1.7949,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5370644070023319,
+      "grad_norm": 0.7983983755111694,
+      "learning_rate": 8.211845102505696e-05,
+      "loss": 1.8027,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5375899103359937,
+      "grad_norm": 0.7051743865013123,
+      "learning_rate": 8.210092868407219e-05,
+      "loss": 1.8062,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5381154136696554,
+      "grad_norm": 0.7773457169532776,
+      "learning_rate": 8.208340634308744e-05,
+      "loss": 1.7559,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5386409170033173,
+      "grad_norm": 1.415022373199463,
+      "learning_rate": 8.206588400210269e-05,
+      "loss": 1.8263,
+      "step": 2050
+    },
+    {
+      "epoch": 0.539166420336979,
+      "grad_norm": 0.8128517270088196,
+      "learning_rate": 8.204836166111794e-05,
+      "loss": 1.8131,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5396919236706408,
+      "grad_norm": 0.8620209097862244,
+      "learning_rate": 8.203083932013317e-05,
+      "loss": 1.7608,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5402174270043025,
+      "grad_norm": 0.7338502407073975,
+      "learning_rate": 8.201331697914842e-05,
+      "loss": 1.8006,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5407429303379643,
+      "grad_norm": 0.8432105183601379,
+      "learning_rate": 8.199579463816366e-05,
+      "loss": 1.7716,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5412684336716261,
+      "grad_norm": 0.600993812084198,
+      "learning_rate": 8.19782722971789e-05,
+      "loss": 1.7858,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5417939370052879,
+      "grad_norm": 0.7413330674171448,
+      "learning_rate": 8.196074995619415e-05,
+      "loss": 1.7714,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5423194403389496,
+      "grad_norm": 0.7860500812530518,
+      "learning_rate": 8.19432276152094e-05,
+      "loss": 1.8065,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5428449436726114,
+      "grad_norm": 0.8075912594795227,
+      "learning_rate": 8.192570527422464e-05,
+      "loss": 1.7946,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5433704470062732,
+      "grad_norm": 0.7681949734687805,
+      "learning_rate": 8.190818293323989e-05,
+      "loss": 1.8246,
+      "step": 2068
+    },
+    {
+      "epoch": 0.543895950339935,
+      "grad_norm": 0.7330343127250671,
+      "learning_rate": 8.189066059225514e-05,
+      "loss": 1.7884,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5444214536735967,
+      "grad_norm": 0.8136972784996033,
+      "learning_rate": 8.187313825127037e-05,
+      "loss": 1.7783,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5449469570072585,
+      "grad_norm": 0.9508219361305237,
+      "learning_rate": 8.185561591028562e-05,
+      "loss": 1.822,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5454724603409203,
+      "grad_norm": 1.0187771320343018,
+      "learning_rate": 8.183809356930087e-05,
+      "loss": 1.7858,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5459979636745821,
+      "grad_norm": 0.872322678565979,
+      "learning_rate": 8.182057122831611e-05,
+      "loss": 1.7715,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5465234670082438,
+      "grad_norm": 0.9391134977340698,
+      "learning_rate": 8.180304888733135e-05,
+      "loss": 1.8014,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5470489703419056,
+      "grad_norm": 0.7798128128051758,
+      "learning_rate": 8.17855265463466e-05,
+      "loss": 1.7803,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5475744736755673,
+      "grad_norm": 0.9748620390892029,
+      "learning_rate": 8.176800420536184e-05,
+      "loss": 1.8054,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5480999770092292,
+      "grad_norm": 0.9456170201301575,
+      "learning_rate": 8.175048186437708e-05,
+      "loss": 1.8084,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5486254803428909,
+      "grad_norm": 1.295296549797058,
+      "learning_rate": 8.173295952339232e-05,
+      "loss": 1.8044,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5491509836765527,
+      "grad_norm": 0.7323461174964905,
+      "learning_rate": 8.171543718240757e-05,
+      "loss": 1.787,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5496764870102144,
+      "grad_norm": 1.0989707708358765,
+      "learning_rate": 8.169791484142282e-05,
+      "loss": 1.7789,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5502019903438763,
+      "grad_norm": 0.9566003680229187,
+      "learning_rate": 8.168039250043807e-05,
+      "loss": 1.7832,
+      "step": 2094
+    },
+    {
+      "epoch": 0.550727493677538,
+      "grad_norm": 0.8293377757072449,
+      "learning_rate": 8.166287015945331e-05,
+      "loss": 1.8081,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5512529970111998,
+      "grad_norm": 0.9931288361549377,
+      "learning_rate": 8.164534781846855e-05,
+      "loss": 1.81,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5517785003448615,
+      "grad_norm": 0.7140156626701355,
+      "learning_rate": 8.16278254774838e-05,
+      "loss": 1.7824,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5523040036785233,
+      "grad_norm": 0.8644457459449768,
+      "learning_rate": 8.161030313649904e-05,
+      "loss": 1.8058,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5528295070121851,
+      "grad_norm": 0.9265533685684204,
+      "learning_rate": 8.159278079551429e-05,
+      "loss": 1.7759,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5533550103458469,
+      "grad_norm": 0.7051352262496948,
+      "learning_rate": 8.157525845452952e-05,
+      "loss": 1.7986,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5538805136795086,
+      "grad_norm": 0.8235836625099182,
+      "learning_rate": 8.155773611354477e-05,
+      "loss": 1.8038,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5544060170131704,
+      "grad_norm": 0.7595791816711426,
+      "learning_rate": 8.154021377256002e-05,
+      "loss": 1.8321,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5549315203468322,
+      "grad_norm": 0.7582104802131653,
+      "learning_rate": 8.152269143157525e-05,
+      "loss": 1.8297,
+      "step": 2112
+    },
+    {
+      "epoch": 0.555457023680494,
+      "grad_norm": 0.6678735613822937,
+      "learning_rate": 8.15051690905905e-05,
+      "loss": 1.7559,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5559825270141557,
+      "grad_norm": 0.8775631189346313,
+      "learning_rate": 8.148764674960575e-05,
+      "loss": 1.8064,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5565080303478175,
+      "grad_norm": 0.8533100485801697,
+      "learning_rate": 8.1470124408621e-05,
+      "loss": 1.7946,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5570335336814793,
+      "grad_norm": 0.6834318041801453,
+      "learning_rate": 8.145260206763624e-05,
+      "loss": 1.7833,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5575590370151411,
+      "grad_norm": 0.9778022766113281,
+      "learning_rate": 8.143507972665149e-05,
+      "loss": 1.8012,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5580845403488028,
+      "grad_norm": 0.7117162942886353,
+      "learning_rate": 8.141755738566673e-05,
+      "loss": 1.8246,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5586100436824646,
+      "grad_norm": 0.8154255747795105,
+      "learning_rate": 8.140003504468197e-05,
+      "loss": 1.7692,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5591355470161263,
+      "grad_norm": 0.7914754152297974,
+      "learning_rate": 8.138251270369722e-05,
+      "loss": 1.7623,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5596610503497882,
+      "grad_norm": 0.657900333404541,
+      "learning_rate": 8.136499036271247e-05,
+      "loss": 1.7882,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5601865536834499,
+      "grad_norm": 0.6770361065864563,
+      "learning_rate": 8.13474680217277e-05,
+      "loss": 1.7737,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5607120570171117,
+      "grad_norm": 0.9176309108734131,
+      "learning_rate": 8.132994568074295e-05,
+      "loss": 1.7706,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5612375603507734,
+      "grad_norm": 1.037473201751709,
+      "learning_rate": 8.13124233397582e-05,
+      "loss": 1.8172,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5617630636844353,
+      "grad_norm": 1.409900426864624,
+      "learning_rate": 8.129490099877343e-05,
+      "loss": 1.7831,
+      "step": 2138
+    },
+    {
+      "epoch": 0.562288567018097,
+      "grad_norm": 0.7014243602752686,
+      "learning_rate": 8.127737865778868e-05,
+      "loss": 1.7844,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5628140703517588,
+      "grad_norm": 0.7743593454360962,
+      "learning_rate": 8.125985631680393e-05,
+      "loss": 1.7591,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5633395736854205,
+      "grad_norm": 0.8568558096885681,
+      "learning_rate": 8.124233397581917e-05,
+      "loss": 1.8341,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5638650770190824,
+      "grad_norm": 0.792972981929779,
+      "learning_rate": 8.122481163483442e-05,
+      "loss": 1.7639,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5643905803527441,
+      "grad_norm": 0.983025312423706,
+      "learning_rate": 8.120728929384967e-05,
+      "loss": 1.7952,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5649160836864059,
+      "grad_norm": 0.7422555088996887,
+      "learning_rate": 8.11897669528649e-05,
+      "loss": 1.7856,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5654415870200676,
+      "grad_norm": 0.6807308793067932,
+      "learning_rate": 8.117224461188015e-05,
+      "loss": 1.841,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5659670903537294,
+      "grad_norm": 0.845778226852417,
+      "learning_rate": 8.11547222708954e-05,
+      "loss": 1.7815,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5664925936873912,
+      "grad_norm": 0.8135868906974792,
+      "learning_rate": 8.113719992991063e-05,
+      "loss": 1.8014,
+      "step": 2156
+    },
+    {
+      "epoch": 0.567018097021053,
+      "grad_norm": 0.7737998366355896,
+      "learning_rate": 8.111967758892588e-05,
+      "loss": 1.7906,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5675436003547147,
+      "grad_norm": 0.8078686594963074,
+      "learning_rate": 8.110215524794113e-05,
+      "loss": 1.7612,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5680691036883765,
+      "grad_norm": 0.8356254696846008,
+      "learning_rate": 8.108463290695637e-05,
+      "loss": 1.7736,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5685946070220383,
+      "grad_norm": 0.7324886322021484,
+      "learning_rate": 8.106711056597162e-05,
+      "loss": 1.8031,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5691201103557001,
+      "grad_norm": 0.6843339800834656,
+      "learning_rate": 8.104958822498686e-05,
+      "loss": 1.7733,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5696456136893618,
+      "grad_norm": 1.1155132055282593,
+      "learning_rate": 8.10320658840021e-05,
+      "loss": 1.8057,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5701711170230236,
+      "grad_norm": 0.944843053817749,
+      "learning_rate": 8.101454354301735e-05,
+      "loss": 1.7848,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5706966203566853,
+      "grad_norm": 0.7923617959022522,
+      "learning_rate": 8.09970212020326e-05,
+      "loss": 1.7636,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5712221236903472,
+      "grad_norm": 0.7884588837623596,
+      "learning_rate": 8.097949886104785e-05,
+      "loss": 1.7856,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5717476270240089,
+      "grad_norm": 0.8496378064155579,
+      "learning_rate": 8.09619765200631e-05,
+      "loss": 1.8124,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5722731303576707,
+      "grad_norm": 0.6795907020568848,
+      "learning_rate": 8.094445417907833e-05,
+      "loss": 1.825,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5727986336913324,
+      "grad_norm": 1.0210143327713013,
+      "learning_rate": 8.092693183809358e-05,
+      "loss": 1.8193,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5733241370249943,
+      "grad_norm": 0.9265308380126953,
+      "learning_rate": 8.090940949710881e-05,
+      "loss": 1.7955,
+      "step": 2182
+    },
+    {
+      "epoch": 0.573849640358656,
+      "grad_norm": 0.7709974646568298,
+      "learning_rate": 8.089188715612406e-05,
+      "loss": 1.7748,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5743751436923178,
+      "grad_norm": 1.015137791633606,
+      "learning_rate": 8.08743648151393e-05,
+      "loss": 1.7767,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5749006470259795,
+      "grad_norm": 0.7084217071533203,
+      "learning_rate": 8.085684247415455e-05,
+      "loss": 1.7737,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5754261503596414,
+      "grad_norm": 0.7948693633079529,
+      "learning_rate": 8.08393201331698e-05,
+      "loss": 1.8202,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5759516536933031,
+      "grad_norm": 0.921947181224823,
+      "learning_rate": 8.082179779218503e-05,
+      "loss": 1.7906,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5764771570269649,
+      "grad_norm": 1.3730195760726929,
+      "learning_rate": 8.080427545120028e-05,
+      "loss": 1.7645,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5770026603606266,
+      "grad_norm": 0.7949815392494202,
+      "learning_rate": 8.078675311021553e-05,
+      "loss": 1.7519,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5775281636942884,
+      "grad_norm": 0.8247926831245422,
+      "learning_rate": 8.076923076923078e-05,
+      "loss": 1.7875,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5780536670279502,
+      "grad_norm": 0.8016488552093506,
+      "learning_rate": 8.075170842824602e-05,
+      "loss": 1.7832,
+      "step": 2200
+    },
+    {
+      "epoch": 0.578579170361612,
+      "grad_norm": 0.8356485366821289,
+      "learning_rate": 8.073418608726127e-05,
+      "loss": 1.7964,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5791046736952737,
+      "grad_norm": 0.6765563488006592,
+      "learning_rate": 8.07166637462765e-05,
+      "loss": 1.7864,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5796301770289355,
+      "grad_norm": 0.6917481422424316,
+      "learning_rate": 8.069914140529175e-05,
+      "loss": 1.7662,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5801556803625973,
+      "grad_norm": 0.7653847336769104,
+      "learning_rate": 8.068161906430699e-05,
+      "loss": 1.7586,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5806811836962591,
+      "grad_norm": 0.7687065005302429,
+      "learning_rate": 8.066409672332224e-05,
+      "loss": 1.8408,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5812066870299208,
+      "grad_norm": 1.3475931882858276,
+      "learning_rate": 8.064657438233748e-05,
+      "loss": 1.7714,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5817321903635826,
+      "grad_norm": 0.9233903288841248,
+      "learning_rate": 8.062905204135273e-05,
+      "loss": 1.7366,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5822576936972443,
+      "grad_norm": 1.3624159097671509,
+      "learning_rate": 8.061152970036798e-05,
+      "loss": 1.8115,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5827831970309062,
+      "grad_norm": 1.2570765018463135,
+      "learning_rate": 8.059400735938321e-05,
+      "loss": 1.8475,
+      "step": 2218
+    },
+    {
+      "epoch": 0.583308700364568,
+      "grad_norm": 1.3805052042007446,
+      "learning_rate": 8.057648501839846e-05,
+      "loss": 1.7814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5838342036982297,
+      "grad_norm": 1.1308529376983643,
+      "learning_rate": 8.055896267741371e-05,
+      "loss": 1.7531,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5843597070318914,
+      "grad_norm": 0.8926995992660522,
+      "learning_rate": 8.054144033642895e-05,
+      "loss": 1.7973,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5848852103655533,
+      "grad_norm": 1.3856247663497925,
+      "learning_rate": 8.05239179954442e-05,
+      "loss": 1.8255,
+      "step": 2226
+    },
+    {
+      "epoch": 0.585410713699215,
+      "grad_norm": 1.143256664276123,
+      "learning_rate": 8.050639565445945e-05,
+      "loss": 1.7983,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5859362170328768,
+      "grad_norm": 0.7704371809959412,
+      "learning_rate": 8.048887331347468e-05,
+      "loss": 1.787,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5864617203665385,
+      "grad_norm": 0.8830547332763672,
+      "learning_rate": 8.047135097248993e-05,
+      "loss": 1.8016,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5869872237002004,
+      "grad_norm": 1.3716325759887695,
+      "learning_rate": 8.045382863150517e-05,
+      "loss": 1.8184,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5875127270338621,
+      "grad_norm": 0.6707213521003723,
+      "learning_rate": 8.043630629052041e-05,
+      "loss": 1.7814,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5880382303675239,
+      "grad_norm": 0.8658749461174011,
+      "learning_rate": 8.041878394953566e-05,
+      "loss": 1.7797,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5885637337011856,
+      "grad_norm": 0.8602432608604431,
+      "learning_rate": 8.040126160855091e-05,
+      "loss": 1.7646,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5890892370348474,
+      "grad_norm": 0.6430072784423828,
+      "learning_rate": 8.038373926756616e-05,
+      "loss": 1.7872,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5896147403685092,
+      "grad_norm": 0.8540019989013672,
+      "learning_rate": 8.036621692658139e-05,
+      "loss": 1.7961,
+      "step": 2244
+    },
+    {
+      "epoch": 0.590140243702171,
+      "grad_norm": 0.7394554615020752,
+      "learning_rate": 8.034869458559664e-05,
+      "loss": 1.814,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5906657470358327,
+      "grad_norm": 0.6837593913078308,
+      "learning_rate": 8.033117224461188e-05,
+      "loss": 1.7913,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5911912503694945,
+      "grad_norm": 0.7128705382347107,
+      "learning_rate": 8.031364990362713e-05,
+      "loss": 1.7655,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5917167537031564,
+      "grad_norm": 0.6362258791923523,
+      "learning_rate": 8.029612756264238e-05,
+      "loss": 1.8152,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5922422570368181,
+      "grad_norm": 0.8071087002754211,
+      "learning_rate": 8.027860522165763e-05,
+      "loss": 1.774,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5927677603704798,
+      "grad_norm": 0.713575005531311,
+      "learning_rate": 8.026108288067286e-05,
+      "loss": 1.7543,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5932932637041416,
+      "grad_norm": 1.3430129289627075,
+      "learning_rate": 8.02435605396881e-05,
+      "loss": 1.8361,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5938187670378033,
+      "grad_norm": 0.7131674885749817,
+      "learning_rate": 8.022603819870334e-05,
+      "loss": 1.7964,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5943442703714652,
+      "grad_norm": 0.8277941942214966,
+      "learning_rate": 8.020851585771859e-05,
+      "loss": 1.7872,
+      "step": 2262
+    },
+    {
+      "epoch": 0.594869773705127,
+      "grad_norm": 1.3098149299621582,
+      "learning_rate": 8.019099351673384e-05,
+      "loss": 1.7578,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5953952770387887,
+      "grad_norm": 0.7214488983154297,
+      "learning_rate": 8.017347117574909e-05,
+      "loss": 1.8064,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5959207803724504,
+      "grad_norm": 0.6886647343635559,
+      "learning_rate": 8.015594883476433e-05,
+      "loss": 1.8442,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5964462837061123,
+      "grad_norm": 0.7065162062644958,
+      "learning_rate": 8.013842649377957e-05,
+      "loss": 1.7493,
+      "step": 2270
+    },
+    {
+      "epoch": 0.596971787039774,
+      "grad_norm": 0.647866427898407,
+      "learning_rate": 8.012090415279481e-05,
+      "loss": 1.7716,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5974972903734358,
+      "grad_norm": 1.0947537422180176,
+      "learning_rate": 8.010338181181006e-05,
+      "loss": 1.7847,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5980227937070975,
+      "grad_norm": 0.7569209337234497,
+      "learning_rate": 8.008585947082531e-05,
+      "loss": 1.8247,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5985482970407594,
+      "grad_norm": 0.8344804644584656,
+      "learning_rate": 8.006833712984056e-05,
+      "loss": 1.7553,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5990738003744212,
+      "grad_norm": 0.7028648257255554,
+      "learning_rate": 8.00508147888558e-05,
+      "loss": 1.7851,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5995993037080829,
+      "grad_norm": 0.8421052098274231,
+      "learning_rate": 8.003329244787104e-05,
+      "loss": 1.7872,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6001248070417446,
+      "grad_norm": 0.6548582911491394,
+      "learning_rate": 8.001577010688627e-05,
+      "loss": 1.7568,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6006503103754064,
+      "grad_norm": 0.7493446469306946,
+      "learning_rate": 7.999824776590152e-05,
+      "loss": 1.766,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6011758137090683,
+      "grad_norm": 0.6293224096298218,
+      "learning_rate": 7.998072542491677e-05,
+      "loss": 1.7701,
+      "step": 2288
+    },
+    {
+      "epoch": 0.60170131704273,
+      "grad_norm": 0.8307104706764221,
+      "learning_rate": 7.996320308393202e-05,
+      "loss": 1.7783,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6022268203763917,
+      "grad_norm": 0.696723997592926,
+      "learning_rate": 7.994568074294726e-05,
+      "loss": 1.8009,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6027523237100535,
+      "grad_norm": 0.6556825637817383,
+      "learning_rate": 7.992815840196251e-05,
+      "loss": 1.7693,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6032778270437154,
+      "grad_norm": 0.9338749051094055,
+      "learning_rate": 7.991063606097774e-05,
+      "loss": 1.7614,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6038033303773771,
+      "grad_norm": 0.8844968676567078,
+      "learning_rate": 7.989311371999299e-05,
+      "loss": 1.7809,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6043288337110388,
+      "grad_norm": 0.6624906063079834,
+      "learning_rate": 7.987559137900824e-05,
+      "loss": 1.7607,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6048543370447006,
+      "grad_norm": 0.7565969824790955,
+      "learning_rate": 7.985806903802349e-05,
+      "loss": 1.7374,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6053798403783625,
+      "grad_norm": 1.3834869861602783,
+      "learning_rate": 7.984054669703873e-05,
+      "loss": 1.8123,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6059053437120242,
+      "grad_norm": 0.9589686989784241,
+      "learning_rate": 7.982302435605398e-05,
+      "loss": 1.7766,
+      "step": 2306
+    },
+    {
+      "epoch": 0.606430847045686,
+      "grad_norm": 1.1011096239089966,
+      "learning_rate": 7.980550201506922e-05,
+      "loss": 1.7737,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6069563503793477,
+      "grad_norm": 1.1299936771392822,
+      "learning_rate": 7.978797967408445e-05,
+      "loss": 1.8042,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6074818537130094,
+      "grad_norm": 0.7697176933288574,
+      "learning_rate": 7.97704573330997e-05,
+      "loss": 1.742,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6080073570466713,
+      "grad_norm": 0.991256833076477,
+      "learning_rate": 7.975293499211495e-05,
+      "loss": 1.8302,
+      "step": 2314
+    },
+    {
+      "epoch": 0.608532860380333,
+      "grad_norm": 0.7879564166069031,
+      "learning_rate": 7.97354126511302e-05,
+      "loss": 1.7624,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6090583637139948,
+      "grad_norm": 0.745040774345398,
+      "learning_rate": 7.971789031014544e-05,
+      "loss": 1.807,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6095838670476565,
+      "grad_norm": 0.9064005613327026,
+      "learning_rate": 7.970036796916069e-05,
+      "loss": 1.7371,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6101093703813184,
+      "grad_norm": 0.9049443602561951,
+      "learning_rate": 7.968284562817592e-05,
+      "loss": 1.8108,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6106348737149802,
+      "grad_norm": 0.8754010200500488,
+      "learning_rate": 7.966532328719117e-05,
+      "loss": 1.7804,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6111603770486419,
+      "grad_norm": 0.7384723424911499,
+      "learning_rate": 7.964780094620642e-05,
+      "loss": 1.7683,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6116858803823036,
+      "grad_norm": 0.739629328250885,
+      "learning_rate": 7.963027860522167e-05,
+      "loss": 1.7826,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6122113837159654,
+      "grad_norm": 0.824571967124939,
+      "learning_rate": 7.961275626423691e-05,
+      "loss": 1.7686,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6127368870496273,
+      "grad_norm": 0.9385930299758911,
+      "learning_rate": 7.959523392325216e-05,
+      "loss": 1.7893,
+      "step": 2332
+    },
+    {
+      "epoch": 0.613262390383289,
+      "grad_norm": 0.7331735491752625,
+      "learning_rate": 7.95777115822674e-05,
+      "loss": 1.7835,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6137878937169507,
+      "grad_norm": 0.6689608693122864,
+      "learning_rate": 7.956018924128263e-05,
+      "loss": 1.755,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6143133970506125,
+      "grad_norm": 0.7041564583778381,
+      "learning_rate": 7.954266690029788e-05,
+      "loss": 1.7593,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6148389003842744,
+      "grad_norm": 0.818277895450592,
+      "learning_rate": 7.952514455931312e-05,
+      "loss": 1.8126,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6153644037179361,
+      "grad_norm": 1.1610822677612305,
+      "learning_rate": 7.950762221832837e-05,
+      "loss": 1.8324,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6158899070515979,
+      "grad_norm": 0.9594135284423828,
+      "learning_rate": 7.949009987734362e-05,
+      "loss": 1.7777,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6164154103852596,
+      "grad_norm": 0.9519843459129333,
+      "learning_rate": 7.947257753635887e-05,
+      "loss": 1.7905,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6169409137189215,
+      "grad_norm": 0.7015102505683899,
+      "learning_rate": 7.94550551953741e-05,
+      "loss": 1.7753,
+      "step": 2348
+    },
+    {
+      "epoch": 0.6174664170525832,
+      "grad_norm": 0.6667357087135315,
+      "learning_rate": 7.943753285438935e-05,
+      "loss": 1.7925,
+      "step": 2350
+    },
+    {
+      "epoch": 0.617991920386245,
+      "grad_norm": 0.6917058229446411,
+      "learning_rate": 7.94200105134046e-05,
+      "loss": 1.7737,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6185174237199067,
+      "grad_norm": 0.8922196626663208,
+      "learning_rate": 7.940248817241984e-05,
+      "loss": 1.77,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6190429270535684,
+      "grad_norm": 0.8834502696990967,
+      "learning_rate": 7.938496583143509e-05,
+      "loss": 1.7626,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6195684303872303,
+      "grad_norm": 0.6824911832809448,
+      "learning_rate": 7.936744349045034e-05,
+      "loss": 1.7694,
+      "step": 2358
+    },
+    {
+      "epoch": 0.620093933720892,
+      "grad_norm": 0.8218874931335449,
+      "learning_rate": 7.934992114946557e-05,
+      "loss": 1.753,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6206194370545538,
+      "grad_norm": 0.808892011642456,
+      "learning_rate": 7.93323988084808e-05,
+      "loss": 1.7619,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6211449403882155,
+      "grad_norm": 0.783028781414032,
+      "learning_rate": 7.931487646749605e-05,
+      "loss": 1.8033,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6216704437218774,
+      "grad_norm": 0.8071235418319702,
+      "learning_rate": 7.92973541265113e-05,
+      "loss": 1.7536,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6221959470555392,
+      "grad_norm": 0.7900059819221497,
+      "learning_rate": 7.927983178552655e-05,
+      "loss": 1.7715,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6227214503892009,
+      "grad_norm": 0.8198074102401733,
+      "learning_rate": 7.92623094445418e-05,
+      "loss": 1.7898,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6232469537228627,
+      "grad_norm": 0.6880433559417725,
+      "learning_rate": 7.924478710355704e-05,
+      "loss": 1.7617,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6237724570565244,
+      "grad_norm": 0.7786495685577393,
+      "learning_rate": 7.922726476257228e-05,
+      "loss": 1.7574,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6242979603901863,
+      "grad_norm": 0.8043944239616394,
+      "learning_rate": 7.920974242158753e-05,
+      "loss": 1.797,
+      "step": 2376
+    },
+    {
+      "epoch": 0.624823463723848,
+      "grad_norm": 0.9602116942405701,
+      "learning_rate": 7.919222008060277e-05,
+      "loss": 1.7835,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6253489670575098,
+      "grad_norm": 0.6723161339759827,
+      "learning_rate": 7.917469773961802e-05,
+      "loss": 1.751,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6258744703911715,
+      "grad_norm": 0.7045361399650574,
+      "learning_rate": 7.915717539863327e-05,
+      "loss": 1.7788,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6263999737248334,
+      "grad_norm": 0.7056633234024048,
+      "learning_rate": 7.913965305764852e-05,
+      "loss": 1.7581,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6269254770584951,
+      "grad_norm": 0.8192391395568848,
+      "learning_rate": 7.912213071666375e-05,
+      "loss": 1.751,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6274509803921569,
+      "grad_norm": 0.8521485924720764,
+      "learning_rate": 7.910460837567898e-05,
+      "loss": 1.811,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6279764837258186,
+      "grad_norm": 0.7382224202156067,
+      "learning_rate": 7.908708603469423e-05,
+      "loss": 1.7848,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6285019870594805,
+      "grad_norm": 0.6544625163078308,
+      "learning_rate": 7.906956369370948e-05,
+      "loss": 1.7902,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6290274903931422,
+      "grad_norm": 0.7634027600288391,
+      "learning_rate": 7.905204135272473e-05,
+      "loss": 1.79,
+      "step": 2394
+    },
+    {
+      "epoch": 0.629552993726804,
+      "grad_norm": 1.0319316387176514,
+      "learning_rate": 7.903451901173997e-05,
+      "loss": 1.8045,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6300784970604657,
+      "grad_norm": 0.6364408731460571,
+      "learning_rate": 7.901699667075522e-05,
+      "loss": 1.7666,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6306040003941275,
+      "grad_norm": 0.919385552406311,
+      "learning_rate": 7.899947432977046e-05,
+      "loss": 1.7625,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6306040003941275,
+      "eval_loss": 1.7536036968231201,
+      "eval_runtime": 487.1509,
+      "eval_samples_per_second": 250.003,
+      "eval_steps_per_second": 31.251,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6311295037277893,
+      "grad_norm": 0.9996768832206726,
+      "learning_rate": 7.89819519887857e-05,
+      "loss": 1.8133,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6316550070614511,
+      "grad_norm": 0.7376594543457031,
+      "learning_rate": 7.896442964780095e-05,
+      "loss": 1.8095,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6321805103951128,
+      "grad_norm": 0.970077633857727,
+      "learning_rate": 7.89469073068162e-05,
+      "loss": 1.7818,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6327060137287746,
+      "grad_norm": 0.8934677839279175,
+      "learning_rate": 7.892938496583145e-05,
+      "loss": 1.7905,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6332315170624364,
+      "grad_norm": 0.8888778686523438,
+      "learning_rate": 7.891186262484669e-05,
+      "loss": 1.7646,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6337570203960982,
+      "grad_norm": 0.7317706942558289,
+      "learning_rate": 7.889434028386193e-05,
+      "loss": 1.8039,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6342825237297599,
+      "grad_norm": 0.768997848033905,
+      "learning_rate": 7.887681794287717e-05,
+      "loss": 1.8152,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6348080270634217,
+      "grad_norm": 0.8444989323616028,
+      "learning_rate": 7.885929560189241e-05,
+      "loss": 1.7889,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6353335303970834,
+      "grad_norm": 0.7109376788139343,
+      "learning_rate": 7.884177326090766e-05,
+      "loss": 1.7729,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6358590337307453,
+      "grad_norm": 0.630806565284729,
+      "learning_rate": 7.88242509199229e-05,
+      "loss": 1.7954,
+      "step": 2420
+    },
+    {
+      "epoch": 0.636384537064407,
+      "grad_norm": 0.6395316123962402,
+      "learning_rate": 7.880672857893815e-05,
+      "loss": 1.7797,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6369100403980688,
+      "grad_norm": 1.0122566223144531,
+      "learning_rate": 7.87892062379534e-05,
+      "loss": 1.7841,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6374355437317305,
+      "grad_norm": 0.7840449810028076,
+      "learning_rate": 7.877168389696865e-05,
+      "loss": 1.7888,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6379610470653924,
+      "grad_norm": 0.7690210342407227,
+      "learning_rate": 7.875416155598388e-05,
+      "loss": 1.8122,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6384865503990541,
+      "grad_norm": 0.7418575286865234,
+      "learning_rate": 7.873663921499913e-05,
+      "loss": 1.8056,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6390120537327159,
+      "grad_norm": 0.8744335770606995,
+      "learning_rate": 7.871911687401438e-05,
+      "loss": 1.7914,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6395375570663776,
+      "grad_norm": 0.9662806391716003,
+      "learning_rate": 7.870159453302962e-05,
+      "loss": 1.7712,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6400630604000395,
+      "grad_norm": 1.1848862171173096,
+      "learning_rate": 7.868407219204487e-05,
+      "loss": 1.7627,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6405885637337012,
+      "grad_norm": 0.7062596082687378,
+      "learning_rate": 7.86665498510601e-05,
+      "loss": 1.7691,
+      "step": 2438
+    },
+    {
+      "epoch": 0.641114067067363,
+      "grad_norm": 0.9317710399627686,
+      "learning_rate": 7.864902751007535e-05,
+      "loss": 1.7697,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6416395704010247,
+      "grad_norm": 0.8188003301620483,
+      "learning_rate": 7.863150516909059e-05,
+      "loss": 1.7934,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6421650737346865,
+      "grad_norm": 1.0256218910217285,
+      "learning_rate": 7.861398282810583e-05,
+      "loss": 1.8175,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6426905770683483,
+      "grad_norm": 0.7887519598007202,
+      "learning_rate": 7.859646048712108e-05,
+      "loss": 1.7842,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6432160804020101,
+      "grad_norm": 0.906284749507904,
+      "learning_rate": 7.857893814613633e-05,
+      "loss": 1.7407,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6437415837356718,
+      "grad_norm": 1.160643458366394,
+      "learning_rate": 7.856141580515158e-05,
+      "loss": 1.7891,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6442670870693336,
+      "grad_norm": 0.7816227078437805,
+      "learning_rate": 7.854389346416682e-05,
+      "loss": 1.802,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6447925904029954,
+      "grad_norm": 0.6715728640556335,
+      "learning_rate": 7.852637112318206e-05,
+      "loss": 1.754,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6453180937366572,
+      "grad_norm": 0.7650018334388733,
+      "learning_rate": 7.85088487821973e-05,
+      "loss": 1.7936,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6458435970703189,
+      "grad_norm": 0.8200324773788452,
+      "learning_rate": 7.849132644121255e-05,
+      "loss": 1.7879,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6463691004039807,
+      "grad_norm": 0.7117056846618652,
+      "learning_rate": 7.84738041002278e-05,
+      "loss": 1.7869,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6468946037376425,
+      "grad_norm": 0.7513619661331177,
+      "learning_rate": 7.845628175924303e-05,
+      "loss": 1.7929,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6474201070713043,
+      "grad_norm": 0.6291913986206055,
+      "learning_rate": 7.843875941825828e-05,
+      "loss": 1.7597,
+      "step": 2464
+    },
+    {
+      "epoch": 0.647945610404966,
+      "grad_norm": 0.7736865878105164,
+      "learning_rate": 7.842123707727353e-05,
+      "loss": 1.8084,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6484711137386278,
+      "grad_norm": 0.9892921447753906,
+      "learning_rate": 7.840371473628876e-05,
+      "loss": 1.7724,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6489966170722895,
+      "grad_norm": 0.5670979619026184,
+      "learning_rate": 7.838619239530401e-05,
+      "loss": 1.7806,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6495221204059514,
+      "grad_norm": 0.728164792060852,
+      "learning_rate": 7.836867005431926e-05,
+      "loss": 1.7687,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6500476237396131,
+      "grad_norm": 1.0032124519348145,
+      "learning_rate": 7.835114771333451e-05,
+      "loss": 1.7992,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6505731270732749,
+      "grad_norm": 0.70088130235672,
+      "learning_rate": 7.833362537234975e-05,
+      "loss": 1.7664,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6510986304069366,
+      "grad_norm": 0.8623471856117249,
+      "learning_rate": 7.8316103031365e-05,
+      "loss": 1.8202,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6516241337405985,
+      "grad_norm": 0.8462334275245667,
+      "learning_rate": 7.829858069038024e-05,
+      "loss": 1.8086,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6521496370742602,
+      "grad_norm": 0.6222977638244629,
+      "learning_rate": 7.828105834939548e-05,
+      "loss": 1.7396,
+      "step": 2482
+    },
+    {
+      "epoch": 0.652675140407922,
+      "grad_norm": 0.7277782559394836,
+      "learning_rate": 7.826353600841073e-05,
+      "loss": 1.787,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6532006437415837,
+      "grad_norm": 0.6451889276504517,
+      "learning_rate": 7.824601366742598e-05,
+      "loss": 1.7687,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6537261470752455,
+      "grad_norm": 0.6663830280303955,
+      "learning_rate": 7.822849132644121e-05,
+      "loss": 1.7627,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6542516504089073,
+      "grad_norm": 0.7570757269859314,
+      "learning_rate": 7.821096898545646e-05,
+      "loss": 1.7423,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6547771537425691,
+      "grad_norm": 0.5967277884483337,
+      "learning_rate": 7.819344664447171e-05,
+      "loss": 1.7834,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6553026570762308,
+      "grad_norm": 0.6351729035377502,
+      "learning_rate": 7.817592430348694e-05,
+      "loss": 1.7541,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6558281604098926,
+      "grad_norm": 0.9284831881523132,
+      "learning_rate": 7.815840196250219e-05,
+      "loss": 1.7658,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6563536637435544,
+      "grad_norm": 0.754885196685791,
+      "learning_rate": 7.814087962151744e-05,
+      "loss": 1.7602,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6568791670772162,
+      "grad_norm": 0.7284504771232605,
+      "learning_rate": 7.812335728053268e-05,
+      "loss": 1.7983,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6574046704108779,
+      "grad_norm": 0.6399169564247131,
+      "learning_rate": 7.810583493954793e-05,
+      "loss": 1.7851,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6579301737445397,
+      "grad_norm": 1.2791913747787476,
+      "learning_rate": 7.808831259856318e-05,
+      "loss": 1.8401,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6584556770782015,
+      "grad_norm": 0.8345859050750732,
+      "learning_rate": 7.807079025757841e-05,
+      "loss": 1.7636,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6589811804118633,
+      "grad_norm": 0.6579688191413879,
+      "learning_rate": 7.805326791659366e-05,
+      "loss": 1.815,
+      "step": 2508
+    },
+    {
+      "epoch": 0.659506683745525,
+      "grad_norm": 0.744471549987793,
+      "learning_rate": 7.803574557560891e-05,
+      "loss": 1.787,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6600321870791868,
+      "grad_norm": 0.6032891869544983,
+      "learning_rate": 7.801822323462416e-05,
+      "loss": 1.7615,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6605576904128485,
+      "grad_norm": 0.6453471183776855,
+      "learning_rate": 7.800070089363939e-05,
+      "loss": 1.7519,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6610831937465104,
+      "grad_norm": 0.6828714609146118,
+      "learning_rate": 7.798317855265464e-05,
+      "loss": 1.7563,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6616086970801721,
+      "grad_norm": 0.7225235104560852,
+      "learning_rate": 7.796565621166989e-05,
+      "loss": 1.7685,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6621342004138339,
+      "grad_norm": 0.7070510387420654,
+      "learning_rate": 7.794813387068512e-05,
+      "loss": 1.7816,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6626597037474956,
+      "grad_norm": 0.7075088620185852,
+      "learning_rate": 7.793061152970037e-05,
+      "loss": 1.7918,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6631852070811575,
+      "grad_norm": 0.8367542028427124,
+      "learning_rate": 7.791308918871561e-05,
+      "loss": 1.7769,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6637107104148192,
+      "grad_norm": 0.7209259867668152,
+      "learning_rate": 7.789556684773086e-05,
+      "loss": 1.7997,
+      "step": 2526
+    },
+    {
+      "epoch": 0.664236213748481,
+      "grad_norm": 0.9258558750152588,
+      "learning_rate": 7.787804450674611e-05,
+      "loss": 1.7706,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6647617170821427,
+      "grad_norm": 0.7765336632728577,
+      "learning_rate": 7.786052216576136e-05,
+      "loss": 1.7774,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6652872204158045,
+      "grad_norm": 0.8592368960380554,
+      "learning_rate": 7.784299982477659e-05,
+      "loss": 1.7496,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6658127237494663,
+      "grad_norm": 0.7928656339645386,
+      "learning_rate": 7.782547748379184e-05,
+      "loss": 1.7735,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6663382270831281,
+      "grad_norm": 0.7750053405761719,
+      "learning_rate": 7.780795514280709e-05,
+      "loss": 1.7732,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6668637304167898,
+      "grad_norm": 1.0451760292053223,
+      "learning_rate": 7.779043280182233e-05,
+      "loss": 1.7658,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6673892337504516,
+      "grad_norm": 0.6935849189758301,
+      "learning_rate": 7.777291046083757e-05,
+      "loss": 1.7494,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6679147370841134,
+      "grad_norm": 0.9037797451019287,
+      "learning_rate": 7.775538811985282e-05,
+      "loss": 1.7664,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6684402404177752,
+      "grad_norm": 0.6649421453475952,
+      "learning_rate": 7.773786577886806e-05,
+      "loss": 1.7503,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6689657437514369,
+      "grad_norm": 0.6808927059173584,
+      "learning_rate": 7.77203434378833e-05,
+      "loss": 1.7481,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6694912470850987,
+      "grad_norm": 0.697090208530426,
+      "learning_rate": 7.770282109689854e-05,
+      "loss": 1.7843,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6700167504187605,
+      "grad_norm": 0.9084567427635193,
+      "learning_rate": 7.768529875591379e-05,
+      "loss": 1.7643,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6705422537524223,
+      "grad_norm": 0.6704044938087463,
+      "learning_rate": 7.766777641492904e-05,
+      "loss": 1.7687,
+      "step": 2552
+    },
+    {
+      "epoch": 0.671067757086084,
+      "grad_norm": 0.7542858719825745,
+      "learning_rate": 7.765025407394429e-05,
+      "loss": 1.771,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6715932604197458,
+      "grad_norm": 0.8199552893638611,
+      "learning_rate": 7.763273173295953e-05,
+      "loss": 1.8042,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6721187637534075,
+      "grad_norm": 0.8224305510520935,
+      "learning_rate": 7.761520939197477e-05,
+      "loss": 1.7625,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6726442670870694,
+      "grad_norm": 0.7174823880195618,
+      "learning_rate": 7.759768705099002e-05,
+      "loss": 1.7857,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6731697704207311,
+      "grad_norm": 0.6753933429718018,
+      "learning_rate": 7.758016471000526e-05,
+      "loss": 1.7732,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6736952737543929,
+      "grad_norm": 0.985587477684021,
+      "learning_rate": 7.75626423690205e-05,
+      "loss": 1.797,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6742207770880546,
+      "grad_norm": 0.892760694026947,
+      "learning_rate": 7.754512002803575e-05,
+      "loss": 1.7881,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6747462804217165,
+      "grad_norm": 0.7832928895950317,
+      "learning_rate": 7.752759768705099e-05,
+      "loss": 1.781,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6752717837553782,
+      "grad_norm": 0.7372546195983887,
+      "learning_rate": 7.751007534606624e-05,
+      "loss": 1.7774,
+      "step": 2570
+    },
+    {
+      "epoch": 0.67579728708904,
+      "grad_norm": 0.7375915050506592,
+      "learning_rate": 7.749255300508147e-05,
+      "loss": 1.8019,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6763227904227017,
+      "grad_norm": 0.77280193567276,
+      "learning_rate": 7.747503066409672e-05,
+      "loss": 1.7743,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6768482937563635,
+      "grad_norm": 0.7224514484405518,
+      "learning_rate": 7.745750832311197e-05,
+      "loss": 1.7741,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6773737970900253,
+      "grad_norm": 0.6910998821258545,
+      "learning_rate": 7.743998598212722e-05,
+      "loss": 1.7893,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6778993004236871,
+      "grad_norm": 0.61247718334198,
+      "learning_rate": 7.742246364114247e-05,
+      "loss": 1.7628,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6784248037573488,
+      "grad_norm": 0.6819799542427063,
+      "learning_rate": 7.740494130015771e-05,
+      "loss": 1.7355,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6789503070910106,
+      "grad_norm": 0.6858199834823608,
+      "learning_rate": 7.738741895917295e-05,
+      "loss": 1.772,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6794758104246724,
+      "grad_norm": 0.7759047150611877,
+      "learning_rate": 7.73698966181882e-05,
+      "loss": 1.7586,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6800013137583342,
+      "grad_norm": 0.6968771815299988,
+      "learning_rate": 7.735237427720344e-05,
+      "loss": 1.7886,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6805268170919959,
+      "grad_norm": 0.7313429117202759,
+      "learning_rate": 7.733485193621868e-05,
+      "loss": 1.7884,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6810523204256577,
+      "grad_norm": 0.6731216311454773,
+      "learning_rate": 7.731732959523392e-05,
+      "loss": 1.8149,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6815778237593195,
+      "grad_norm": 0.6619842648506165,
+      "learning_rate": 7.729980725424917e-05,
+      "loss": 1.7993,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6821033270929813,
+      "grad_norm": 0.6684291362762451,
+      "learning_rate": 7.728228491326442e-05,
+      "loss": 1.7386,
+      "step": 2596
+    },
+    {
+      "epoch": 0.682628830426643,
+      "grad_norm": 0.6439480781555176,
+      "learning_rate": 7.726476257227965e-05,
+      "loss": 1.7711,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6831543337603048,
+      "grad_norm": 0.9060842990875244,
+      "learning_rate": 7.72472402312949e-05,
+      "loss": 1.7901,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6836798370939665,
+      "grad_norm": 0.8999879360198975,
+      "learning_rate": 7.722971789031015e-05,
+      "loss": 1.755,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6842053404276284,
+      "grad_norm": 0.931769609451294,
+      "learning_rate": 7.72121955493254e-05,
+      "loss": 1.7682,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6847308437612901,
+      "grad_norm": 0.6413145065307617,
+      "learning_rate": 7.719467320834064e-05,
+      "loss": 1.7774,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6852563470949519,
+      "grad_norm": 0.7932469248771667,
+      "learning_rate": 7.717715086735589e-05,
+      "loss": 1.7702,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6857818504286136,
+      "grad_norm": 0.9040171504020691,
+      "learning_rate": 7.715962852637112e-05,
+      "loss": 1.7799,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6863073537622755,
+      "grad_norm": 0.8220160603523254,
+      "learning_rate": 7.714210618538637e-05,
+      "loss": 1.776,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6868328570959372,
+      "grad_norm": 0.727171778678894,
+      "learning_rate": 7.712458384440162e-05,
+      "loss": 1.7711,
+      "step": 2614
+    },
+    {
+      "epoch": 0.687358360429599,
+      "grad_norm": 0.877223789691925,
+      "learning_rate": 7.710706150341685e-05,
+      "loss": 1.7487,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6878838637632607,
+      "grad_norm": 0.7638031244277954,
+      "learning_rate": 7.70895391624321e-05,
+      "loss": 1.7925,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6884093670969226,
+      "grad_norm": 0.8082540035247803,
+      "learning_rate": 7.707201682144735e-05,
+      "loss": 1.756,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6889348704305843,
+      "grad_norm": 0.8931254148483276,
+      "learning_rate": 7.70544944804626e-05,
+      "loss": 1.7783,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6894603737642461,
+      "grad_norm": 0.760176420211792,
+      "learning_rate": 7.703697213947783e-05,
+      "loss": 1.7603,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6899858770979078,
+      "grad_norm": 0.7421066761016846,
+      "learning_rate": 7.701944979849308e-05,
+      "loss": 1.7872,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6905113804315696,
+      "grad_norm": 0.7226428389549255,
+      "learning_rate": 7.700192745750833e-05,
+      "loss": 1.7947,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6910368837652314,
+      "grad_norm": 0.7266933917999268,
+      "learning_rate": 7.698440511652357e-05,
+      "loss": 1.7512,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6915623870988932,
+      "grad_norm": 0.6181286573410034,
+      "learning_rate": 7.696688277553882e-05,
+      "loss": 1.7382,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6920878904325549,
+      "grad_norm": 0.7923069000244141,
+      "learning_rate": 7.694936043455407e-05,
+      "loss": 1.772,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6926133937662167,
+      "grad_norm": 0.7019143104553223,
+      "learning_rate": 7.69318380935693e-05,
+      "loss": 1.7768,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6931388970998785,
+      "grad_norm": 0.6440560817718506,
+      "learning_rate": 7.691431575258455e-05,
+      "loss": 1.784,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6936644004335403,
+      "grad_norm": 0.6912879943847656,
+      "learning_rate": 7.689679341159978e-05,
+      "loss": 1.7714,
+      "step": 2640
+    },
+    {
+      "epoch": 0.694189903767202,
+      "grad_norm": 0.7078375816345215,
+      "learning_rate": 7.687927107061503e-05,
+      "loss": 1.7774,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6947154071008638,
+      "grad_norm": 0.744601309299469,
+      "learning_rate": 7.686174872963028e-05,
+      "loss": 1.781,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6952409104345255,
+      "grad_norm": 0.6456273198127747,
+      "learning_rate": 7.684422638864553e-05,
+      "loss": 1.7704,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6957664137681874,
+      "grad_norm": 1.024562954902649,
+      "learning_rate": 7.682670404766077e-05,
+      "loss": 1.7578,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6962919171018491,
+      "grad_norm": 0.7917899489402771,
+      "learning_rate": 7.680918170667601e-05,
+      "loss": 1.7853,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6968174204355109,
+      "grad_norm": 0.8094285130500793,
+      "learning_rate": 7.679165936569126e-05,
+      "loss": 1.7531,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6973429237691726,
+      "grad_norm": 0.5724372863769531,
+      "learning_rate": 7.67741370247065e-05,
+      "loss": 1.7458,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6978684271028345,
+      "grad_norm": 0.8631569743156433,
+      "learning_rate": 7.675661468372175e-05,
+      "loss": 1.7559,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6983939304364962,
+      "grad_norm": 0.8773946762084961,
+      "learning_rate": 7.6739092342737e-05,
+      "loss": 1.7658,
+      "step": 2658
+    },
+    {
+      "epoch": 0.698919433770158,
+      "grad_norm": 0.7718110084533691,
+      "learning_rate": 7.672157000175225e-05,
+      "loss": 1.8134,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6994449371038197,
+      "grad_norm": 0.7242169380187988,
+      "learning_rate": 7.670404766076748e-05,
+      "loss": 1.7982,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6999704404374816,
+      "grad_norm": 0.8516372442245483,
+      "learning_rate": 7.668652531978273e-05,
+      "loss": 1.7878,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7004959437711433,
+      "grad_norm": 0.7234377861022949,
+      "learning_rate": 7.666900297879796e-05,
+      "loss": 1.808,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7010214471048051,
+      "grad_norm": 0.6989150643348694,
+      "learning_rate": 7.665148063781321e-05,
+      "loss": 1.7998,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7015469504384668,
+      "grad_norm": 0.8885288834571838,
+      "learning_rate": 7.663395829682846e-05,
+      "loss": 1.7768,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7020724537721286,
+      "grad_norm": 0.7263723015785217,
+      "learning_rate": 7.66164359558437e-05,
+      "loss": 1.7789,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7025979571057904,
+      "grad_norm": 0.7461345195770264,
+      "learning_rate": 7.659891361485895e-05,
+      "loss": 1.7583,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7031234604394522,
+      "grad_norm": 1.028860330581665,
+      "learning_rate": 7.65813912738742e-05,
+      "loss": 1.7666,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7036489637731139,
+      "grad_norm": 0.7666818499565125,
+      "learning_rate": 7.656386893288943e-05,
+      "loss": 1.7621,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7041744671067757,
+      "grad_norm": 0.8842204809188843,
+      "learning_rate": 7.654634659190468e-05,
+      "loss": 1.7664,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7046999704404375,
+      "grad_norm": 1.2368742227554321,
+      "learning_rate": 7.652882425091993e-05,
+      "loss": 1.7572,
+      "step": 2682
+    },
+    {
+      "epoch": 0.7052254737740993,
+      "grad_norm": 0.9390943646430969,
+      "learning_rate": 7.651130190993518e-05,
+      "loss": 1.7482,
+      "step": 2684
+    },
+    {
+      "epoch": 0.705750977107761,
+      "grad_norm": 0.7266187071800232,
+      "learning_rate": 7.649377956895042e-05,
+      "loss": 1.7961,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7062764804414228,
+      "grad_norm": 0.6799050569534302,
+      "learning_rate": 7.647625722796567e-05,
+      "loss": 1.7712,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7068019837750845,
+      "grad_norm": 0.8757466077804565,
+      "learning_rate": 7.64587348869809e-05,
+      "loss": 1.7721,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7073274871087464,
+      "grad_norm": 0.8347486257553101,
+      "learning_rate": 7.644121254599614e-05,
+      "loss": 1.7481,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7078529904424081,
+      "grad_norm": 0.7270652055740356,
+      "learning_rate": 7.642369020501139e-05,
+      "loss": 1.7818,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7083784937760699,
+      "grad_norm": 0.563240110874176,
+      "learning_rate": 7.640616786402663e-05,
+      "loss": 1.769,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7089039971097316,
+      "grad_norm": 0.8982949256896973,
+      "learning_rate": 7.638864552304188e-05,
+      "loss": 1.7806,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7094295004433935,
+      "grad_norm": 0.723839282989502,
+      "learning_rate": 7.637112318205713e-05,
+      "loss": 1.7947,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7099550037770552,
+      "grad_norm": 0.7281327247619629,
+      "learning_rate": 7.635360084107238e-05,
+      "loss": 1.7763,
+      "step": 2702
+    },
+    {
+      "epoch": 0.710480507110717,
+      "grad_norm": 0.8034355044364929,
+      "learning_rate": 7.633607850008761e-05,
+      "loss": 1.7894,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7110060104443787,
+      "grad_norm": 0.7230488061904907,
+      "learning_rate": 7.631855615910286e-05,
+      "loss": 1.7715,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7115315137780406,
+      "grad_norm": 0.9625870585441589,
+      "learning_rate": 7.63010338181181e-05,
+      "loss": 1.7966,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7120570171117023,
+      "grad_norm": 0.6187042593955994,
+      "learning_rate": 7.628351147713335e-05,
+      "loss": 1.791,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7125825204453641,
+      "grad_norm": 0.7607492208480835,
+      "learning_rate": 7.62659891361486e-05,
+      "loss": 1.8201,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7131080237790258,
+      "grad_norm": 0.8220197558403015,
+      "learning_rate": 7.624846679516385e-05,
+      "loss": 1.7512,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7136335271126876,
+      "grad_norm": 0.7219741344451904,
+      "learning_rate": 7.623094445417908e-05,
+      "loss": 1.775,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7141590304463494,
+      "grad_norm": 0.9267223477363586,
+      "learning_rate": 7.621342211319432e-05,
+      "loss": 1.7833,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7146845337800112,
+      "grad_norm": 0.6037401556968689,
+      "learning_rate": 7.619589977220956e-05,
+      "loss": 1.7956,
+      "step": 2720
+    },
+    {
+      "epoch": 0.7152100371136729,
+      "grad_norm": 0.7281090021133423,
+      "learning_rate": 7.617837743122481e-05,
+      "loss": 1.7341,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7157355404473347,
+      "grad_norm": 0.8907596468925476,
+      "learning_rate": 7.616085509024006e-05,
+      "loss": 1.754,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7162610437809965,
+      "grad_norm": 0.7699323296546936,
+      "learning_rate": 7.61433327492553e-05,
+      "loss": 1.7591,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7167865471146583,
+      "grad_norm": 0.7458289265632629,
+      "learning_rate": 7.612581040827055e-05,
+      "loss": 1.8033,
+      "step": 2728
+    },
+    {
+      "epoch": 0.71731205044832,
+      "grad_norm": 1.1340159177780151,
+      "learning_rate": 7.610828806728579e-05,
+      "loss": 1.7821,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7178375537819818,
+      "grad_norm": 0.6701779365539551,
+      "learning_rate": 7.609076572630104e-05,
+      "loss": 1.7681,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7183630571156436,
+      "grad_norm": 0.7182425260543823,
+      "learning_rate": 7.607324338531628e-05,
+      "loss": 1.7735,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7188885604493054,
+      "grad_norm": 0.8409538865089417,
+      "learning_rate": 7.605572104433153e-05,
+      "loss": 1.7535,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7194140637829671,
+      "grad_norm": 0.7824596166610718,
+      "learning_rate": 7.603819870334678e-05,
+      "loss": 1.7853,
+      "step": 2738
+    },
+    {
+      "epoch": 0.7199395671166289,
+      "grad_norm": 1.0866069793701172,
+      "learning_rate": 7.602067636236203e-05,
+      "loss": 1.7729,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7204650704502906,
+      "grad_norm": 0.7940483093261719,
+      "learning_rate": 7.600315402137726e-05,
+      "loss": 1.7643,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7209905737839525,
+      "grad_norm": 1.0263162851333618,
+      "learning_rate": 7.59856316803925e-05,
+      "loss": 1.7533,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7215160771176142,
+      "grad_norm": 0.7818773984909058,
+      "learning_rate": 7.596810933940774e-05,
+      "loss": 1.7526,
+      "step": 2746
+    },
+    {
+      "epoch": 0.722041580451276,
+      "grad_norm": 0.7003962397575378,
+      "learning_rate": 7.595058699842299e-05,
+      "loss": 1.7635,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7225670837849377,
+      "grad_norm": 0.7112312316894531,
+      "learning_rate": 7.593306465743824e-05,
+      "loss": 1.7608,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7230925871185996,
+      "grad_norm": 0.8331362009048462,
+      "learning_rate": 7.591554231645348e-05,
+      "loss": 1.7899,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7236180904522613,
+      "grad_norm": 0.7973011136054993,
+      "learning_rate": 7.589801997546873e-05,
+      "loss": 1.757,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7241435937859231,
+      "grad_norm": 0.6794710159301758,
+      "learning_rate": 7.588049763448397e-05,
+      "loss": 1.7516,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7246690971195848,
+      "grad_norm": 0.9882004857063293,
+      "learning_rate": 7.586297529349921e-05,
+      "loss": 1.7499,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7251946004532466,
+      "grad_norm": 0.6301134824752808,
+      "learning_rate": 7.584545295251446e-05,
+      "loss": 1.7785,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7257201037869084,
+      "grad_norm": 0.8858461976051331,
+      "learning_rate": 7.582793061152971e-05,
+      "loss": 1.7956,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7262456071205702,
+      "grad_norm": 0.7689917087554932,
+      "learning_rate": 7.581040827054496e-05,
+      "loss": 1.7588,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7267711104542319,
+      "grad_norm": 0.7277519702911377,
+      "learning_rate": 7.57928859295602e-05,
+      "loss": 1.7878,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7272966137878937,
+      "grad_norm": 0.7690980434417725,
+      "learning_rate": 7.577536358857544e-05,
+      "loss": 1.7738,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7278221171215555,
+      "grad_norm": 0.9652357697486877,
+      "learning_rate": 7.575784124759067e-05,
+      "loss": 1.8106,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7283476204552173,
+      "grad_norm": 0.7968404293060303,
+      "learning_rate": 7.574031890660592e-05,
+      "loss": 1.7581,
+      "step": 2772
+    },
+    {
+      "epoch": 0.728873123788879,
+      "grad_norm": 0.7790765762329102,
+      "learning_rate": 7.572279656562117e-05,
+      "loss": 1.7257,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7293986271225408,
+      "grad_norm": 0.93352872133255,
+      "learning_rate": 7.570527422463641e-05,
+      "loss": 1.7343,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7299241304562026,
+      "grad_norm": 0.8516173958778381,
+      "learning_rate": 7.568775188365166e-05,
+      "loss": 1.7442,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7304496337898644,
+      "grad_norm": 0.6152732968330383,
+      "learning_rate": 7.567022954266691e-05,
+      "loss": 1.7601,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7309751371235261,
+      "grad_norm": 0.6822034120559692,
+      "learning_rate": 7.565270720168214e-05,
+      "loss": 1.7915,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7315006404571879,
+      "grad_norm": 0.6148113012313843,
+      "learning_rate": 7.563518486069739e-05,
+      "loss": 1.7642,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7320261437908496,
+      "grad_norm": 0.9175143241882324,
+      "learning_rate": 7.561766251971264e-05,
+      "loss": 1.7588,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7325516471245115,
+      "grad_norm": 0.6481335759162903,
+      "learning_rate": 7.560014017872789e-05,
+      "loss": 1.7838,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7330771504581732,
+      "grad_norm": 0.712563693523407,
+      "learning_rate": 7.558261783774313e-05,
+      "loss": 1.7449,
+      "step": 2790
+    },
+    {
+      "epoch": 0.733602653791835,
+      "grad_norm": 0.7027430534362793,
+      "learning_rate": 7.556509549675838e-05,
+      "loss": 1.7544,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7341281571254967,
+      "grad_norm": 0.7059524059295654,
+      "learning_rate": 7.554757315577362e-05,
+      "loss": 1.7766,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7346536604591586,
+      "grad_norm": 0.7603355050086975,
+      "learning_rate": 7.553005081478885e-05,
+      "loss": 1.7778,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7351791637928203,
+      "grad_norm": 0.8169555068016052,
+      "learning_rate": 7.55125284738041e-05,
+      "loss": 1.768,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7357046671264821,
+      "grad_norm": 0.7924453616142273,
+      "learning_rate": 7.549500613281934e-05,
+      "loss": 1.7617,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7357046671264821,
+      "eval_loss": 1.736789345741272,
+      "eval_runtime": 487.2587,
+      "eval_samples_per_second": 249.947,
+      "eval_steps_per_second": 31.244,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7362301704601438,
+      "grad_norm": 0.6122123003005981,
+      "learning_rate": 7.547748379183459e-05,
+      "loss": 1.7861,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7367556737938056,
+      "grad_norm": 0.7799514532089233,
+      "learning_rate": 7.545996145084984e-05,
+      "loss": 1.7612,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7372811771274674,
+      "grad_norm": 0.6509242057800293,
+      "learning_rate": 7.544243910986509e-05,
+      "loss": 1.8034,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7378066804611292,
+      "grad_norm": 0.6152432560920715,
+      "learning_rate": 7.542491676888032e-05,
+      "loss": 1.7565,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7383321837947909,
+      "grad_norm": 0.6861807703971863,
+      "learning_rate": 7.540739442789557e-05,
+      "loss": 1.7622,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7388576871284527,
+      "grad_norm": 0.6434677839279175,
+      "learning_rate": 7.538987208691082e-05,
+      "loss": 1.7778,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7393831904621145,
+      "grad_norm": 0.8894173502922058,
+      "learning_rate": 7.537234974592606e-05,
+      "loss": 1.7921,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7399086937957763,
+      "grad_norm": 0.6787139177322388,
+      "learning_rate": 7.535482740494131e-05,
+      "loss": 1.76,
+      "step": 2816
+    },
+    {
+      "epoch": 0.740434197129438,
+      "grad_norm": 0.6434260010719299,
+      "learning_rate": 7.533730506395656e-05,
+      "loss": 1.7702,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7409597004630998,
+      "grad_norm": 0.6635193228721619,
+      "learning_rate": 7.531978272297179e-05,
+      "loss": 1.7631,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7414852037967616,
+      "grad_norm": 0.6992824077606201,
+      "learning_rate": 7.530226038198703e-05,
+      "loss": 1.8017,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7420107071304234,
+      "grad_norm": 0.6800288558006287,
+      "learning_rate": 7.528473804100227e-05,
+      "loss": 1.764,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7425362104640851,
+      "grad_norm": 1.0732684135437012,
+      "learning_rate": 7.526721570001752e-05,
+      "loss": 1.7818,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7430617137977469,
+      "grad_norm": 0.7014878392219543,
+      "learning_rate": 7.524969335903277e-05,
+      "loss": 1.7774,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7435872171314086,
+      "grad_norm": 0.5895276069641113,
+      "learning_rate": 7.523217101804802e-05,
+      "loss": 1.7622,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7441127204650705,
+      "grad_norm": 0.8901596069335938,
+      "learning_rate": 7.521464867706326e-05,
+      "loss": 1.7579,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7446382237987322,
+      "grad_norm": 0.9512175917625427,
+      "learning_rate": 7.51971263360785e-05,
+      "loss": 1.7481,
+      "step": 2834
+    },
+    {
+      "epoch": 0.745163727132394,
+      "grad_norm": 0.8138533234596252,
+      "learning_rate": 7.517960399509375e-05,
+      "loss": 1.7612,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7456892304660557,
+      "grad_norm": 0.7234623432159424,
+      "learning_rate": 7.5162081654109e-05,
+      "loss": 1.7587,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7462147337997176,
+      "grad_norm": 1.0871793031692505,
+      "learning_rate": 7.514455931312424e-05,
+      "loss": 1.7368,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7467402371333793,
+      "grad_norm": 0.7515408992767334,
+      "learning_rate": 7.512703697213949e-05,
+      "loss": 1.8096,
+      "step": 2842
+    },
+    {
+      "epoch": 0.7472657404670411,
+      "grad_norm": 0.6623795032501221,
+      "learning_rate": 7.510951463115472e-05,
+      "loss": 1.7683,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7477912438007028,
+      "grad_norm": 0.648363471031189,
+      "learning_rate": 7.509199229016997e-05,
+      "loss": 1.8195,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7483167471343646,
+      "grad_norm": 0.7021984457969666,
+      "learning_rate": 7.50744699491852e-05,
+      "loss": 1.7918,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7488422504680264,
+      "grad_norm": 0.7864859104156494,
+      "learning_rate": 7.505694760820045e-05,
+      "loss": 1.8118,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7493677538016882,
+      "grad_norm": 0.6326330900192261,
+      "learning_rate": 7.50394252672157e-05,
+      "loss": 1.7781,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7498932571353499,
+      "grad_norm": 0.6461377143859863,
+      "learning_rate": 7.502190292623095e-05,
+      "loss": 1.77,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7504187604690117,
+      "grad_norm": 0.6641056537628174,
+      "learning_rate": 7.50043805852462e-05,
+      "loss": 1.7442,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7509442638026735,
+      "grad_norm": 0.6882733702659607,
+      "learning_rate": 7.498685824426144e-05,
+      "loss": 1.7583,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7514697671363353,
+      "grad_norm": 0.9753492474555969,
+      "learning_rate": 7.496933590327668e-05,
+      "loss": 1.7557,
+      "step": 2860
+    },
+    {
+      "epoch": 0.751995270469997,
+      "grad_norm": 0.6425254940986633,
+      "learning_rate": 7.495181356229192e-05,
+      "loss": 1.8,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7525207738036588,
+      "grad_norm": 0.7045446038246155,
+      "learning_rate": 7.493429122130717e-05,
+      "loss": 1.7532,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7530462771373206,
+      "grad_norm": 0.6855698227882385,
+      "learning_rate": 7.491676888032242e-05,
+      "loss": 1.7652,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7535717804709824,
+      "grad_norm": 0.7373823523521423,
+      "learning_rate": 7.489924653933767e-05,
+      "loss": 1.7896,
+      "step": 2868
+    },
+    {
+      "epoch": 0.7540972838046441,
+      "grad_norm": 0.772221565246582,
+      "learning_rate": 7.48817241983529e-05,
+      "loss": 1.7712,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7546227871383059,
+      "grad_norm": 0.7220898270606995,
+      "learning_rate": 7.486420185736815e-05,
+      "loss": 1.7298,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7551482904719676,
+      "grad_norm": 0.6807803511619568,
+      "learning_rate": 7.484667951638338e-05,
+      "loss": 1.7767,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7556737938056295,
+      "grad_norm": 0.7383838891983032,
+      "learning_rate": 7.482915717539863e-05,
+      "loss": 1.7722,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7561992971392912,
+      "grad_norm": 0.7259317636489868,
+      "learning_rate": 7.481163483441388e-05,
+      "loss": 1.7623,
+      "step": 2878
+    },
+    {
+      "epoch": 0.756724800472953,
+      "grad_norm": 0.7161348462104797,
+      "learning_rate": 7.479411249342913e-05,
+      "loss": 1.7379,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7572503038066147,
+      "grad_norm": 0.5866290330886841,
+      "learning_rate": 7.477659015244437e-05,
+      "loss": 1.7687,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7577758071402766,
+      "grad_norm": 0.6606796979904175,
+      "learning_rate": 7.475906781145962e-05,
+      "loss": 1.7833,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7583013104739383,
+      "grad_norm": 0.6400638818740845,
+      "learning_rate": 7.474154547047485e-05,
+      "loss": 1.75,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7588268138076001,
+      "grad_norm": 0.6338980793952942,
+      "learning_rate": 7.47240231294901e-05,
+      "loss": 1.7836,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7593523171412618,
+      "grad_norm": 0.6896232962608337,
+      "learning_rate": 7.470650078850535e-05,
+      "loss": 1.7473,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7598778204749237,
+      "grad_norm": 0.7504851222038269,
+      "learning_rate": 7.46889784475206e-05,
+      "loss": 1.7587,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7604033238085854,
+      "grad_norm": 0.6796631217002869,
+      "learning_rate": 7.467145610653584e-05,
+      "loss": 1.7469,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7609288271422472,
+      "grad_norm": 0.6032044887542725,
+      "learning_rate": 7.465393376555108e-05,
+      "loss": 1.7651,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7614543304759089,
+      "grad_norm": 0.6399370431900024,
+      "learning_rate": 7.463641142456633e-05,
+      "loss": 1.8041,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7619798338095707,
+      "grad_norm": 0.6167407631874084,
+      "learning_rate": 7.461888908358156e-05,
+      "loss": 1.7743,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7625053371432325,
+      "grad_norm": 0.5767862200737,
+      "learning_rate": 7.460136674259681e-05,
+      "loss": 1.7598,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7630308404768943,
+      "grad_norm": 0.6222682595252991,
+      "learning_rate": 7.458384440161206e-05,
+      "loss": 1.7659,
+      "step": 2904
+    },
+    {
+      "epoch": 0.763556343810556,
+      "grad_norm": 0.6252115368843079,
+      "learning_rate": 7.45663220606273e-05,
+      "loss": 1.7671,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7640818471442178,
+      "grad_norm": 0.6266006231307983,
+      "learning_rate": 7.454879971964255e-05,
+      "loss": 1.7786,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7646073504778796,
+      "grad_norm": 0.7761850357055664,
+      "learning_rate": 7.45312773786578e-05,
+      "loss": 1.7185,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7651328538115414,
+      "grad_norm": 0.6215353608131409,
+      "learning_rate": 7.451375503767303e-05,
+      "loss": 1.7478,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7656583571452031,
+      "grad_norm": 0.7758198380470276,
+      "learning_rate": 7.449623269668828e-05,
+      "loss": 1.7437,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7661838604788649,
+      "grad_norm": 0.5999752879142761,
+      "learning_rate": 7.447871035570353e-05,
+      "loss": 1.7534,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7667093638125266,
+      "grad_norm": 0.6463642120361328,
+      "learning_rate": 7.446118801471877e-05,
+      "loss": 1.7589,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7672348671461885,
+      "grad_norm": 0.7435876727104187,
+      "learning_rate": 7.444366567373402e-05,
+      "loss": 1.8053,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7677603704798502,
+      "grad_norm": 0.7085327506065369,
+      "learning_rate": 7.442614333274926e-05,
+      "loss": 1.748,
+      "step": 2922
+    },
+    {
+      "epoch": 0.768285873813512,
+      "grad_norm": 0.6690971255302429,
+      "learning_rate": 7.44086209917645e-05,
+      "loss": 1.808,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7688113771471737,
+      "grad_norm": 0.652035653591156,
+      "learning_rate": 7.439109865077975e-05,
+      "loss": 1.7491,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7693368804808356,
+      "grad_norm": 0.6042243242263794,
+      "learning_rate": 7.437357630979499e-05,
+      "loss": 1.751,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7698623838144973,
+      "grad_norm": 0.5847947597503662,
+      "learning_rate": 7.435605396881023e-05,
+      "loss": 1.7539,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7703878871481591,
+      "grad_norm": 0.690543532371521,
+      "learning_rate": 7.433853162782548e-05,
+      "loss": 1.7448,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7709133904818208,
+      "grad_norm": 0.7835954427719116,
+      "learning_rate": 7.432100928684073e-05,
+      "loss": 1.7352,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7714388938154827,
+      "grad_norm": 0.8784381747245789,
+      "learning_rate": 7.430348694585598e-05,
+      "loss": 1.7802,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7719643971491444,
+      "grad_norm": 0.8104349970817566,
+      "learning_rate": 7.428596460487122e-05,
+      "loss": 1.755,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7724899004828062,
+      "grad_norm": 0.6043236255645752,
+      "learning_rate": 7.426844226388646e-05,
+      "loss": 1.8118,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7730154038164679,
+      "grad_norm": 1.0382120609283447,
+      "learning_rate": 7.42509199229017e-05,
+      "loss": 1.7832,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7735409071501297,
+      "grad_norm": 0.6753326654434204,
+      "learning_rate": 7.423339758191695e-05,
+      "loss": 1.7886,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7740664104837915,
+      "grad_norm": 1.226515531539917,
+      "learning_rate": 7.421587524093219e-05,
+      "loss": 1.7406,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7745919138174533,
+      "grad_norm": 0.7555555105209351,
+      "learning_rate": 7.419835289994743e-05,
+      "loss": 1.726,
+      "step": 2948
+    },
+    {
+      "epoch": 0.775117417151115,
+      "grad_norm": 1.073789119720459,
+      "learning_rate": 7.418083055896268e-05,
+      "loss": 1.7905,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7756429204847768,
+      "grad_norm": 0.883283793926239,
+      "learning_rate": 7.416330821797793e-05,
+      "loss": 1.7516,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7761684238184386,
+      "grad_norm": 0.8123281598091125,
+      "learning_rate": 7.414578587699316e-05,
+      "loss": 1.7689,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7766939271521004,
+      "grad_norm": 0.6473522782325745,
+      "learning_rate": 7.412826353600841e-05,
+      "loss": 1.7966,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7772194304857621,
+      "grad_norm": 0.6950981616973877,
+      "learning_rate": 7.411074119502366e-05,
+      "loss": 1.7494,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7777449338194239,
+      "grad_norm": 0.7919045686721802,
+      "learning_rate": 7.40932188540389e-05,
+      "loss": 1.7992,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7782704371530856,
+      "grad_norm": 0.8082287311553955,
+      "learning_rate": 7.407569651305415e-05,
+      "loss": 1.758,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7787959404867475,
+      "grad_norm": 0.634069561958313,
+      "learning_rate": 7.40581741720694e-05,
+      "loss": 1.7855,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7793214438204092,
+      "grad_norm": 1.0830199718475342,
+      "learning_rate": 7.404065183108463e-05,
+      "loss": 1.7509,
+      "step": 2966
+    },
+    {
+      "epoch": 0.779846947154071,
+      "grad_norm": 0.6961039304733276,
+      "learning_rate": 7.402312949009988e-05,
+      "loss": 1.7491,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7803724504877327,
+      "grad_norm": 0.7842292189598083,
+      "learning_rate": 7.400560714911513e-05,
+      "loss": 1.736,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7808979538213946,
+      "grad_norm": 0.6680390238761902,
+      "learning_rate": 7.398808480813036e-05,
+      "loss": 1.7594,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7814234571550563,
+      "grad_norm": 0.8992615342140198,
+      "learning_rate": 7.397056246714561e-05,
+      "loss": 1.7811,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7819489604887181,
+      "grad_norm": 0.6779314279556274,
+      "learning_rate": 7.395304012616086e-05,
+      "loss": 1.7675,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7824744638223798,
+      "grad_norm": 0.5508474111557007,
+      "learning_rate": 7.39355177851761e-05,
+      "loss": 1.8152,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7829999671560417,
+      "grad_norm": 0.6748946905136108,
+      "learning_rate": 7.391799544419134e-05,
+      "loss": 1.7596,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7835254704897034,
+      "grad_norm": 0.5707883834838867,
+      "learning_rate": 7.390047310320659e-05,
+      "loss": 1.765,
+      "step": 2982
+    },
+    {
+      "epoch": 0.7840509738233652,
+      "grad_norm": 0.6725517511367798,
+      "learning_rate": 7.388295076222184e-05,
+      "loss": 1.7528,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7845764771570269,
+      "grad_norm": 0.6516979336738586,
+      "learning_rate": 7.386542842123708e-05,
+      "loss": 1.7613,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7851019804906887,
+      "grad_norm": 0.6247593760490417,
+      "learning_rate": 7.384790608025233e-05,
+      "loss": 1.7519,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7856274838243505,
+      "grad_norm": 0.6806461811065674,
+      "learning_rate": 7.383038373926758e-05,
+      "loss": 1.7501,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7861529871580123,
+      "grad_norm": 0.8175075054168701,
+      "learning_rate": 7.381286139828281e-05,
+      "loss": 1.7543,
+      "step": 2992
+    },
+    {
+      "epoch": 0.786678490491674,
+      "grad_norm": 0.8868039846420288,
+      "learning_rate": 7.379533905729806e-05,
+      "loss": 1.7285,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7872039938253358,
+      "grad_norm": 0.9624283909797668,
+      "learning_rate": 7.377781671631331e-05,
+      "loss": 1.7679,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7877294971589976,
+      "grad_norm": 0.8716928958892822,
+      "learning_rate": 7.376029437532854e-05,
+      "loss": 1.7642,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7882550004926594,
+      "grad_norm": 0.6427202224731445,
+      "learning_rate": 7.374277203434379e-05,
+      "loss": 1.777,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7887805038263211,
+      "grad_norm": 0.6569937467575073,
+      "learning_rate": 7.372524969335904e-05,
+      "loss": 1.7414,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7893060071599829,
+      "grad_norm": 1.117759346961975,
+      "learning_rate": 7.370772735237428e-05,
+      "loss": 1.7709,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7898315104936446,
+      "grad_norm": 0.6267141103744507,
+      "learning_rate": 7.369020501138952e-05,
+      "loss": 1.7871,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7903570138273065,
+      "grad_norm": 0.8071964979171753,
+      "learning_rate": 7.367268267040477e-05,
+      "loss": 1.7471,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7908825171609682,
+      "grad_norm": 0.702384352684021,
+      "learning_rate": 7.365516032942001e-05,
+      "loss": 1.7434,
+      "step": 3010
+    },
+    {
+      "epoch": 0.79140802049463,
+      "grad_norm": 0.6770474314689636,
+      "learning_rate": 7.363763798843526e-05,
+      "loss": 1.782,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7919335238282917,
+      "grad_norm": 0.6293635964393616,
+      "learning_rate": 7.362011564745051e-05,
+      "loss": 1.7682,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7924590271619536,
+      "grad_norm": 0.8292271494865417,
+      "learning_rate": 7.360259330646576e-05,
+      "loss": 1.7729,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7929845304956153,
+      "grad_norm": 0.6828389167785645,
+      "learning_rate": 7.358507096548099e-05,
+      "loss": 1.7521,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7935100338292771,
+      "grad_norm": 0.6849939823150635,
+      "learning_rate": 7.356754862449624e-05,
+      "loss": 1.7619,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7940355371629388,
+      "grad_norm": 0.720439612865448,
+      "learning_rate": 7.355002628351149e-05,
+      "loss": 1.7725,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7945610404966007,
+      "grad_norm": 0.6729586124420166,
+      "learning_rate": 7.353250394252672e-05,
+      "loss": 1.7711,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7950865438302624,
+      "grad_norm": 1.0918195247650146,
+      "learning_rate": 7.351498160154197e-05,
+      "loss": 1.7764,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7956120471639242,
+      "grad_norm": 0.658743143081665,
+      "learning_rate": 7.349745926055721e-05,
+      "loss": 1.7641,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7961375504975859,
+      "grad_norm": 0.5973094701766968,
+      "learning_rate": 7.347993691957246e-05,
+      "loss": 1.7554,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7966630538312477,
+      "grad_norm": 0.6641756296157837,
+      "learning_rate": 7.34624145785877e-05,
+      "loss": 1.7656,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7971885571649096,
+      "grad_norm": 0.7535148859024048,
+      "learning_rate": 7.344489223760294e-05,
+      "loss": 1.7521,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7977140604985713,
+      "grad_norm": 0.9680157899856567,
+      "learning_rate": 7.342736989661819e-05,
+      "loss": 1.7794,
+      "step": 3036
+    },
+    {
+      "epoch": 0.798239563832233,
+      "grad_norm": 0.7855157852172852,
+      "learning_rate": 7.340984755563344e-05,
+      "loss": 1.7853,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7987650671658948,
+      "grad_norm": 0.6831044554710388,
+      "learning_rate": 7.339232521464869e-05,
+      "loss": 1.7443,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7992905704995567,
+      "grad_norm": 0.7435096502304077,
+      "learning_rate": 7.337480287366393e-05,
+      "loss": 1.7585,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7998160738332184,
+      "grad_norm": 0.6614308953285217,
+      "learning_rate": 7.335728053267917e-05,
+      "loss": 1.7872,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8003415771668801,
+      "grad_norm": 0.5782405138015747,
+      "learning_rate": 7.333975819169442e-05,
+      "loss": 1.7504,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8008670805005419,
+      "grad_norm": 0.6519070267677307,
+      "learning_rate": 7.332223585070965e-05,
+      "loss": 1.7396,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8013925838342038,
+      "grad_norm": 0.5866365432739258,
+      "learning_rate": 7.33047135097249e-05,
+      "loss": 1.7591,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8019180871678655,
+      "grad_norm": 0.6096078753471375,
+      "learning_rate": 7.328719116874014e-05,
+      "loss": 1.7871,
+      "step": 3052
+    },
+    {
+      "epoch": 0.8024435905015272,
+      "grad_norm": 0.6846382021903992,
+      "learning_rate": 7.326966882775539e-05,
+      "loss": 1.7773,
+      "step": 3054
+    },
+    {
+      "epoch": 0.802969093835189,
+      "grad_norm": 0.6193353533744812,
+      "learning_rate": 7.325214648677064e-05,
+      "loss": 1.753,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8034945971688507,
+      "grad_norm": 0.6320629119873047,
+      "learning_rate": 7.323462414578587e-05,
+      "loss": 1.7679,
+      "step": 3058
+    },
+    {
+      "epoch": 0.8040201005025126,
+      "grad_norm": 0.5982667803764343,
+      "learning_rate": 7.321710180480112e-05,
+      "loss": 1.7307,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8045456038361744,
+      "grad_norm": 0.7248689532279968,
+      "learning_rate": 7.319957946381637e-05,
+      "loss": 1.7933,
+      "step": 3062
+    },
+    {
+      "epoch": 0.8050711071698361,
+      "grad_norm": 0.7433560490608215,
+      "learning_rate": 7.318205712283162e-05,
+      "loss": 1.7393,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8055966105034978,
+      "grad_norm": 0.6755779981613159,
+      "learning_rate": 7.316453478184686e-05,
+      "loss": 1.7589,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8061221138371597,
+      "grad_norm": 0.6949239373207092,
+      "learning_rate": 7.314701244086211e-05,
+      "loss": 1.7677,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8066476171708215,
+      "grad_norm": 0.6781786680221558,
+      "learning_rate": 7.312949009987735e-05,
+      "loss": 1.7027,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8071731205044832,
+      "grad_norm": 0.683310866355896,
+      "learning_rate": 7.311196775889259e-05,
+      "loss": 1.7205,
+      "step": 3072
+    },
+    {
+      "epoch": 0.807698623838145,
+      "grad_norm": 0.5861634016036987,
+      "learning_rate": 7.309444541790783e-05,
+      "loss": 1.7412,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8082241271718067,
+      "grad_norm": 0.7344016432762146,
+      "learning_rate": 7.307692307692307e-05,
+      "loss": 1.7988,
+      "step": 3076
+    },
+    {
+      "epoch": 0.8087496305054686,
+      "grad_norm": 0.5996577143669128,
+      "learning_rate": 7.305940073593832e-05,
+      "loss": 1.765,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8092751338391303,
+      "grad_norm": 0.5766566395759583,
+      "learning_rate": 7.304187839495357e-05,
+      "loss": 1.8002,
+      "step": 3080
+    },
+    {
+      "epoch": 0.809800637172792,
+      "grad_norm": 0.6364811658859253,
+      "learning_rate": 7.302435605396882e-05,
+      "loss": 1.7687,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8103261405064538,
+      "grad_norm": 0.839227557182312,
+      "learning_rate": 7.300683371298405e-05,
+      "loss": 1.7501,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8108516438401157,
+      "grad_norm": 0.6285102367401123,
+      "learning_rate": 7.29893113719993e-05,
+      "loss": 1.7503,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8113771471737774,
+      "grad_norm": 0.6087677478790283,
+      "learning_rate": 7.297178903101455e-05,
+      "loss": 1.7873,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8119026505074391,
+      "grad_norm": 0.9094337821006775,
+      "learning_rate": 7.29542666900298e-05,
+      "loss": 1.7501,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8124281538411009,
+      "grad_norm": 0.6166443228721619,
+      "learning_rate": 7.293674434904504e-05,
+      "loss": 1.7809,
+      "step": 3092
+    },
+    {
+      "epoch": 0.8129536571747628,
+      "grad_norm": 0.6993762850761414,
+      "learning_rate": 7.291922200806029e-05,
+      "loss": 1.776,
+      "step": 3094
+    },
+    {
+      "epoch": 0.8134791605084245,
+      "grad_norm": 0.6359695792198181,
+      "learning_rate": 7.290169966707552e-05,
+      "loss": 1.7177,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8140046638420863,
+      "grad_norm": 0.8014838695526123,
+      "learning_rate": 7.288417732609077e-05,
+      "loss": 1.7488,
+      "step": 3098
+    },
+    {
+      "epoch": 0.814530167175748,
+      "grad_norm": 0.6601728200912476,
+      "learning_rate": 7.2866654985106e-05,
+      "loss": 1.8006,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8150556705094097,
+      "grad_norm": 0.9497177004814148,
+      "learning_rate": 7.284913264412125e-05,
+      "loss": 1.8234,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8155811738430716,
+      "grad_norm": 0.7122120261192322,
+      "learning_rate": 7.28316103031365e-05,
+      "loss": 1.7306,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8161066771767334,
+      "grad_norm": 0.7118192911148071,
+      "learning_rate": 7.281408796215175e-05,
+      "loss": 1.7178,
+      "step": 3106
+    },
+    {
+      "epoch": 0.8166321805103951,
+      "grad_norm": 0.727682888507843,
+      "learning_rate": 7.2796565621167e-05,
+      "loss": 1.7857,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8171576838440568,
+      "grad_norm": 0.6266892552375793,
+      "learning_rate": 7.277904328018223e-05,
+      "loss": 1.7531,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8176831871777187,
+      "grad_norm": 1.1099108457565308,
+      "learning_rate": 7.276152093919748e-05,
+      "loss": 1.7579,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8182086905113805,
+      "grad_norm": 0.7087392807006836,
+      "learning_rate": 7.274399859821272e-05,
+      "loss": 1.7393,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8187341938450422,
+      "grad_norm": 0.9023381471633911,
+      "learning_rate": 7.272647625722797e-05,
+      "loss": 1.7488,
+      "step": 3116
+    },
+    {
+      "epoch": 0.819259697178704,
+      "grad_norm": 0.6250995993614197,
+      "learning_rate": 7.270895391624322e-05,
+      "loss": 1.7567,
+      "step": 3118
+    },
+    {
+      "epoch": 0.8197852005123657,
+      "grad_norm": 0.7343935370445251,
+      "learning_rate": 7.269143157525847e-05,
+      "loss": 1.7801,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8203107038460276,
+      "grad_norm": 0.7513467073440552,
+      "learning_rate": 7.26739092342737e-05,
+      "loss": 1.7638,
+      "step": 3122
+    },
+    {
+      "epoch": 0.8208362071796893,
+      "grad_norm": 0.7709615230560303,
+      "learning_rate": 7.265638689328895e-05,
+      "loss": 1.7477,
+      "step": 3124
+    },
+    {
+      "epoch": 0.821361710513351,
+      "grad_norm": 0.6068372130393982,
+      "learning_rate": 7.263886455230418e-05,
+      "loss": 1.7597,
+      "step": 3126
+    },
+    {
+      "epoch": 0.8218872138470128,
+      "grad_norm": 0.6252472400665283,
+      "learning_rate": 7.262134221131943e-05,
+      "loss": 1.7413,
+      "step": 3128
+    },
+    {
+      "epoch": 0.8224127171806747,
+      "grad_norm": 0.6601640582084656,
+      "learning_rate": 7.260381987033468e-05,
+      "loss": 1.7298,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8229382205143364,
+      "grad_norm": 0.6942901015281677,
+      "learning_rate": 7.258629752934993e-05,
+      "loss": 1.7788,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8234637238479982,
+      "grad_norm": 0.7625568509101868,
+      "learning_rate": 7.256877518836517e-05,
+      "loss": 1.752,
+      "step": 3134
+    },
+    {
+      "epoch": 0.8239892271816599,
+      "grad_norm": 0.7857789993286133,
+      "learning_rate": 7.25512528473804e-05,
+      "loss": 1.7903,
+      "step": 3136
+    },
+    {
+      "epoch": 0.8245147305153218,
+      "grad_norm": 0.76043301820755,
+      "learning_rate": 7.253373050639565e-05,
+      "loss": 1.7619,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8250402338489835,
+      "grad_norm": 0.7142301201820374,
+      "learning_rate": 7.25162081654109e-05,
+      "loss": 1.7328,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8255657371826453,
+      "grad_norm": 0.8170753717422485,
+      "learning_rate": 7.249868582442615e-05,
+      "loss": 1.7779,
+      "step": 3142
+    },
+    {
+      "epoch": 0.826091240516307,
+      "grad_norm": 0.6277784109115601,
+      "learning_rate": 7.24811634834414e-05,
+      "loss": 1.7421,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8266167438499687,
+      "grad_norm": 0.8075504899024963,
+      "learning_rate": 7.246364114245664e-05,
+      "loss": 1.7504,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8271422471836306,
+      "grad_norm": 0.5615305304527283,
+      "learning_rate": 7.244611880147188e-05,
+      "loss": 1.7537,
+      "step": 3148
+    },
+    {
+      "epoch": 0.8276677505172924,
+      "grad_norm": 0.7479259967803955,
+      "learning_rate": 7.242859646048711e-05,
+      "loss": 1.7801,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8281932538509541,
+      "grad_norm": 0.7294136881828308,
+      "learning_rate": 7.241107411950236e-05,
+      "loss": 1.7752,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8287187571846159,
+      "grad_norm": 0.6307454705238342,
+      "learning_rate": 7.239355177851761e-05,
+      "loss": 1.8088,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8292442605182777,
+      "grad_norm": 0.641791582107544,
+      "learning_rate": 7.237602943753286e-05,
+      "loss": 1.758,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8297697638519395,
+      "grad_norm": 0.7693712711334229,
+      "learning_rate": 7.23585070965481e-05,
+      "loss": 1.7816,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8302952671856012,
+      "grad_norm": 0.6319524049758911,
+      "learning_rate": 7.234098475556335e-05,
+      "loss": 1.7644,
+      "step": 3160
+    },
+    {
+      "epoch": 0.830820770519263,
+      "grad_norm": 0.6433089375495911,
+      "learning_rate": 7.232346241457858e-05,
+      "loss": 1.7876,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8313462738529247,
+      "grad_norm": 0.5543965697288513,
+      "learning_rate": 7.230594007359383e-05,
+      "loss": 1.7678,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8318717771865866,
+      "grad_norm": 0.7125136852264404,
+      "learning_rate": 7.228841773260908e-05,
+      "loss": 1.7816,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8323972805202483,
+      "grad_norm": 0.6893459558486938,
+      "learning_rate": 7.227089539162433e-05,
+      "loss": 1.7556,
+      "step": 3168
+    },
+    {
+      "epoch": 0.83292278385391,
+      "grad_norm": 0.7028675675392151,
+      "learning_rate": 7.225337305063957e-05,
+      "loss": 1.752,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8334482871875718,
+      "grad_norm": 0.6112826466560364,
+      "learning_rate": 7.223585070965482e-05,
+      "loss": 1.7367,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8339737905212337,
+      "grad_norm": 0.6377979516983032,
+      "learning_rate": 7.221832836867006e-05,
+      "loss": 1.7452,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8344992938548954,
+      "grad_norm": 0.6925122141838074,
+      "learning_rate": 7.22008060276853e-05,
+      "loss": 1.7776,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8350247971885572,
+      "grad_norm": 0.6226949095726013,
+      "learning_rate": 7.218328368670054e-05,
+      "loss": 1.748,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8355503005222189,
+      "grad_norm": 0.7138300538063049,
+      "learning_rate": 7.216576134571579e-05,
+      "loss": 1.7662,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8360758038558808,
+      "grad_norm": 0.6248802542686462,
+      "learning_rate": 7.214823900473103e-05,
+      "loss": 1.7652,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8366013071895425,
+      "grad_norm": 0.6834786534309387,
+      "learning_rate": 7.213071666374628e-05,
+      "loss": 1.7537,
+      "step": 3184
+    },
+    {
+      "epoch": 0.8371268105232043,
+      "grad_norm": 0.6411855220794678,
+      "learning_rate": 7.211319432276153e-05,
+      "loss": 1.7582,
+      "step": 3186
+    },
+    {
+      "epoch": 0.837652313856866,
+      "grad_norm": 0.6991042494773865,
+      "learning_rate": 7.209567198177678e-05,
+      "loss": 1.7643,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8381778171905278,
+      "grad_norm": 0.6581319570541382,
+      "learning_rate": 7.207814964079201e-05,
+      "loss": 1.7283,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8387033205241896,
+      "grad_norm": 0.6132378578186035,
+      "learning_rate": 7.206062729980726e-05,
+      "loss": 1.7724,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8392288238578514,
+      "grad_norm": 0.6994782090187073,
+      "learning_rate": 7.20431049588225e-05,
+      "loss": 1.7765,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8397543271915131,
+      "grad_norm": 0.7247324585914612,
+      "learning_rate": 7.202558261783775e-05,
+      "loss": 1.7665,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8402798305251749,
+      "grad_norm": 0.8014911413192749,
+      "learning_rate": 7.2008060276853e-05,
+      "loss": 1.7471,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8408053338588367,
+      "grad_norm": 0.7040480971336365,
+      "learning_rate": 7.199053793586825e-05,
+      "loss": 1.7527,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8408053338588367,
+      "eval_loss": 1.7256534099578857,
+      "eval_runtime": 487.1811,
+      "eval_samples_per_second": 249.987,
+      "eval_steps_per_second": 31.249,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8413308371924985,
+      "grad_norm": 0.8999149799346924,
+      "learning_rate": 7.197301559488348e-05,
+      "loss": 1.7836,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8418563405261602,
+      "grad_norm": 0.6944252252578735,
+      "learning_rate": 7.195549325389872e-05,
+      "loss": 1.7336,
+      "step": 3204
+    },
+    {
+      "epoch": 0.842381843859822,
+      "grad_norm": 0.5566868185997009,
+      "learning_rate": 7.193797091291396e-05,
+      "loss": 1.7681,
+      "step": 3206
+    },
+    {
+      "epoch": 0.8429073471934838,
+      "grad_norm": 0.769062340259552,
+      "learning_rate": 7.192044857192921e-05,
+      "loss": 1.7477,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8434328505271456,
+      "grad_norm": 0.6222507953643799,
+      "learning_rate": 7.190292623094446e-05,
+      "loss": 1.7576,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8439583538608073,
+      "grad_norm": 0.6584329009056091,
+      "learning_rate": 7.18854038899597e-05,
+      "loss": 1.7658,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8444838571944691,
+      "grad_norm": 0.6822264194488525,
+      "learning_rate": 7.186788154897495e-05,
+      "loss": 1.7709,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8450093605281308,
+      "grad_norm": 0.6141505241394043,
+      "learning_rate": 7.185035920799019e-05,
+      "loss": 1.776,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8455348638617927,
+      "grad_norm": 0.5974141359329224,
+      "learning_rate": 7.183283686700543e-05,
+      "loss": 1.7246,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8460603671954544,
+      "grad_norm": 0.6883708238601685,
+      "learning_rate": 7.181531452602068e-05,
+      "loss": 1.7609,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8465858705291162,
+      "grad_norm": 0.6883281469345093,
+      "learning_rate": 7.179779218503593e-05,
+      "loss": 1.7315,
+      "step": 3222
+    },
+    {
+      "epoch": 0.8471113738627779,
+      "grad_norm": 0.6654126048088074,
+      "learning_rate": 7.178026984405118e-05,
+      "loss": 1.7469,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8476368771964398,
+      "grad_norm": 0.8119237422943115,
+      "learning_rate": 7.176274750306641e-05,
+      "loss": 1.7882,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8481623805301015,
+      "grad_norm": 0.7793521285057068,
+      "learning_rate": 7.174522516208166e-05,
+      "loss": 1.7826,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8486878838637633,
+      "grad_norm": 0.8359899520874023,
+      "learning_rate": 7.172770282109689e-05,
+      "loss": 1.7612,
+      "step": 3230
+    },
+    {
+      "epoch": 0.849213387197425,
+      "grad_norm": 0.5970791578292847,
+      "learning_rate": 7.171018048011214e-05,
+      "loss": 1.7498,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8497388905310868,
+      "grad_norm": 0.6869466304779053,
+      "learning_rate": 7.169265813912739e-05,
+      "loss": 1.7787,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8502643938647486,
+      "grad_norm": 0.6755763292312622,
+      "learning_rate": 7.167513579814264e-05,
+      "loss": 1.7883,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8507898971984104,
+      "grad_norm": 0.8182030320167542,
+      "learning_rate": 7.165761345715788e-05,
+      "loss": 1.7821,
+      "step": 3238
+    },
+    {
+      "epoch": 0.8513154005320721,
+      "grad_norm": 0.6454432010650635,
+      "learning_rate": 7.164009111617313e-05,
+      "loss": 1.7527,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8518409038657339,
+      "grad_norm": 0.7202356457710266,
+      "learning_rate": 7.162256877518836e-05,
+      "loss": 1.7821,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8523664071993957,
+      "grad_norm": 0.7236865758895874,
+      "learning_rate": 7.160504643420361e-05,
+      "loss": 1.7579,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8528919105330575,
+      "grad_norm": 0.6406378746032715,
+      "learning_rate": 7.158752409321886e-05,
+      "loss": 1.7459,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8534174138667192,
+      "grad_norm": 0.755293607711792,
+      "learning_rate": 7.157000175223411e-05,
+      "loss": 1.7638,
+      "step": 3248
+    },
+    {
+      "epoch": 0.853942917200381,
+      "grad_norm": 0.7986418604850769,
+      "learning_rate": 7.155247941124936e-05,
+      "loss": 1.7686,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8544684205340428,
+      "grad_norm": 0.7703737616539001,
+      "learning_rate": 7.153495707026459e-05,
+      "loss": 1.7372,
+      "step": 3252
+    },
+    {
+      "epoch": 0.8549939238677046,
+      "grad_norm": 0.6074422597885132,
+      "learning_rate": 7.151743472927984e-05,
+      "loss": 1.7765,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8555194272013663,
+      "grad_norm": 0.7662899494171143,
+      "learning_rate": 7.149991238829507e-05,
+      "loss": 1.7563,
+      "step": 3256
+    },
+    {
+      "epoch": 0.8560449305350281,
+      "grad_norm": 0.7859123945236206,
+      "learning_rate": 7.148239004731032e-05,
+      "loss": 1.7462,
+      "step": 3258
+    },
+    {
+      "epoch": 0.8565704338686898,
+      "grad_norm": 0.5845335721969604,
+      "learning_rate": 7.146486770632557e-05,
+      "loss": 1.7524,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8570959372023517,
+      "grad_norm": 0.6083472967147827,
+      "learning_rate": 7.144734536534081e-05,
+      "loss": 1.7705,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8576214405360134,
+      "grad_norm": 0.779712438583374,
+      "learning_rate": 7.142982302435606e-05,
+      "loss": 1.7768,
+      "step": 3264
+    },
+    {
+      "epoch": 0.8581469438696752,
+      "grad_norm": 0.7343006134033203,
+      "learning_rate": 7.141230068337131e-05,
+      "loss": 1.7489,
+      "step": 3266
+    },
+    {
+      "epoch": 0.8586724472033369,
+      "grad_norm": 0.578289270401001,
+      "learning_rate": 7.139477834238654e-05,
+      "loss": 1.7436,
+      "step": 3268
+    },
+    {
+      "epoch": 0.8591979505369988,
+      "grad_norm": 0.6657706499099731,
+      "learning_rate": 7.137725600140179e-05,
+      "loss": 1.7674,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8597234538706605,
+      "grad_norm": 0.7936644554138184,
+      "learning_rate": 7.135973366041704e-05,
+      "loss": 1.7482,
+      "step": 3272
+    },
+    {
+      "epoch": 0.8602489572043223,
+      "grad_norm": 0.5892787575721741,
+      "learning_rate": 7.134221131943229e-05,
+      "loss": 1.7448,
+      "step": 3274
+    },
+    {
+      "epoch": 0.860774460537984,
+      "grad_norm": 0.6341478228569031,
+      "learning_rate": 7.132468897844753e-05,
+      "loss": 1.7613,
+      "step": 3276
+    },
+    {
+      "epoch": 0.8612999638716458,
+      "grad_norm": 0.684622585773468,
+      "learning_rate": 7.130716663746277e-05,
+      "loss": 1.7685,
+      "step": 3278
+    },
+    {
+      "epoch": 0.8618254672053076,
+      "grad_norm": 0.5765745639801025,
+      "learning_rate": 7.128964429647801e-05,
+      "loss": 1.7889,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8623509705389694,
+      "grad_norm": 0.5164791941642761,
+      "learning_rate": 7.127212195549325e-05,
+      "loss": 1.7573,
+      "step": 3282
+    },
+    {
+      "epoch": 0.8628764738726311,
+      "grad_norm": 0.5809277892112732,
+      "learning_rate": 7.12545996145085e-05,
+      "loss": 1.8003,
+      "step": 3284
+    },
+    {
+      "epoch": 0.8634019772062929,
+      "grad_norm": 0.6859455108642578,
+      "learning_rate": 7.123707727352374e-05,
+      "loss": 1.726,
+      "step": 3286
+    },
+    {
+      "epoch": 0.8639274805399547,
+      "grad_norm": 0.665773332118988,
+      "learning_rate": 7.121955493253899e-05,
+      "loss": 1.7629,
+      "step": 3288
+    },
+    {
+      "epoch": 0.8644529838736165,
+      "grad_norm": 0.7473315596580505,
+      "learning_rate": 7.120203259155424e-05,
+      "loss": 1.7538,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8649784872072782,
+      "grad_norm": 0.7632318735122681,
+      "learning_rate": 7.118451025056949e-05,
+      "loss": 1.7755,
+      "step": 3292
+    },
+    {
+      "epoch": 0.86550399054094,
+      "grad_norm": 0.7813208699226379,
+      "learning_rate": 7.116698790958472e-05,
+      "loss": 1.7835,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8660294938746018,
+      "grad_norm": 0.9293962121009827,
+      "learning_rate": 7.114946556859997e-05,
+      "loss": 1.7332,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8665549972082636,
+      "grad_norm": 0.9213319420814514,
+      "learning_rate": 7.113194322761522e-05,
+      "loss": 1.709,
+      "step": 3298
+    },
+    {
+      "epoch": 0.8670805005419253,
+      "grad_norm": 0.5981359481811523,
+      "learning_rate": 7.111442088663046e-05,
+      "loss": 1.7406,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8676060038755871,
+      "grad_norm": 0.645785927772522,
+      "learning_rate": 7.109689854564571e-05,
+      "loss": 1.7659,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8681315072092488,
+      "grad_norm": 0.9009891748428345,
+      "learning_rate": 7.107937620466094e-05,
+      "loss": 1.7335,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8686570105429107,
+      "grad_norm": 0.7231364250183105,
+      "learning_rate": 7.106185386367619e-05,
+      "loss": 1.7378,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8691825138765724,
+      "grad_norm": 0.8442168235778809,
+      "learning_rate": 7.104433152269143e-05,
+      "loss": 1.7722,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8697080172102342,
+      "grad_norm": 0.7314670085906982,
+      "learning_rate": 7.102680918170667e-05,
+      "loss": 1.7594,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8702335205438959,
+      "grad_norm": 0.6394951343536377,
+      "learning_rate": 7.100928684072192e-05,
+      "loss": 1.7693,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8707590238775578,
+      "grad_norm": 0.612392008304596,
+      "learning_rate": 7.099176449973717e-05,
+      "loss": 1.7777,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8712845272112195,
+      "grad_norm": 0.6011560559272766,
+      "learning_rate": 7.097424215875242e-05,
+      "loss": 1.7429,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8718100305448813,
+      "grad_norm": 0.9169111847877502,
+      "learning_rate": 7.095671981776766e-05,
+      "loss": 1.7723,
+      "step": 3318
+    },
+    {
+      "epoch": 0.872335533878543,
+      "grad_norm": 0.5939310193061829,
+      "learning_rate": 7.09391974767829e-05,
+      "loss": 1.718,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8728610372122048,
+      "grad_norm": 0.8067646026611328,
+      "learning_rate": 7.092167513579815e-05,
+      "loss": 1.7679,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8733865405458666,
+      "grad_norm": 0.6771594882011414,
+      "learning_rate": 7.090415279481339e-05,
+      "loss": 1.7405,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8739120438795284,
+      "grad_norm": 0.7373068332672119,
+      "learning_rate": 7.088663045382864e-05,
+      "loss": 1.7363,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8744375472131901,
+      "grad_norm": 0.5647407174110413,
+      "learning_rate": 7.086910811284387e-05,
+      "loss": 1.8027,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8749630505468519,
+      "grad_norm": 0.5459885001182556,
+      "learning_rate": 7.085158577185912e-05,
+      "loss": 1.7588,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8754885538805137,
+      "grad_norm": 0.6479038596153259,
+      "learning_rate": 7.083406343087437e-05,
+      "loss": 1.8066,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8760140572141755,
+      "grad_norm": 0.6865916848182678,
+      "learning_rate": 7.08165410898896e-05,
+      "loss": 1.7605,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8765395605478372,
+      "grad_norm": 0.6146122813224792,
+      "learning_rate": 7.079901874890485e-05,
+      "loss": 1.7711,
+      "step": 3336
+    },
+    {
+      "epoch": 0.877065063881499,
+      "grad_norm": 0.6822938323020935,
+      "learning_rate": 7.07814964079201e-05,
+      "loss": 1.7355,
+      "step": 3338
+    },
+    {
+      "epoch": 0.8775905672151608,
+      "grad_norm": 0.8722227215766907,
+      "learning_rate": 7.076397406693535e-05,
+      "loss": 1.7405,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8781160705488226,
+      "grad_norm": 0.6625831127166748,
+      "learning_rate": 7.07464517259506e-05,
+      "loss": 1.748,
+      "step": 3342
+    },
+    {
+      "epoch": 0.8786415738824843,
+      "grad_norm": 0.5351794362068176,
+      "learning_rate": 7.072892938496584e-05,
+      "loss": 1.747,
+      "step": 3344
+    },
+    {
+      "epoch": 0.8791670772161461,
+      "grad_norm": 0.7087706923484802,
+      "learning_rate": 7.071140704398108e-05,
+      "loss": 1.7475,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8796925805498078,
+      "grad_norm": 0.6210707426071167,
+      "learning_rate": 7.069388470299632e-05,
+      "loss": 1.7666,
+      "step": 3348
+    },
+    {
+      "epoch": 0.8802180838834697,
+      "grad_norm": 0.5860678553581238,
+      "learning_rate": 7.067636236201157e-05,
+      "loss": 1.728,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8807435872171314,
+      "grad_norm": 0.5733693242073059,
+      "learning_rate": 7.065884002102682e-05,
+      "loss": 1.8006,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8812690905507932,
+      "grad_norm": 0.6897549629211426,
+      "learning_rate": 7.064131768004205e-05,
+      "loss": 1.7615,
+      "step": 3354
+    },
+    {
+      "epoch": 0.8817945938844549,
+      "grad_norm": 0.8119019865989685,
+      "learning_rate": 7.06237953390573e-05,
+      "loss": 1.7428,
+      "step": 3356
+    },
+    {
+      "epoch": 0.8823200972181168,
+      "grad_norm": 0.6543797254562378,
+      "learning_rate": 7.060627299807255e-05,
+      "loss": 1.7592,
+      "step": 3358
+    },
+    {
+      "epoch": 0.8828456005517785,
+      "grad_norm": 0.6914211511611938,
+      "learning_rate": 7.058875065708778e-05,
+      "loss": 1.7549,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8833711038854403,
+      "grad_norm": 0.7868301868438721,
+      "learning_rate": 7.057122831610303e-05,
+      "loss": 1.7475,
+      "step": 3362
+    },
+    {
+      "epoch": 0.883896607219102,
+      "grad_norm": 0.7719436287879944,
+      "learning_rate": 7.055370597511828e-05,
+      "loss": 1.7549,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8844221105527639,
+      "grad_norm": 0.6737284064292908,
+      "learning_rate": 7.053618363413352e-05,
+      "loss": 1.7392,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8849476138864256,
+      "grad_norm": 1.0450035333633423,
+      "learning_rate": 7.051866129314877e-05,
+      "loss": 1.736,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8854731172200874,
+      "grad_norm": 0.6924079060554504,
+      "learning_rate": 7.050113895216402e-05,
+      "loss": 1.7431,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8859986205537491,
+      "grad_norm": 0.689011812210083,
+      "learning_rate": 7.048361661117925e-05,
+      "loss": 1.7567,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8865241238874109,
+      "grad_norm": 0.6025006771087646,
+      "learning_rate": 7.04660942701945e-05,
+      "loss": 1.7339,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8870496272210727,
+      "grad_norm": 0.6246035695075989,
+      "learning_rate": 7.044857192920975e-05,
+      "loss": 1.7728,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8875751305547345,
+      "grad_norm": 0.92098069190979,
+      "learning_rate": 7.0431049588225e-05,
+      "loss": 1.7452,
+      "step": 3378
+    },
+    {
+      "epoch": 0.8881006338883962,
+      "grad_norm": 0.6286795139312744,
+      "learning_rate": 7.041352724724023e-05,
+      "loss": 1.752,
+      "step": 3380
+    },
+    {
+      "epoch": 0.888626137222058,
+      "grad_norm": 0.5665922164916992,
+      "learning_rate": 7.039600490625548e-05,
+      "loss": 1.7561,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8891516405557198,
+      "grad_norm": 0.7606804966926575,
+      "learning_rate": 7.037848256527073e-05,
+      "loss": 1.7903,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8896771438893816,
+      "grad_norm": 0.8399646282196045,
+      "learning_rate": 7.036096022428596e-05,
+      "loss": 1.8045,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8902026472230433,
+      "grad_norm": 0.6141343116760254,
+      "learning_rate": 7.03434378833012e-05,
+      "loss": 1.7665,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8907281505567051,
+      "grad_norm": 0.7554699778556824,
+      "learning_rate": 7.032591554231645e-05,
+      "loss": 1.7219,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8912536538903668,
+      "grad_norm": 0.7068594694137573,
+      "learning_rate": 7.03083932013317e-05,
+      "loss": 1.7689,
+      "step": 3392
+    },
+    {
+      "epoch": 0.8917791572240287,
+      "grad_norm": 0.8859004378318787,
+      "learning_rate": 7.029087086034695e-05,
+      "loss": 1.7261,
+      "step": 3394
+    },
+    {
+      "epoch": 0.8923046605576904,
+      "grad_norm": 0.7646594047546387,
+      "learning_rate": 7.02733485193622e-05,
+      "loss": 1.7813,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8928301638913522,
+      "grad_norm": 0.7105104923248291,
+      "learning_rate": 7.025582617837743e-05,
+      "loss": 1.7528,
+      "step": 3398
+    },
+    {
+      "epoch": 0.8933556672250139,
+      "grad_norm": 0.5750377178192139,
+      "learning_rate": 7.023830383739268e-05,
+      "loss": 1.7324,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8938811705586758,
+      "grad_norm": 0.6942424178123474,
+      "learning_rate": 7.022078149640793e-05,
+      "loss": 1.7352,
+      "step": 3402
+    },
+    {
+      "epoch": 0.8944066738923375,
+      "grad_norm": 0.7094271779060364,
+      "learning_rate": 7.020325915542317e-05,
+      "loss": 1.7449,
+      "step": 3404
+    },
+    {
+      "epoch": 0.8949321772259993,
+      "grad_norm": 0.6620864868164062,
+      "learning_rate": 7.018573681443841e-05,
+      "loss": 1.7718,
+      "step": 3406
+    },
+    {
+      "epoch": 0.895457680559661,
+      "grad_norm": 0.6577492952346802,
+      "learning_rate": 7.016821447345366e-05,
+      "loss": 1.745,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8959831838933229,
+      "grad_norm": 0.6020835041999817,
+      "learning_rate": 7.01506921324689e-05,
+      "loss": 1.7486,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8965086872269846,
+      "grad_norm": 0.6429753303527832,
+      "learning_rate": 7.013316979148414e-05,
+      "loss": 1.7656,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8970341905606464,
+      "grad_norm": 0.6767374277114868,
+      "learning_rate": 7.011564745049938e-05,
+      "loss": 1.7193,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8975596938943081,
+      "grad_norm": 0.7432959079742432,
+      "learning_rate": 7.009812510951463e-05,
+      "loss": 1.7336,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8980851972279699,
+      "grad_norm": 0.6830999851226807,
+      "learning_rate": 7.008060276852988e-05,
+      "loss": 1.7438,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8986107005616317,
+      "grad_norm": 0.5865710377693176,
+      "learning_rate": 7.006308042754513e-05,
+      "loss": 1.8067,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8991362038952935,
+      "grad_norm": 1.0097007751464844,
+      "learning_rate": 7.004555808656037e-05,
+      "loss": 1.7458,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8996617072289552,
+      "grad_norm": 0.7260006666183472,
+      "learning_rate": 7.002803574557561e-05,
+      "loss": 1.7707,
+      "step": 3424
+    },
+    {
+      "epoch": 0.900187210562617,
+      "grad_norm": 0.7570676803588867,
+      "learning_rate": 7.001051340459086e-05,
+      "loss": 1.7631,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9007127138962788,
+      "grad_norm": 1.079424500465393,
+      "learning_rate": 6.99929910636061e-05,
+      "loss": 1.7318,
+      "step": 3428
+    },
+    {
+      "epoch": 0.9012382172299406,
+      "grad_norm": 0.637510359287262,
+      "learning_rate": 6.997546872262134e-05,
+      "loss": 1.7427,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9017637205636023,
+      "grad_norm": 0.9443916082382202,
+      "learning_rate": 6.995794638163659e-05,
+      "loss": 1.7182,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9022892238972641,
+      "grad_norm": 0.7777066826820374,
+      "learning_rate": 6.994042404065183e-05,
+      "loss": 1.7543,
+      "step": 3434
+    },
+    {
+      "epoch": 0.9028147272309258,
+      "grad_norm": 0.6093029975891113,
+      "learning_rate": 6.992290169966708e-05,
+      "loss": 1.7501,
+      "step": 3436
+    },
+    {
+      "epoch": 0.9033402305645877,
+      "grad_norm": 0.6203290820121765,
+      "learning_rate": 6.990537935868233e-05,
+      "loss": 1.8313,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9038657338982494,
+      "grad_norm": 0.6452112793922424,
+      "learning_rate": 6.988785701769756e-05,
+      "loss": 1.7691,
+      "step": 3440
+    },
+    {
+      "epoch": 0.9043912372319112,
+      "grad_norm": 0.6220587491989136,
+      "learning_rate": 6.987033467671281e-05,
+      "loss": 1.7458,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9049167405655729,
+      "grad_norm": 0.6492230296134949,
+      "learning_rate": 6.985281233572806e-05,
+      "loss": 1.7624,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9054422438992348,
+      "grad_norm": 0.8058958053588867,
+      "learning_rate": 6.98352899947433e-05,
+      "loss": 1.7341,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9059677472328965,
+      "grad_norm": 0.765034019947052,
+      "learning_rate": 6.981776765375855e-05,
+      "loss": 1.7312,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9064932505665583,
+      "grad_norm": 0.9301319122314453,
+      "learning_rate": 6.98002453127738e-05,
+      "loss": 1.7785,
+      "step": 3450
+    },
+    {
+      "epoch": 0.90701875390022,
+      "grad_norm": 0.723552942276001,
+      "learning_rate": 6.978272297178903e-05,
+      "loss": 1.7547,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9075442572338819,
+      "grad_norm": 0.8970544338226318,
+      "learning_rate": 6.976520063080428e-05,
+      "loss": 1.7424,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9080697605675436,
+      "grad_norm": 0.6986632943153381,
+      "learning_rate": 6.974767828981952e-05,
+      "loss": 1.76,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9085952639012054,
+      "grad_norm": 0.6767635941505432,
+      "learning_rate": 6.973015594883476e-05,
+      "loss": 1.7517,
+      "step": 3458
+    },
+    {
+      "epoch": 0.9091207672348671,
+      "grad_norm": 0.7813493013381958,
+      "learning_rate": 6.971263360785001e-05,
+      "loss": 1.7303,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9096462705685289,
+      "grad_norm": 0.7122093439102173,
+      "learning_rate": 6.969511126686526e-05,
+      "loss": 1.7735,
+      "step": 3462
+    },
+    {
+      "epoch": 0.9101717739021907,
+      "grad_norm": 0.9538240432739258,
+      "learning_rate": 6.96775889258805e-05,
+      "loss": 1.7279,
+      "step": 3464
+    },
+    {
+      "epoch": 0.9106972772358525,
+      "grad_norm": 0.9474038481712341,
+      "learning_rate": 6.966006658489574e-05,
+      "loss": 1.7636,
+      "step": 3466
+    },
+    {
+      "epoch": 0.9112227805695142,
+      "grad_norm": 0.6011683344841003,
+      "learning_rate": 6.964254424391099e-05,
+      "loss": 1.7414,
+      "step": 3468
+    },
+    {
+      "epoch": 0.911748283903176,
+      "grad_norm": 0.7591129541397095,
+      "learning_rate": 6.962502190292623e-05,
+      "loss": 1.7654,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9122737872368378,
+      "grad_norm": 0.8848958015441895,
+      "learning_rate": 6.960749956194148e-05,
+      "loss": 1.7581,
+      "step": 3472
+    },
+    {
+      "epoch": 0.9127992905704996,
+      "grad_norm": 0.7266464233398438,
+      "learning_rate": 6.958997722095673e-05,
+      "loss": 1.769,
+      "step": 3474
+    },
+    {
+      "epoch": 0.9133247939041613,
+      "grad_norm": 0.8937695026397705,
+      "learning_rate": 6.957245487997198e-05,
+      "loss": 1.7574,
+      "step": 3476
+    },
+    {
+      "epoch": 0.9138502972378231,
+      "grad_norm": 0.6543369293212891,
+      "learning_rate": 6.955493253898721e-05,
+      "loss": 1.7587,
+      "step": 3478
+    },
+    {
+      "epoch": 0.9143758005714848,
+      "grad_norm": 1.0419197082519531,
+      "learning_rate": 6.953741019800246e-05,
+      "loss": 1.7311,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9149013039051467,
+      "grad_norm": 0.6583675742149353,
+      "learning_rate": 6.951988785701769e-05,
+      "loss": 1.7415,
+      "step": 3482
+    },
+    {
+      "epoch": 0.9154268072388084,
+      "grad_norm": 0.7123555541038513,
+      "learning_rate": 6.950236551603294e-05,
+      "loss": 1.7227,
+      "step": 3484
+    },
+    {
+      "epoch": 0.9159523105724702,
+      "grad_norm": 0.606636106967926,
+      "learning_rate": 6.948484317504819e-05,
+      "loss": 1.7555,
+      "step": 3486
+    },
+    {
+      "epoch": 0.9164778139061319,
+      "grad_norm": 1.0369200706481934,
+      "learning_rate": 6.946732083406344e-05,
+      "loss": 1.7494,
+      "step": 3488
+    },
+    {
+      "epoch": 0.9170033172397938,
+      "grad_norm": 0.6828787922859192,
+      "learning_rate": 6.944979849307868e-05,
+      "loss": 1.7302,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9175288205734555,
+      "grad_norm": 0.7840218544006348,
+      "learning_rate": 6.943227615209392e-05,
+      "loss": 1.734,
+      "step": 3492
+    },
+    {
+      "epoch": 0.9180543239071173,
+      "grad_norm": 0.6639379858970642,
+      "learning_rate": 6.941475381110916e-05,
+      "loss": 1.7197,
+      "step": 3494
+    },
+    {
+      "epoch": 0.918579827240779,
+      "grad_norm": 0.6590544581413269,
+      "learning_rate": 6.939723147012441e-05,
+      "loss": 1.758,
+      "step": 3496
+    },
+    {
+      "epoch": 0.9191053305744409,
+      "grad_norm": 0.5985316038131714,
+      "learning_rate": 6.937970912913966e-05,
+      "loss": 1.7545,
+      "step": 3498
+    },
+    {
+      "epoch": 0.9196308339081026,
+      "grad_norm": 0.6269810795783997,
+      "learning_rate": 6.936218678815491e-05,
+      "loss": 1.7476,
+      "step": 3500
+    },
+    {
+      "epoch": 0.9201563372417644,
+      "grad_norm": 0.7149941921234131,
+      "learning_rate": 6.934466444717016e-05,
+      "loss": 1.7629,
+      "step": 3502
+    },
+    {
+      "epoch": 0.9206818405754261,
+      "grad_norm": 0.641220211982727,
+      "learning_rate": 6.932714210618539e-05,
+      "loss": 1.7613,
+      "step": 3504
+    },
+    {
+      "epoch": 0.9212073439090879,
+      "grad_norm": 0.7803055644035339,
+      "learning_rate": 6.930961976520064e-05,
+      "loss": 1.7401,
+      "step": 3506
+    },
+    {
+      "epoch": 0.9217328472427497,
+      "grad_norm": 0.6284143924713135,
+      "learning_rate": 6.929209742421587e-05,
+      "loss": 1.7267,
+      "step": 3508
+    },
+    {
+      "epoch": 0.9222583505764115,
+      "grad_norm": 0.594203531742096,
+      "learning_rate": 6.927457508323112e-05,
+      "loss": 1.7553,
+      "step": 3510
+    },
+    {
+      "epoch": 0.9227838539100732,
+      "grad_norm": 0.7015509009361267,
+      "learning_rate": 6.925705274224637e-05,
+      "loss": 1.7122,
+      "step": 3512
+    },
+    {
+      "epoch": 0.923309357243735,
+      "grad_norm": 0.6864806413650513,
+      "learning_rate": 6.923953040126161e-05,
+      "loss": 1.7099,
+      "step": 3514
+    },
+    {
+      "epoch": 0.9238348605773968,
+      "grad_norm": 0.7701146602630615,
+      "learning_rate": 6.922200806027686e-05,
+      "loss": 1.7508,
+      "step": 3516
+    },
+    {
+      "epoch": 0.9243603639110586,
+      "grad_norm": 0.6888076663017273,
+      "learning_rate": 6.92044857192921e-05,
+      "loss": 1.7455,
+      "step": 3518
+    },
+    {
+      "epoch": 0.9248858672447203,
+      "grad_norm": 0.6871370077133179,
+      "learning_rate": 6.918696337830734e-05,
+      "loss": 1.7597,
+      "step": 3520
+    },
+    {
+      "epoch": 0.9254113705783821,
+      "grad_norm": 1.1456379890441895,
+      "learning_rate": 6.916944103732259e-05,
+      "loss": 1.7585,
+      "step": 3522
+    },
+    {
+      "epoch": 0.9259368739120439,
+      "grad_norm": 0.6293717622756958,
+      "learning_rate": 6.915191869633784e-05,
+      "loss": 1.7279,
+      "step": 3524
+    },
+    {
+      "epoch": 0.9264623772457057,
+      "grad_norm": 0.6523435711860657,
+      "learning_rate": 6.913439635535309e-05,
+      "loss": 1.7486,
+      "step": 3526
+    },
+    {
+      "epoch": 0.9269878805793674,
+      "grad_norm": 0.6591430306434631,
+      "learning_rate": 6.911687401436833e-05,
+      "loss": 1.7544,
+      "step": 3528
+    },
+    {
+      "epoch": 0.9275133839130292,
+      "grad_norm": 0.7993916273117065,
+      "learning_rate": 6.909935167338357e-05,
+      "loss": 1.7468,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9280388872466909,
+      "grad_norm": 0.5962069630622864,
+      "learning_rate": 6.90818293323988e-05,
+      "loss": 1.73,
+      "step": 3532
+    },
+    {
+      "epoch": 0.9285643905803528,
+      "grad_norm": 0.7057903409004211,
+      "learning_rate": 6.906430699141405e-05,
+      "loss": 1.7613,
+      "step": 3534
+    },
+    {
+      "epoch": 0.9290898939140145,
+      "grad_norm": 0.8017992377281189,
+      "learning_rate": 6.90467846504293e-05,
+      "loss": 1.7576,
+      "step": 3536
+    },
+    {
+      "epoch": 0.9296153972476763,
+      "grad_norm": 0.717413067817688,
+      "learning_rate": 6.902926230944454e-05,
+      "loss": 1.6982,
+      "step": 3538
+    },
+    {
+      "epoch": 0.930140900581338,
+      "grad_norm": 0.7504727840423584,
+      "learning_rate": 6.901173996845979e-05,
+      "loss": 1.7657,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9306664039149999,
+      "grad_norm": 1.548189401626587,
+      "learning_rate": 6.899421762747504e-05,
+      "loss": 1.7614,
+      "step": 3542
+    },
+    {
+      "epoch": 0.9311919072486616,
+      "grad_norm": 0.7025728821754456,
+      "learning_rate": 6.897669528649027e-05,
+      "loss": 1.748,
+      "step": 3544
+    },
+    {
+      "epoch": 0.9317174105823234,
+      "grad_norm": 0.8109622001647949,
+      "learning_rate": 6.895917294550552e-05,
+      "loss": 1.7346,
+      "step": 3546
+    },
+    {
+      "epoch": 0.9322429139159851,
+      "grad_norm": 0.6339906454086304,
+      "learning_rate": 6.894165060452077e-05,
+      "loss": 1.7594,
+      "step": 3548
+    },
+    {
+      "epoch": 0.9327684172496469,
+      "grad_norm": 0.7020452618598938,
+      "learning_rate": 6.892412826353602e-05,
+      "loss": 1.7461,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9332939205833087,
+      "grad_norm": 0.755463182926178,
+      "learning_rate": 6.890660592255126e-05,
+      "loss": 1.7522,
+      "step": 3552
+    },
+    {
+      "epoch": 0.9338194239169705,
+      "grad_norm": 0.8311240673065186,
+      "learning_rate": 6.888908358156651e-05,
+      "loss": 1.7438,
+      "step": 3554
+    },
+    {
+      "epoch": 0.9343449272506322,
+      "grad_norm": 0.8520714044570923,
+      "learning_rate": 6.887156124058174e-05,
+      "loss": 1.7354,
+      "step": 3556
+    },
+    {
+      "epoch": 0.934870430584294,
+      "grad_norm": 0.5846522450447083,
+      "learning_rate": 6.885403889959698e-05,
+      "loss": 1.7784,
+      "step": 3558
+    },
+    {
+      "epoch": 0.9353959339179558,
+      "grad_norm": 0.6541776657104492,
+      "learning_rate": 6.883651655861223e-05,
+      "loss": 1.7437,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9359214372516176,
+      "grad_norm": 0.8012898564338684,
+      "learning_rate": 6.881899421762747e-05,
+      "loss": 1.7559,
+      "step": 3562
+    },
+    {
+      "epoch": 0.9364469405852793,
+      "grad_norm": 0.6521446704864502,
+      "learning_rate": 6.880147187664272e-05,
+      "loss": 1.7359,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9369724439189411,
+      "grad_norm": 0.6322072744369507,
+      "learning_rate": 6.878394953565797e-05,
+      "loss": 1.7682,
+      "step": 3566
+    },
+    {
+      "epoch": 0.9374979472526029,
+      "grad_norm": 0.8116897940635681,
+      "learning_rate": 6.876642719467322e-05,
+      "loss": 1.7303,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9380234505862647,
+      "grad_norm": 0.6734224557876587,
+      "learning_rate": 6.874890485368845e-05,
+      "loss": 1.7101,
+      "step": 3570
+    },
+    {
+      "epoch": 0.9385489539199264,
+      "grad_norm": 0.7724919319152832,
+      "learning_rate": 6.87313825127037e-05,
+      "loss": 1.754,
+      "step": 3572
+    },
+    {
+      "epoch": 0.9390744572535882,
+      "grad_norm": 0.6816089749336243,
+      "learning_rate": 6.871386017171895e-05,
+      "loss": 1.7592,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9395999605872499,
+      "grad_norm": 0.6413043737411499,
+      "learning_rate": 6.869633783073419e-05,
+      "loss": 1.7446,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9401254639209118,
+      "grad_norm": 0.8377385139465332,
+      "learning_rate": 6.867881548974944e-05,
+      "loss": 1.7528,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9406509672545735,
+      "grad_norm": 0.7805231809616089,
+      "learning_rate": 6.866129314876469e-05,
+      "loss": 1.8016,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.5867004990577698,
+      "learning_rate": 6.864377080777992e-05,
+      "loss": 1.7655,
+      "step": 3582
+    },
+    {
+      "epoch": 0.941701973921897,
+      "grad_norm": 0.6897924542427063,
+      "learning_rate": 6.862624846679516e-05,
+      "loss": 1.7547,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9422274772555589,
+      "grad_norm": 0.8123984336853027,
+      "learning_rate": 6.86087261258104e-05,
+      "loss": 1.7478,
+      "step": 3586
+    },
+    {
+      "epoch": 0.9427529805892206,
+      "grad_norm": 0.6391046643257141,
+      "learning_rate": 6.859120378482565e-05,
+      "loss": 1.7409,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9432784839228824,
+      "grad_norm": 0.7633985280990601,
+      "learning_rate": 6.85736814438409e-05,
+      "loss": 1.742,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9438039872565441,
+      "grad_norm": 0.6767701506614685,
+      "learning_rate": 6.855615910285615e-05,
+      "loss": 1.723,
+      "step": 3592
+    },
+    {
+      "epoch": 0.9443294905902059,
+      "grad_norm": 0.7079693675041199,
+      "learning_rate": 6.85386367618714e-05,
+      "loss": 1.7801,
+      "step": 3594
+    },
+    {
+      "epoch": 0.9448549939238677,
+      "grad_norm": 0.6894028782844543,
+      "learning_rate": 6.852111442088663e-05,
+      "loss": 1.7508,
+      "step": 3596
+    },
+    {
+      "epoch": 0.9453804972575295,
+      "grad_norm": 0.6596964597702026,
+      "learning_rate": 6.850359207990188e-05,
+      "loss": 1.7529,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9459060005911912,
+      "grad_norm": 0.6702237725257874,
+      "learning_rate": 6.848606973891712e-05,
+      "loss": 1.7714,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9459060005911912,
+      "eval_loss": 1.7171640396118164,
+      "eval_runtime": 487.1544,
+      "eval_samples_per_second": 250.001,
+      "eval_steps_per_second": 31.251,
+      "step": 3600
+    },
+    {
+      "epoch": 0.946431503924853,
+      "grad_norm": 0.5935298800468445,
+      "learning_rate": 6.846854739793237e-05,
+      "loss": 1.7409,
+      "step": 3602
+    },
+    {
+      "epoch": 0.9469570072585148,
+      "grad_norm": 0.844529926776886,
+      "learning_rate": 6.845102505694762e-05,
+      "loss": 1.7598,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9474825105921766,
+      "grad_norm": 0.6452075839042664,
+      "learning_rate": 6.843350271596287e-05,
+      "loss": 1.7641,
+      "step": 3606
+    },
+    {
+      "epoch": 0.9480080139258383,
+      "grad_norm": 0.7184598445892334,
+      "learning_rate": 6.84159803749781e-05,
+      "loss": 1.7575,
+      "step": 3608
+    },
+    {
+      "epoch": 0.9485335172595001,
+      "grad_norm": 0.6628120541572571,
+      "learning_rate": 6.839845803399333e-05,
+      "loss": 1.7427,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9490590205931619,
+      "grad_norm": 0.6025474667549133,
+      "learning_rate": 6.838093569300858e-05,
+      "loss": 1.7132,
+      "step": 3612
+    },
+    {
+      "epoch": 0.9495845239268237,
+      "grad_norm": 0.6190858483314514,
+      "learning_rate": 6.836341335202383e-05,
+      "loss": 1.7773,
+      "step": 3614
+    },
+    {
+      "epoch": 0.9501100272604854,
+      "grad_norm": 0.6773670315742493,
+      "learning_rate": 6.834589101103908e-05,
+      "loss": 1.7139,
+      "step": 3616
+    },
+    {
+      "epoch": 0.9506355305941472,
+      "grad_norm": 0.6356403827667236,
+      "learning_rate": 6.832836867005432e-05,
+      "loss": 1.7366,
+      "step": 3618
+    },
+    {
+      "epoch": 0.9511610339278089,
+      "grad_norm": 0.7546253800392151,
+      "learning_rate": 6.831084632906957e-05,
+      "loss": 1.7313,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9516865372614708,
+      "grad_norm": 0.6066844463348389,
+      "learning_rate": 6.82933239880848e-05,
+      "loss": 1.7141,
+      "step": 3622
+    },
+    {
+      "epoch": 0.9522120405951325,
+      "grad_norm": 0.5842781066894531,
+      "learning_rate": 6.827580164710005e-05,
+      "loss": 1.7677,
+      "step": 3624
+    },
+    {
+      "epoch": 0.9527375439287943,
+      "grad_norm": 0.6750701069831848,
+      "learning_rate": 6.82582793061153e-05,
+      "loss": 1.7065,
+      "step": 3626
+    },
+    {
+      "epoch": 0.953263047262456,
+      "grad_norm": 0.5796899199485779,
+      "learning_rate": 6.824075696513055e-05,
+      "loss": 1.7272,
+      "step": 3628
+    },
+    {
+      "epoch": 0.9537885505961179,
+      "grad_norm": 0.7888158559799194,
+      "learning_rate": 6.82232346241458e-05,
+      "loss": 1.7441,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9543140539297796,
+      "grad_norm": 0.5940207839012146,
+      "learning_rate": 6.820571228316104e-05,
+      "loss": 1.7261,
+      "step": 3632
+    },
+    {
+      "epoch": 0.9548395572634414,
+      "grad_norm": 0.7521408200263977,
+      "learning_rate": 6.818818994217628e-05,
+      "loss": 1.7826,
+      "step": 3634
+    },
+    {
+      "epoch": 0.9553650605971031,
+      "grad_norm": 0.6173054575920105,
+      "learning_rate": 6.817066760119151e-05,
+      "loss": 1.7113,
+      "step": 3636
+    },
+    {
+      "epoch": 0.9558905639307649,
+      "grad_norm": 0.6263679265975952,
+      "learning_rate": 6.815314526020676e-05,
+      "loss": 1.6943,
+      "step": 3638
+    },
+    {
+      "epoch": 0.9564160672644267,
+      "grad_norm": 0.6220889687538147,
+      "learning_rate": 6.8135622919222e-05,
+      "loss": 1.7146,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9569415705980885,
+      "grad_norm": 0.6978549957275391,
+      "learning_rate": 6.811810057823725e-05,
+      "loss": 1.791,
+      "step": 3642
+    },
+    {
+      "epoch": 0.9574670739317502,
+      "grad_norm": 0.6536712050437927,
+      "learning_rate": 6.81005782372525e-05,
+      "loss": 1.7397,
+      "step": 3644
+    },
+    {
+      "epoch": 0.957992577265412,
+      "grad_norm": 0.6998410820960999,
+      "learning_rate": 6.808305589626775e-05,
+      "loss": 1.736,
+      "step": 3646
+    },
+    {
+      "epoch": 0.9585180805990738,
+      "grad_norm": 0.6182027459144592,
+      "learning_rate": 6.806553355528298e-05,
+      "loss": 1.7462,
+      "step": 3648
+    },
+    {
+      "epoch": 0.9590435839327356,
+      "grad_norm": 0.7864842414855957,
+      "learning_rate": 6.804801121429823e-05,
+      "loss": 1.7495,
+      "step": 3650
+    },
+    {
+      "epoch": 0.9595690872663973,
+      "grad_norm": 0.6455209255218506,
+      "learning_rate": 6.803048887331348e-05,
+      "loss": 1.7324,
+      "step": 3652
+    },
+    {
+      "epoch": 0.9600945906000591,
+      "grad_norm": 0.6225829720497131,
+      "learning_rate": 6.801296653232873e-05,
+      "loss": 1.748,
+      "step": 3654
+    },
+    {
+      "epoch": 0.9606200939337209,
+      "grad_norm": 0.6527931094169617,
+      "learning_rate": 6.799544419134397e-05,
+      "loss": 1.7133,
+      "step": 3656
+    },
+    {
+      "epoch": 0.9611455972673827,
+      "grad_norm": 0.619773805141449,
+      "learning_rate": 6.797792185035922e-05,
+      "loss": 1.7876,
+      "step": 3658
+    },
+    {
+      "epoch": 0.9616711006010444,
+      "grad_norm": 0.681759774684906,
+      "learning_rate": 6.796039950937446e-05,
+      "loss": 1.7069,
+      "step": 3660
+    },
+    {
+      "epoch": 0.9621966039347062,
+      "grad_norm": 0.746255099773407,
+      "learning_rate": 6.794287716838969e-05,
+      "loss": 1.7784,
+      "step": 3662
+    },
+    {
+      "epoch": 0.9627221072683679,
+      "grad_norm": 0.5940551161766052,
+      "learning_rate": 6.792535482740494e-05,
+      "loss": 1.7495,
+      "step": 3664
+    },
+    {
+      "epoch": 0.9632476106020298,
+      "grad_norm": 0.6246922016143799,
+      "learning_rate": 6.790783248642018e-05,
+      "loss": 1.7758,
+      "step": 3666
+    },
+    {
+      "epoch": 0.9637731139356915,
+      "grad_norm": 0.6583105325698853,
+      "learning_rate": 6.789031014543543e-05,
+      "loss": 1.7431,
+      "step": 3668
+    },
+    {
+      "epoch": 0.9642986172693533,
+      "grad_norm": 0.6988399624824524,
+      "learning_rate": 6.787278780445068e-05,
+      "loss": 1.7269,
+      "step": 3670
+    },
+    {
+      "epoch": 0.964824120603015,
+      "grad_norm": 0.60069739818573,
+      "learning_rate": 6.785526546346593e-05,
+      "loss": 1.7466,
+      "step": 3672
+    },
+    {
+      "epoch": 0.9653496239366769,
+      "grad_norm": 0.615967869758606,
+      "learning_rate": 6.783774312248116e-05,
+      "loss": 1.7208,
+      "step": 3674
+    },
+    {
+      "epoch": 0.9658751272703386,
+      "grad_norm": 0.7129417061805725,
+      "learning_rate": 6.782022078149641e-05,
+      "loss": 1.7441,
+      "step": 3676
+    },
+    {
+      "epoch": 0.9664006306040004,
+      "grad_norm": 0.6287668347358704,
+      "learning_rate": 6.780269844051166e-05,
+      "loss": 1.7413,
+      "step": 3678
+    },
+    {
+      "epoch": 0.9669261339376621,
+      "grad_norm": 0.7385637760162354,
+      "learning_rate": 6.77851760995269e-05,
+      "loss": 1.7569,
+      "step": 3680
+    },
+    {
+      "epoch": 0.967451637271324,
+      "grad_norm": 0.6797763109207153,
+      "learning_rate": 6.776765375854215e-05,
+      "loss": 1.7417,
+      "step": 3682
+    },
+    {
+      "epoch": 0.9679771406049857,
+      "grad_norm": 0.667385458946228,
+      "learning_rate": 6.77501314175574e-05,
+      "loss": 1.7554,
+      "step": 3684
+    },
+    {
+      "epoch": 0.9685026439386475,
+      "grad_norm": 0.7547754645347595,
+      "learning_rate": 6.773260907657263e-05,
+      "loss": 1.6905,
+      "step": 3686
+    },
+    {
+      "epoch": 0.9690281472723092,
+      "grad_norm": 0.6124489903450012,
+      "learning_rate": 6.771508673558788e-05,
+      "loss": 1.7276,
+      "step": 3688
+    },
+    {
+      "epoch": 0.969553650605971,
+      "grad_norm": 0.6314408779144287,
+      "learning_rate": 6.769756439460311e-05,
+      "loss": 1.7375,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9700791539396328,
+      "grad_norm": 1.1786826848983765,
+      "learning_rate": 6.768004205361836e-05,
+      "loss": 1.7797,
+      "step": 3692
+    },
+    {
+      "epoch": 0.9706046572732946,
+      "grad_norm": 0.6495144963264465,
+      "learning_rate": 6.766251971263361e-05,
+      "loss": 1.7767,
+      "step": 3694
+    },
+    {
+      "epoch": 0.9711301606069563,
+      "grad_norm": 0.8259857892990112,
+      "learning_rate": 6.764499737164886e-05,
+      "loss": 1.7568,
+      "step": 3696
+    },
+    {
+      "epoch": 0.9716556639406181,
+      "grad_norm": 0.9718241095542908,
+      "learning_rate": 6.76274750306641e-05,
+      "loss": 1.7382,
+      "step": 3698
+    },
+    {
+      "epoch": 0.97218116727428,
+      "grad_norm": 0.7467637062072754,
+      "learning_rate": 6.760995268967935e-05,
+      "loss": 1.7168,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9727066706079417,
+      "grad_norm": 0.6963891386985779,
+      "learning_rate": 6.759243034869459e-05,
+      "loss": 1.75,
+      "step": 3702
+    },
+    {
+      "epoch": 0.9732321739416034,
+      "grad_norm": 0.5882383584976196,
+      "learning_rate": 6.757490800770983e-05,
+      "loss": 1.7664,
+      "step": 3704
+    },
+    {
+      "epoch": 0.9737576772752652,
+      "grad_norm": 0.5909221768379211,
+      "learning_rate": 6.755738566672508e-05,
+      "loss": 1.7569,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9742831806089269,
+      "grad_norm": 0.6651691794395447,
+      "learning_rate": 6.753986332574033e-05,
+      "loss": 1.7332,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9748086839425888,
+      "grad_norm": 0.6473085880279541,
+      "learning_rate": 6.752234098475558e-05,
+      "loss": 1.7668,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9753341872762505,
+      "grad_norm": 0.6437013745307922,
+      "learning_rate": 6.750481864377081e-05,
+      "loss": 1.747,
+      "step": 3712
+    },
+    {
+      "epoch": 0.9758596906099123,
+      "grad_norm": 0.6409528255462646,
+      "learning_rate": 6.748729630278606e-05,
+      "loss": 1.7105,
+      "step": 3714
+    },
+    {
+      "epoch": 0.976385193943574,
+      "grad_norm": 0.7331600785255432,
+      "learning_rate": 6.746977396180129e-05,
+      "loss": 1.7328,
+      "step": 3716
+    },
+    {
+      "epoch": 0.9769106972772359,
+      "grad_norm": 0.7643489837646484,
+      "learning_rate": 6.745225162081654e-05,
+      "loss": 1.7464,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9774362006108976,
+      "grad_norm": 0.8822628259658813,
+      "learning_rate": 6.743472927983179e-05,
+      "loss": 1.7681,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9779617039445594,
+      "grad_norm": 0.9610887765884399,
+      "learning_rate": 6.741720693884703e-05,
+      "loss": 1.7329,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9784872072782211,
+      "grad_norm": 0.7626636028289795,
+      "learning_rate": 6.739968459786228e-05,
+      "loss": 1.7161,
+      "step": 3724
+    },
+    {
+      "epoch": 0.979012710611883,
+      "grad_norm": 0.6538355946540833,
+      "learning_rate": 6.738216225687753e-05,
+      "loss": 1.7716,
+      "step": 3726
+    },
+    {
+      "epoch": 0.9795382139455447,
+      "grad_norm": 0.7273457050323486,
+      "learning_rate": 6.736463991589276e-05,
+      "loss": 1.7367,
+      "step": 3728
+    },
+    {
+      "epoch": 0.9800637172792065,
+      "grad_norm": 0.7992196679115295,
+      "learning_rate": 6.734711757490801e-05,
+      "loss": 1.7294,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9805892206128682,
+      "grad_norm": 0.6318385601043701,
+      "learning_rate": 6.732959523392326e-05,
+      "loss": 1.7498,
+      "step": 3732
+    },
+    {
+      "epoch": 0.98111472394653,
+      "grad_norm": 0.8519952893257141,
+      "learning_rate": 6.73120728929385e-05,
+      "loss": 1.7366,
+      "step": 3734
+    },
+    {
+      "epoch": 0.9816402272801918,
+      "grad_norm": 0.7661817073822021,
+      "learning_rate": 6.729455055195374e-05,
+      "loss": 1.7565,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9821657306138536,
+      "grad_norm": 0.5940839052200317,
+      "learning_rate": 6.727702821096899e-05,
+      "loss": 1.7384,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9826912339475153,
+      "grad_norm": 0.6861841082572937,
+      "learning_rate": 6.725950586998424e-05,
+      "loss": 1.7129,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9832167372811771,
+      "grad_norm": 0.9989137649536133,
+      "learning_rate": 6.724198352899947e-05,
+      "loss": 1.7758,
+      "step": 3742
+    },
+    {
+      "epoch": 0.983742240614839,
+      "grad_norm": 0.6522616147994995,
+      "learning_rate": 6.722446118801472e-05,
+      "loss": 1.7114,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9842677439485007,
+      "grad_norm": 0.6522443294525146,
+      "learning_rate": 6.720693884702996e-05,
+      "loss": 1.7375,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9847932472821624,
+      "grad_norm": 0.5641542077064514,
+      "learning_rate": 6.718941650604521e-05,
+      "loss": 1.7371,
+      "step": 3748
+    },
+    {
+      "epoch": 0.9853187506158242,
+      "grad_norm": 0.5788604021072388,
+      "learning_rate": 6.717189416506046e-05,
+      "loss": 1.7184,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9858442539494859,
+      "grad_norm": 0.5921624302864075,
+      "learning_rate": 6.715437182407571e-05,
+      "loss": 1.7304,
+      "step": 3752
+    },
+    {
+      "epoch": 0.9863697572831478,
+      "grad_norm": 0.6481342911720276,
+      "learning_rate": 6.713684948309094e-05,
+      "loss": 1.7838,
+      "step": 3754
+    },
+    {
+      "epoch": 0.9868952606168095,
+      "grad_norm": 0.6901116967201233,
+      "learning_rate": 6.711932714210619e-05,
+      "loss": 1.7189,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9874207639504713,
+      "grad_norm": 0.7430614233016968,
+      "learning_rate": 6.710180480112144e-05,
+      "loss": 1.7532,
+      "step": 3758
+    },
+    {
+      "epoch": 0.987946267284133,
+      "grad_norm": 0.8305982947349548,
+      "learning_rate": 6.708428246013668e-05,
+      "loss": 1.7417,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9884717706177949,
+      "grad_norm": 0.801034152507782,
+      "learning_rate": 6.706676011915192e-05,
+      "loss": 1.7058,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9889972739514566,
+      "grad_norm": 0.6561713218688965,
+      "learning_rate": 6.704923777816717e-05,
+      "loss": 1.7347,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9895227772851184,
+      "grad_norm": 0.6591217517852783,
+      "learning_rate": 6.703171543718241e-05,
+      "loss": 1.7693,
+      "step": 3766
+    },
+    {
+      "epoch": 0.9900482806187801,
+      "grad_norm": 0.6448448300361633,
+      "learning_rate": 6.701419309619765e-05,
+      "loss": 1.7327,
+      "step": 3768
+    },
+    {
+      "epoch": 0.990573783952442,
+      "grad_norm": 0.7278120517730713,
+      "learning_rate": 6.69966707552129e-05,
+      "loss": 1.7049,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9910992872861037,
+      "grad_norm": 0.5780648589134216,
+      "learning_rate": 6.697914841422814e-05,
+      "loss": 1.7643,
+      "step": 3772
+    },
+    {
+      "epoch": 0.9916247906197655,
+      "grad_norm": 0.599757969379425,
+      "learning_rate": 6.696162607324339e-05,
+      "loss": 1.7419,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9921502939534272,
+      "grad_norm": 0.6809077858924866,
+      "learning_rate": 6.694410373225864e-05,
+      "loss": 1.7638,
+      "step": 3776
+    },
+    {
+      "epoch": 0.992675797287089,
+      "grad_norm": 0.7125533223152161,
+      "learning_rate": 6.692658139127389e-05,
+      "loss": 1.7598,
+      "step": 3778
+    },
+    {
+      "epoch": 0.9932013006207508,
+      "grad_norm": 0.7388641834259033,
+      "learning_rate": 6.690905905028912e-05,
+      "loss": 1.7484,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9937268039544126,
+      "grad_norm": 0.622369647026062,
+      "learning_rate": 6.689153670930437e-05,
+      "loss": 1.7587,
+      "step": 3782
+    },
+    {
+      "epoch": 0.9942523072880743,
+      "grad_norm": 0.7785460948944092,
+      "learning_rate": 6.687401436831961e-05,
+      "loss": 1.7575,
+      "step": 3784
+    },
+    {
+      "epoch": 0.9947778106217361,
+      "grad_norm": 0.6789509057998657,
+      "learning_rate": 6.685649202733486e-05,
+      "loss": 1.7346,
+      "step": 3786
+    },
+    {
+      "epoch": 0.995303313955398,
+      "grad_norm": 0.6360666155815125,
+      "learning_rate": 6.68389696863501e-05,
+      "loss": 1.7838,
+      "step": 3788
+    },
+    {
+      "epoch": 0.9958288172890597,
+      "grad_norm": 0.6261754035949707,
+      "learning_rate": 6.682144734536534e-05,
+      "loss": 1.7563,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9963543206227214,
+      "grad_norm": 0.6586030125617981,
+      "learning_rate": 6.680392500438059e-05,
+      "loss": 1.7034,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9968798239563832,
+      "grad_norm": 0.7084933519363403,
+      "learning_rate": 6.678640266339583e-05,
+      "loss": 1.7995,
+      "step": 3794
+    },
+    {
+      "epoch": 0.9974053272900449,
+      "grad_norm": 0.8571730256080627,
+      "learning_rate": 6.676888032241107e-05,
+      "loss": 1.7874,
+      "step": 3796
+    },
+    {
+      "epoch": 0.9979308306237068,
+      "grad_norm": 0.6537512540817261,
+      "learning_rate": 6.675135798142632e-05,
+      "loss": 1.7672,
+      "step": 3798
+    },
+    {
+      "epoch": 0.9984563339573685,
+      "grad_norm": 0.6276223659515381,
+      "learning_rate": 6.673383564044157e-05,
+      "loss": 1.7106,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9989818372910303,
+      "grad_norm": 0.6775636076927185,
+      "learning_rate": 6.671631329945682e-05,
+      "loss": 1.7228,
+      "step": 3802
+    },
+    {
+      "epoch": 0.999507340624692,
+      "grad_norm": 0.776142954826355,
+      "learning_rate": 6.669879095847206e-05,
+      "loss": 1.7674,
+      "step": 3804
+    },
+    {
+      "epoch": 1.000032843958354,
+      "grad_norm": 0.6469699740409851,
+      "learning_rate": 6.66812686174873e-05,
+      "loss": 1.7201,
+      "step": 3806
+    },
+    {
+      "epoch": 1.0005583472920156,
+      "grad_norm": 0.6889809966087341,
+      "learning_rate": 6.666374627650254e-05,
+      "loss": 1.6785,
+      "step": 3808
+    },
+    {
+      "epoch": 1.0010838506256774,
+      "grad_norm": 0.5873193740844727,
+      "learning_rate": 6.664622393551779e-05,
+      "loss": 1.6643,
+      "step": 3810
+    },
+    {
+      "epoch": 1.0016093539593391,
+      "grad_norm": 0.6216626167297363,
+      "learning_rate": 6.662870159453303e-05,
+      "loss": 1.6835,
+      "step": 3812
+    },
+    {
+      "epoch": 1.002134857293001,
+      "grad_norm": 0.621909499168396,
+      "learning_rate": 6.661117925354827e-05,
+      "loss": 1.6972,
+      "step": 3814
+    },
+    {
+      "epoch": 1.0026603606266626,
+      "grad_norm": 0.6056811213493347,
+      "learning_rate": 6.659365691256352e-05,
+      "loss": 1.7327,
+      "step": 3816
+    },
+    {
+      "epoch": 1.0031858639603246,
+      "grad_norm": 0.6430692076683044,
+      "learning_rate": 6.657613457157877e-05,
+      "loss": 1.7151,
+      "step": 3818
+    },
+    {
+      "epoch": 1.0037113672939864,
+      "grad_norm": 0.6576054692268372,
+      "learning_rate": 6.6558612230594e-05,
+      "loss": 1.7058,
+      "step": 3820
+    },
+    {
+      "epoch": 1.004236870627648,
+      "grad_norm": 0.666572093963623,
+      "learning_rate": 6.654108988960925e-05,
+      "loss": 1.6745,
+      "step": 3822
+    },
+    {
+      "epoch": 1.0047623739613099,
+      "grad_norm": 0.873776376247406,
+      "learning_rate": 6.65235675486245e-05,
+      "loss": 1.7123,
+      "step": 3824
+    },
+    {
+      "epoch": 1.0052878772949716,
+      "grad_norm": 0.6568595767021179,
+      "learning_rate": 6.650604520763975e-05,
+      "loss": 1.6813,
+      "step": 3826
+    },
+    {
+      "epoch": 1.0058133806286333,
+      "grad_norm": 0.7071711421012878,
+      "learning_rate": 6.648852286665499e-05,
+      "loss": 1.6926,
+      "step": 3828
+    },
+    {
+      "epoch": 1.006338883962295,
+      "grad_norm": 0.756188154220581,
+      "learning_rate": 6.647100052567024e-05,
+      "loss": 1.6939,
+      "step": 3830
+    },
+    {
+      "epoch": 1.0068643872959568,
+      "grad_norm": 0.6261985301971436,
+      "learning_rate": 6.645347818468547e-05,
+      "loss": 1.6729,
+      "step": 3832
+    },
+    {
+      "epoch": 1.0073898906296186,
+      "grad_norm": 0.6476467847824097,
+      "learning_rate": 6.643595584370072e-05,
+      "loss": 1.6905,
+      "step": 3834
+    },
+    {
+      "epoch": 1.0079153939632806,
+      "grad_norm": 0.7380629777908325,
+      "learning_rate": 6.641843350271597e-05,
+      "loss": 1.6695,
+      "step": 3836
+    },
+    {
+      "epoch": 1.0084408972969423,
+      "grad_norm": 0.7090455293655396,
+      "learning_rate": 6.64009111617312e-05,
+      "loss": 1.7271,
+      "step": 3838
+    },
+    {
+      "epoch": 1.008966400630604,
+      "grad_norm": 0.5697006583213806,
+      "learning_rate": 6.638338882074645e-05,
+      "loss": 1.6869,
+      "step": 3840
+    },
+    {
+      "epoch": 1.0094919039642658,
+      "grad_norm": 0.5765069127082825,
+      "learning_rate": 6.63658664797617e-05,
+      "loss": 1.7133,
+      "step": 3842
+    },
+    {
+      "epoch": 1.0100174072979275,
+      "grad_norm": 0.7223833799362183,
+      "learning_rate": 6.634834413877695e-05,
+      "loss": 1.6566,
+      "step": 3844
+    },
+    {
+      "epoch": 1.0105429106315893,
+      "grad_norm": 0.6132098436355591,
+      "learning_rate": 6.633082179779218e-05,
+      "loss": 1.7448,
+      "step": 3846
+    },
+    {
+      "epoch": 1.011068413965251,
+      "grad_norm": 0.8053067922592163,
+      "learning_rate": 6.631329945680743e-05,
+      "loss": 1.6833,
+      "step": 3848
+    },
+    {
+      "epoch": 1.0115939172989128,
+      "grad_norm": 0.6919856667518616,
+      "learning_rate": 6.629577711582268e-05,
+      "loss": 1.7138,
+      "step": 3850
+    },
+    {
+      "epoch": 1.0121194206325745,
+      "grad_norm": 0.590743899345398,
+      "learning_rate": 6.627825477483792e-05,
+      "loss": 1.7058,
+      "step": 3852
+    },
+    {
+      "epoch": 1.0126449239662365,
+      "grad_norm": 0.6320709586143494,
+      "learning_rate": 6.626073243385317e-05,
+      "loss": 1.6988,
+      "step": 3854
+    },
+    {
+      "epoch": 1.0131704272998983,
+      "grad_norm": 0.5564618706703186,
+      "learning_rate": 6.624321009286842e-05,
+      "loss": 1.7098,
+      "step": 3856
+    },
+    {
+      "epoch": 1.01369593063356,
+      "grad_norm": 0.5974157452583313,
+      "learning_rate": 6.622568775188365e-05,
+      "loss": 1.7168,
+      "step": 3858
+    },
+    {
+      "epoch": 1.0142214339672218,
+      "grad_norm": 0.6708089709281921,
+      "learning_rate": 6.62081654108989e-05,
+      "loss": 1.6849,
+      "step": 3860
+    },
+    {
+      "epoch": 1.0147469373008835,
+      "grad_norm": 0.6684040427207947,
+      "learning_rate": 6.619064306991415e-05,
+      "loss": 1.7061,
+      "step": 3862
+    },
+    {
+      "epoch": 1.0152724406345452,
+      "grad_norm": 0.6342937350273132,
+      "learning_rate": 6.617312072892938e-05,
+      "loss": 1.689,
+      "step": 3864
+    },
+    {
+      "epoch": 1.015797943968207,
+      "grad_norm": 0.5644361972808838,
+      "learning_rate": 6.615559838794463e-05,
+      "loss": 1.6784,
+      "step": 3866
+    },
+    {
+      "epoch": 1.0163234473018687,
+      "grad_norm": 0.7777919173240662,
+      "learning_rate": 6.613807604695988e-05,
+      "loss": 1.6912,
+      "step": 3868
+    },
+    {
+      "epoch": 1.0168489506355305,
+      "grad_norm": 0.7786663770675659,
+      "learning_rate": 6.612055370597512e-05,
+      "loss": 1.6968,
+      "step": 3870
+    },
+    {
+      "epoch": 1.0173744539691925,
+      "grad_norm": 0.7163161039352417,
+      "learning_rate": 6.610303136499036e-05,
+      "loss": 1.7118,
+      "step": 3872
+    },
+    {
+      "epoch": 1.0178999573028542,
+      "grad_norm": 0.731606662273407,
+      "learning_rate": 6.60855090240056e-05,
+      "loss": 1.6944,
+      "step": 3874
+    },
+    {
+      "epoch": 1.018425460636516,
+      "grad_norm": 0.6335828900337219,
+      "learning_rate": 6.606798668302085e-05,
+      "loss": 1.7006,
+      "step": 3876
+    },
+    {
+      "epoch": 1.0189509639701777,
+      "grad_norm": 0.7113467454910278,
+      "learning_rate": 6.60504643420361e-05,
+      "loss": 1.694,
+      "step": 3878
+    },
+    {
+      "epoch": 1.0194764673038395,
+      "grad_norm": 0.84892338514328,
+      "learning_rate": 6.603294200105135e-05,
+      "loss": 1.7098,
+      "step": 3880
+    },
+    {
+      "epoch": 1.0200019706375012,
+      "grad_norm": 0.6938359141349792,
+      "learning_rate": 6.60154196600666e-05,
+      "loss": 1.6912,
+      "step": 3882
+    },
+    {
+      "epoch": 1.020527473971163,
+      "grad_norm": 0.6478989124298096,
+      "learning_rate": 6.599789731908183e-05,
+      "loss": 1.6992,
+      "step": 3884
+    },
+    {
+      "epoch": 1.0210529773048247,
+      "grad_norm": 0.7370628118515015,
+      "learning_rate": 6.598037497809708e-05,
+      "loss": 1.6763,
+      "step": 3886
+    },
+    {
+      "epoch": 1.0215784806384864,
+      "grad_norm": 0.6913176774978638,
+      "learning_rate": 6.596285263711233e-05,
+      "loss": 1.6808,
+      "step": 3888
+    },
+    {
+      "epoch": 1.0221039839721484,
+      "grad_norm": 0.6637833118438721,
+      "learning_rate": 6.594533029612756e-05,
+      "loss": 1.679,
+      "step": 3890
+    },
+    {
+      "epoch": 1.0226294873058102,
+      "grad_norm": 0.7522826194763184,
+      "learning_rate": 6.59278079551428e-05,
+      "loss": 1.6729,
+      "step": 3892
+    },
+    {
+      "epoch": 1.023154990639472,
+      "grad_norm": 0.5955492258071899,
+      "learning_rate": 6.591028561415805e-05,
+      "loss": 1.6915,
+      "step": 3894
+    },
+    {
+      "epoch": 1.0236804939731337,
+      "grad_norm": 0.6156378388404846,
+      "learning_rate": 6.58927632731733e-05,
+      "loss": 1.6887,
+      "step": 3896
+    },
+    {
+      "epoch": 1.0242059973067954,
+      "grad_norm": 0.5954993963241577,
+      "learning_rate": 6.587524093218854e-05,
+      "loss": 1.7028,
+      "step": 3898
+    },
+    {
+      "epoch": 1.0247315006404571,
+      "grad_norm": 0.6089223623275757,
+      "learning_rate": 6.585771859120378e-05,
+      "loss": 1.7172,
+      "step": 3900
+    },
+    {
+      "epoch": 1.025257003974119,
+      "grad_norm": 0.6109156012535095,
+      "learning_rate": 6.584019625021903e-05,
+      "loss": 1.6883,
+      "step": 3902
+    },
+    {
+      "epoch": 1.0257825073077806,
+      "grad_norm": 0.769751787185669,
+      "learning_rate": 6.582267390923428e-05,
+      "loss": 1.7029,
+      "step": 3904
+    },
+    {
+      "epoch": 1.0263080106414426,
+      "grad_norm": 0.579433023929596,
+      "learning_rate": 6.580515156824953e-05,
+      "loss": 1.6728,
+      "step": 3906
+    },
+    {
+      "epoch": 1.0268335139751044,
+      "grad_norm": 0.6194645166397095,
+      "learning_rate": 6.578762922726477e-05,
+      "loss": 1.6949,
+      "step": 3908
+    },
+    {
+      "epoch": 1.027359017308766,
+      "grad_norm": 0.5495603680610657,
+      "learning_rate": 6.577010688628001e-05,
+      "loss": 1.7059,
+      "step": 3910
+    },
+    {
+      "epoch": 1.0278845206424279,
+      "grad_norm": 0.7045862674713135,
+      "learning_rate": 6.575258454529526e-05,
+      "loss": 1.7098,
+      "step": 3912
+    },
+    {
+      "epoch": 1.0284100239760896,
+      "grad_norm": 0.6708394885063171,
+      "learning_rate": 6.573506220431049e-05,
+      "loss": 1.6671,
+      "step": 3914
+    },
+    {
+      "epoch": 1.0289355273097514,
+      "grad_norm": 0.6526671051979065,
+      "learning_rate": 6.571753986332574e-05,
+      "loss": 1.6645,
+      "step": 3916
+    },
+    {
+      "epoch": 1.029461030643413,
+      "grad_norm": 0.5542386174201965,
+      "learning_rate": 6.570001752234098e-05,
+      "loss": 1.6694,
+      "step": 3918
+    },
+    {
+      "epoch": 1.0299865339770748,
+      "grad_norm": 0.6871373057365417,
+      "learning_rate": 6.568249518135623e-05,
+      "loss": 1.7054,
+      "step": 3920
+    },
+    {
+      "epoch": 1.0305120373107366,
+      "grad_norm": 0.6650441288948059,
+      "learning_rate": 6.566497284037148e-05,
+      "loss": 1.7034,
+      "step": 3922
+    },
+    {
+      "epoch": 1.0310375406443986,
+      "grad_norm": 1.027212142944336,
+      "learning_rate": 6.564745049938671e-05,
+      "loss": 1.6684,
+      "step": 3924
+    },
+    {
+      "epoch": 1.0315630439780603,
+      "grad_norm": 0.7262475490570068,
+      "learning_rate": 6.562992815840196e-05,
+      "loss": 1.6994,
+      "step": 3926
+    },
+    {
+      "epoch": 1.032088547311722,
+      "grad_norm": 0.697229266166687,
+      "learning_rate": 6.561240581741721e-05,
+      "loss": 1.6627,
+      "step": 3928
+    },
+    {
+      "epoch": 1.0326140506453838,
+      "grad_norm": 0.6965095400810242,
+      "learning_rate": 6.559488347643246e-05,
+      "loss": 1.7078,
+      "step": 3930
+    },
+    {
+      "epoch": 1.0331395539790456,
+      "grad_norm": 0.6743383407592773,
+      "learning_rate": 6.55773611354477e-05,
+      "loss": 1.6954,
+      "step": 3932
+    },
+    {
+      "epoch": 1.0336650573127073,
+      "grad_norm": 0.7088636159896851,
+      "learning_rate": 6.555983879446295e-05,
+      "loss": 1.7079,
+      "step": 3934
+    },
+    {
+      "epoch": 1.034190560646369,
+      "grad_norm": 0.9612395763397217,
+      "learning_rate": 6.554231645347819e-05,
+      "loss": 1.7034,
+      "step": 3936
+    },
+    {
+      "epoch": 1.0347160639800308,
+      "grad_norm": 0.644659161567688,
+      "learning_rate": 6.552479411249343e-05,
+      "loss": 1.6736,
+      "step": 3938
+    },
+    {
+      "epoch": 1.0352415673136925,
+      "grad_norm": 0.6379082202911377,
+      "learning_rate": 6.550727177150867e-05,
+      "loss": 1.6999,
+      "step": 3940
+    },
+    {
+      "epoch": 1.0357670706473545,
+      "grad_norm": 0.8248457908630371,
+      "learning_rate": 6.548974943052391e-05,
+      "loss": 1.7085,
+      "step": 3942
+    },
+    {
+      "epoch": 1.0362925739810163,
+      "grad_norm": 0.9295015335083008,
+      "learning_rate": 6.547222708953916e-05,
+      "loss": 1.6797,
+      "step": 3944
+    },
+    {
+      "epoch": 1.036818077314678,
+      "grad_norm": 0.614661693572998,
+      "learning_rate": 6.545470474855441e-05,
+      "loss": 1.6928,
+      "step": 3946
+    },
+    {
+      "epoch": 1.0373435806483398,
+      "grad_norm": 0.721056342124939,
+      "learning_rate": 6.543718240756966e-05,
+      "loss": 1.6783,
+      "step": 3948
+    },
+    {
+      "epoch": 1.0378690839820015,
+      "grad_norm": 1.0733940601348877,
+      "learning_rate": 6.54196600665849e-05,
+      "loss": 1.7053,
+      "step": 3950
+    },
+    {
+      "epoch": 1.0383945873156633,
+      "grad_norm": 0.6472097039222717,
+      "learning_rate": 6.540213772560014e-05,
+      "loss": 1.7327,
+      "step": 3952
+    },
+    {
+      "epoch": 1.038920090649325,
+      "grad_norm": 0.7506822943687439,
+      "learning_rate": 6.538461538461539e-05,
+      "loss": 1.6703,
+      "step": 3954
+    },
+    {
+      "epoch": 1.0394455939829867,
+      "grad_norm": 0.8442516326904297,
+      "learning_rate": 6.536709304363063e-05,
+      "loss": 1.695,
+      "step": 3956
+    },
+    {
+      "epoch": 1.0399710973166485,
+      "grad_norm": 0.7090259790420532,
+      "learning_rate": 6.534957070264588e-05,
+      "loss": 1.6954,
+      "step": 3958
+    },
+    {
+      "epoch": 1.0404966006503105,
+      "grad_norm": 0.8459334969520569,
+      "learning_rate": 6.533204836166113e-05,
+      "loss": 1.6729,
+      "step": 3960
+    },
+    {
+      "epoch": 1.0410221039839722,
+      "grad_norm": 0.8890243172645569,
+      "learning_rate": 6.531452602067638e-05,
+      "loss": 1.6778,
+      "step": 3962
+    },
+    {
+      "epoch": 1.041547607317634,
+      "grad_norm": 0.9002764821052551,
+      "learning_rate": 6.529700367969161e-05,
+      "loss": 1.6651,
+      "step": 3964
+    },
+    {
+      "epoch": 1.0420731106512957,
+      "grad_norm": 0.7871319055557251,
+      "learning_rate": 6.527948133870684e-05,
+      "loss": 1.7182,
+      "step": 3966
+    },
+    {
+      "epoch": 1.0425986139849575,
+      "grad_norm": 1.2501089572906494,
+      "learning_rate": 6.526195899772209e-05,
+      "loss": 1.7056,
+      "step": 3968
+    },
+    {
+      "epoch": 1.0431241173186192,
+      "grad_norm": 0.8261802792549133,
+      "learning_rate": 6.524443665673734e-05,
+      "loss": 1.7063,
+      "step": 3970
+    },
+    {
+      "epoch": 1.043649620652281,
+      "grad_norm": 0.8968937993049622,
+      "learning_rate": 6.522691431575259e-05,
+      "loss": 1.6993,
+      "step": 3972
+    },
+    {
+      "epoch": 1.0441751239859427,
+      "grad_norm": 0.8691303730010986,
+      "learning_rate": 6.520939197476783e-05,
+      "loss": 1.6731,
+      "step": 3974
+    },
+    {
+      "epoch": 1.0447006273196044,
+      "grad_norm": 0.9870227575302124,
+      "learning_rate": 6.519186963378308e-05,
+      "loss": 1.6689,
+      "step": 3976
+    },
+    {
+      "epoch": 1.0452261306532664,
+      "grad_norm": 0.6180405616760254,
+      "learning_rate": 6.517434729279832e-05,
+      "loss": 1.69,
+      "step": 3978
+    },
+    {
+      "epoch": 1.0457516339869282,
+      "grad_norm": 0.6060590147972107,
+      "learning_rate": 6.515682495181356e-05,
+      "loss": 1.7048,
+      "step": 3980
+    },
+    {
+      "epoch": 1.04627713732059,
+      "grad_norm": 0.6996809244155884,
+      "learning_rate": 6.513930261082881e-05,
+      "loss": 1.693,
+      "step": 3982
+    },
+    {
+      "epoch": 1.0468026406542517,
+      "grad_norm": 0.8016669154167175,
+      "learning_rate": 6.512178026984406e-05,
+      "loss": 1.7072,
+      "step": 3984
+    },
+    {
+      "epoch": 1.0473281439879134,
+      "grad_norm": 0.6650173664093018,
+      "learning_rate": 6.51042579288593e-05,
+      "loss": 1.6923,
+      "step": 3986
+    },
+    {
+      "epoch": 1.0478536473215752,
+      "grad_norm": 0.921610951423645,
+      "learning_rate": 6.508673558787455e-05,
+      "loss": 1.6857,
+      "step": 3988
+    },
+    {
+      "epoch": 1.048379150655237,
+      "grad_norm": 0.5630477666854858,
+      "learning_rate": 6.506921324688979e-05,
+      "loss": 1.6831,
+      "step": 3990
+    },
+    {
+      "epoch": 1.0489046539888986,
+      "grad_norm": 0.7276068329811096,
+      "learning_rate": 6.505169090590502e-05,
+      "loss": 1.6785,
+      "step": 3992
+    },
+    {
+      "epoch": 1.0494301573225606,
+      "grad_norm": 0.7698312997817993,
+      "learning_rate": 6.503416856492027e-05,
+      "loss": 1.6842,
+      "step": 3994
+    },
+    {
+      "epoch": 1.0499556606562224,
+      "grad_norm": 0.8240602016448975,
+      "learning_rate": 6.501664622393552e-05,
+      "loss": 1.6815,
+      "step": 3996
+    },
+    {
+      "epoch": 1.0504811639898841,
+      "grad_norm": 0.6555003523826599,
+      "learning_rate": 6.499912388295076e-05,
+      "loss": 1.6668,
+      "step": 3998
+    },
+    {
+      "epoch": 1.0510066673235459,
+      "grad_norm": 0.7370006442070007,
+      "learning_rate": 6.498160154196601e-05,
+      "loss": 1.6993,
+      "step": 4000
+    },
+    {
+      "epoch": 1.0510066673235459,
+      "eval_loss": 1.7162247896194458,
+      "eval_runtime": 487.2234,
+      "eval_samples_per_second": 249.965,
+      "eval_steps_per_second": 31.246,
+      "step": 4000
+    },
+    {
+      "epoch": 1.0515321706572076,
+      "grad_norm": 0.7542751431465149,
+      "learning_rate": 6.496407920098126e-05,
+      "loss": 1.6854,
+      "step": 4002
+    },
+    {
+      "epoch": 1.0520576739908694,
+      "grad_norm": 0.7216155529022217,
+      "learning_rate": 6.49465568599965e-05,
+      "loss": 1.7008,
+      "step": 4004
+    },
+    {
+      "epoch": 1.052583177324531,
+      "grad_norm": 0.6681018471717834,
+      "learning_rate": 6.492903451901174e-05,
+      "loss": 1.703,
+      "step": 4006
+    },
+    {
+      "epoch": 1.0531086806581929,
+      "grad_norm": 0.8611218929290771,
+      "learning_rate": 6.491151217802699e-05,
+      "loss": 1.6686,
+      "step": 4008
+    },
+    {
+      "epoch": 1.0536341839918546,
+      "grad_norm": 0.6838074922561646,
+      "learning_rate": 6.489398983704224e-05,
+      "loss": 1.7053,
+      "step": 4010
+    },
+    {
+      "epoch": 1.0541596873255166,
+      "grad_norm": 0.6425184607505798,
+      "learning_rate": 6.487646749605748e-05,
+      "loss": 1.6621,
+      "step": 4012
+    },
+    {
+      "epoch": 1.0546851906591783,
+      "grad_norm": 0.8689895272254944,
+      "learning_rate": 6.485894515507273e-05,
+      "loss": 1.6693,
+      "step": 4014
+    },
+    {
+      "epoch": 1.05521069399284,
+      "grad_norm": 0.7122433185577393,
+      "learning_rate": 6.484142281408797e-05,
+      "loss": 1.6936,
+      "step": 4016
+    },
+    {
+      "epoch": 1.0557361973265018,
+      "grad_norm": 0.7124624252319336,
+      "learning_rate": 6.48239004731032e-05,
+      "loss": 1.6915,
+      "step": 4018
+    },
+    {
+      "epoch": 1.0562617006601636,
+      "grad_norm": 0.6151629686355591,
+      "learning_rate": 6.480637813211845e-05,
+      "loss": 1.6601,
+      "step": 4020
+    },
+    {
+      "epoch": 1.0567872039938253,
+      "grad_norm": 0.5656692385673523,
+      "learning_rate": 6.47888557911337e-05,
+      "loss": 1.6651,
+      "step": 4022
+    },
+    {
+      "epoch": 1.057312707327487,
+      "grad_norm": 0.6214647889137268,
+      "learning_rate": 6.477133345014894e-05,
+      "loss": 1.7278,
+      "step": 4024
+    },
+    {
+      "epoch": 1.0578382106611488,
+      "grad_norm": 0.7187774777412415,
+      "learning_rate": 6.475381110916419e-05,
+      "loss": 1.682,
+      "step": 4026
+    },
+    {
+      "epoch": 1.0583637139948106,
+      "grad_norm": 0.602172315120697,
+      "learning_rate": 6.473628876817944e-05,
+      "loss": 1.6712,
+      "step": 4028
+    },
+    {
+      "epoch": 1.0588892173284725,
+      "grad_norm": 0.7032018303871155,
+      "learning_rate": 6.471876642719467e-05,
+      "loss": 1.7056,
+      "step": 4030
+    },
+    {
+      "epoch": 1.0594147206621343,
+      "grad_norm": 0.720413088798523,
+      "learning_rate": 6.470124408620992e-05,
+      "loss": 1.6658,
+      "step": 4032
+    },
+    {
+      "epoch": 1.059940223995796,
+      "grad_norm": 0.7536730170249939,
+      "learning_rate": 6.468372174522517e-05,
+      "loss": 1.6624,
+      "step": 4034
+    },
+    {
+      "epoch": 1.0604657273294578,
+      "grad_norm": 0.6960994005203247,
+      "learning_rate": 6.466619940424041e-05,
+      "loss": 1.6805,
+      "step": 4036
+    },
+    {
+      "epoch": 1.0609912306631195,
+      "grad_norm": 0.6354637145996094,
+      "learning_rate": 6.464867706325566e-05,
+      "loss": 1.6969,
+      "step": 4038
+    },
+    {
+      "epoch": 1.0615167339967813,
+      "grad_norm": 0.659994900226593,
+      "learning_rate": 6.463115472227091e-05,
+      "loss": 1.7076,
+      "step": 4040
+    },
+    {
+      "epoch": 1.062042237330443,
+      "grad_norm": 0.6840848326683044,
+      "learning_rate": 6.461363238128614e-05,
+      "loss": 1.7255,
+      "step": 4042
+    },
+    {
+      "epoch": 1.0625677406641048,
+      "grad_norm": 0.8560720086097717,
+      "learning_rate": 6.459611004030138e-05,
+      "loss": 1.6797,
+      "step": 4044
+    },
+    {
+      "epoch": 1.0630932439977667,
+      "grad_norm": 0.5914828181266785,
+      "learning_rate": 6.457858769931663e-05,
+      "loss": 1.6948,
+      "step": 4046
+    },
+    {
+      "epoch": 1.0636187473314285,
+      "grad_norm": 0.9306698441505432,
+      "learning_rate": 6.456106535833187e-05,
+      "loss": 1.6749,
+      "step": 4048
+    },
+    {
+      "epoch": 1.0641442506650902,
+      "grad_norm": 0.6383719444274902,
+      "learning_rate": 6.454354301734712e-05,
+      "loss": 1.7245,
+      "step": 4050
+    },
+    {
+      "epoch": 1.064669753998752,
+      "grad_norm": 0.6287466287612915,
+      "learning_rate": 6.452602067636237e-05,
+      "loss": 1.6935,
+      "step": 4052
+    },
+    {
+      "epoch": 1.0651952573324137,
+      "grad_norm": 0.7017025947570801,
+      "learning_rate": 6.450849833537762e-05,
+      "loss": 1.6792,
+      "step": 4054
+    },
+    {
+      "epoch": 1.0657207606660755,
+      "grad_norm": 0.6088765859603882,
+      "learning_rate": 6.449097599439285e-05,
+      "loss": 1.6926,
+      "step": 4056
+    },
+    {
+      "epoch": 1.0662462639997372,
+      "grad_norm": 0.6329763531684875,
+      "learning_rate": 6.44734536534081e-05,
+      "loss": 1.7088,
+      "step": 4058
+    },
+    {
+      "epoch": 1.066771767333399,
+      "grad_norm": 0.6406139731407166,
+      "learning_rate": 6.445593131242334e-05,
+      "loss": 1.6847,
+      "step": 4060
+    },
+    {
+      "epoch": 1.0672972706670607,
+      "grad_norm": 0.5512668490409851,
+      "learning_rate": 6.443840897143859e-05,
+      "loss": 1.6572,
+      "step": 4062
+    },
+    {
+      "epoch": 1.0678227740007227,
+      "grad_norm": 0.6363273859024048,
+      "learning_rate": 6.442088663045384e-05,
+      "loss": 1.7079,
+      "step": 4064
+    },
+    {
+      "epoch": 1.0683482773343844,
+      "grad_norm": 0.5485667586326599,
+      "learning_rate": 6.440336428946909e-05,
+      "loss": 1.7357,
+      "step": 4066
+    },
+    {
+      "epoch": 1.0688737806680462,
+      "grad_norm": 0.6467545628547668,
+      "learning_rate": 6.438584194848432e-05,
+      "loss": 1.6527,
+      "step": 4068
+    },
+    {
+      "epoch": 1.069399284001708,
+      "grad_norm": 0.6813017129898071,
+      "learning_rate": 6.436831960749956e-05,
+      "loss": 1.6907,
+      "step": 4070
+    },
+    {
+      "epoch": 1.0699247873353697,
+      "grad_norm": 0.7942006587982178,
+      "learning_rate": 6.43507972665148e-05,
+      "loss": 1.6908,
+      "step": 4072
+    },
+    {
+      "epoch": 1.0704502906690314,
+      "grad_norm": 0.7022045254707336,
+      "learning_rate": 6.433327492553005e-05,
+      "loss": 1.6829,
+      "step": 4074
+    },
+    {
+      "epoch": 1.0709757940026932,
+      "grad_norm": 0.581889271736145,
+      "learning_rate": 6.43157525845453e-05,
+      "loss": 1.6625,
+      "step": 4076
+    },
+    {
+      "epoch": 1.071501297336355,
+      "grad_norm": 0.979885458946228,
+      "learning_rate": 6.429823024356055e-05,
+      "loss": 1.6818,
+      "step": 4078
+    },
+    {
+      "epoch": 1.0720268006700167,
+      "grad_norm": 0.6776348352432251,
+      "learning_rate": 6.428070790257579e-05,
+      "loss": 1.6806,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0725523040036786,
+      "grad_norm": 0.8267763257026672,
+      "learning_rate": 6.426318556159103e-05,
+      "loss": 1.6728,
+      "step": 4082
+    },
+    {
+      "epoch": 1.0730778073373404,
+      "grad_norm": 0.6556562781333923,
+      "learning_rate": 6.424566322060627e-05,
+      "loss": 1.6894,
+      "step": 4084
+    },
+    {
+      "epoch": 1.0736033106710021,
+      "grad_norm": 0.809785783290863,
+      "learning_rate": 6.422814087962152e-05,
+      "loss": 1.6918,
+      "step": 4086
+    },
+    {
+      "epoch": 1.0741288140046639,
+      "grad_norm": 0.5993219017982483,
+      "learning_rate": 6.421061853863677e-05,
+      "loss": 1.673,
+      "step": 4088
+    },
+    {
+      "epoch": 1.0746543173383256,
+      "grad_norm": 0.6594780087471008,
+      "learning_rate": 6.419309619765202e-05,
+      "loss": 1.7099,
+      "step": 4090
+    },
+    {
+      "epoch": 1.0751798206719874,
+      "grad_norm": 0.8069416284561157,
+      "learning_rate": 6.417557385666726e-05,
+      "loss": 1.7032,
+      "step": 4092
+    },
+    {
+      "epoch": 1.075705324005649,
+      "grad_norm": 0.6383103728294373,
+      "learning_rate": 6.41580515156825e-05,
+      "loss": 1.6763,
+      "step": 4094
+    },
+    {
+      "epoch": 1.0762308273393109,
+      "grad_norm": 0.622296929359436,
+      "learning_rate": 6.414052917469773e-05,
+      "loss": 1.6717,
+      "step": 4096
+    },
+    {
+      "epoch": 1.0767563306729726,
+      "grad_norm": 0.5699704885482788,
+      "learning_rate": 6.412300683371298e-05,
+      "loss": 1.6684,
+      "step": 4098
+    },
+    {
+      "epoch": 1.0772818340066346,
+      "grad_norm": 0.653029203414917,
+      "learning_rate": 6.410548449272823e-05,
+      "loss": 1.6898,
+      "step": 4100
+    },
+    {
+      "epoch": 1.0778073373402963,
+      "grad_norm": 0.5569880604743958,
+      "learning_rate": 6.408796215174348e-05,
+      "loss": 1.6897,
+      "step": 4102
+    },
+    {
+      "epoch": 1.078332840673958,
+      "grad_norm": 0.8268890380859375,
+      "learning_rate": 6.407043981075872e-05,
+      "loss": 1.691,
+      "step": 4104
+    },
+    {
+      "epoch": 1.0788583440076198,
+      "grad_norm": 0.6143434643745422,
+      "learning_rate": 6.405291746977397e-05,
+      "loss": 1.7033,
+      "step": 4106
+    },
+    {
+      "epoch": 1.0793838473412816,
+      "grad_norm": 0.6650940775871277,
+      "learning_rate": 6.40353951287892e-05,
+      "loss": 1.7149,
+      "step": 4108
+    },
+    {
+      "epoch": 1.0799093506749433,
+      "grad_norm": 0.9318827986717224,
+      "learning_rate": 6.401787278780445e-05,
+      "loss": 1.7007,
+      "step": 4110
+    },
+    {
+      "epoch": 1.080434854008605,
+      "grad_norm": 0.870758056640625,
+      "learning_rate": 6.40003504468197e-05,
+      "loss": 1.6928,
+      "step": 4112
+    },
+    {
+      "epoch": 1.0809603573422668,
+      "grad_norm": 0.5847840309143066,
+      "learning_rate": 6.398282810583495e-05,
+      "loss": 1.7171,
+      "step": 4114
+    },
+    {
+      "epoch": 1.0814858606759286,
+      "grad_norm": 0.6585195064544678,
+      "learning_rate": 6.39653057648502e-05,
+      "loss": 1.6983,
+      "step": 4116
+    },
+    {
+      "epoch": 1.0820113640095905,
+      "grad_norm": 0.6032067537307739,
+      "learning_rate": 6.394778342386543e-05,
+      "loss": 1.7266,
+      "step": 4118
+    },
+    {
+      "epoch": 1.0825368673432523,
+      "grad_norm": 0.7615604996681213,
+      "learning_rate": 6.393026108288068e-05,
+      "loss": 1.6821,
+      "step": 4120
+    },
+    {
+      "epoch": 1.083062370676914,
+      "grad_norm": 0.6143482327461243,
+      "learning_rate": 6.391273874189591e-05,
+      "loss": 1.6857,
+      "step": 4122
+    },
+    {
+      "epoch": 1.0835878740105758,
+      "grad_norm": 0.7933993935585022,
+      "learning_rate": 6.389521640091116e-05,
+      "loss": 1.6717,
+      "step": 4124
+    },
+    {
+      "epoch": 1.0841133773442375,
+      "grad_norm": 0.6543474793434143,
+      "learning_rate": 6.38776940599264e-05,
+      "loss": 1.6959,
+      "step": 4126
+    },
+    {
+      "epoch": 1.0846388806778993,
+      "grad_norm": 0.6127591133117676,
+      "learning_rate": 6.386017171894165e-05,
+      "loss": 1.6885,
+      "step": 4128
+    },
+    {
+      "epoch": 1.085164384011561,
+      "grad_norm": 0.8532068133354187,
+      "learning_rate": 6.38426493779569e-05,
+      "loss": 1.6763,
+      "step": 4130
+    },
+    {
+      "epoch": 1.0856898873452228,
+      "grad_norm": 0.5886075496673584,
+      "learning_rate": 6.382512703697215e-05,
+      "loss": 1.6857,
+      "step": 4132
+    },
+    {
+      "epoch": 1.0862153906788845,
+      "grad_norm": 0.6927480697631836,
+      "learning_rate": 6.380760469598738e-05,
+      "loss": 1.6978,
+      "step": 4134
+    },
+    {
+      "epoch": 1.0867408940125465,
+      "grad_norm": 0.5445473790168762,
+      "learning_rate": 6.379008235500263e-05,
+      "loss": 1.6985,
+      "step": 4136
+    },
+    {
+      "epoch": 1.0872663973462082,
+      "grad_norm": 0.6567670702934265,
+      "learning_rate": 6.377256001401788e-05,
+      "loss": 1.6897,
+      "step": 4138
+    },
+    {
+      "epoch": 1.08779190067987,
+      "grad_norm": 0.6687731742858887,
+      "learning_rate": 6.375503767303312e-05,
+      "loss": 1.688,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0883174040135317,
+      "grad_norm": 0.575955331325531,
+      "learning_rate": 6.373751533204837e-05,
+      "loss": 1.7204,
+      "step": 4142
+    },
+    {
+      "epoch": 1.0888429073471935,
+      "grad_norm": 0.6957133412361145,
+      "learning_rate": 6.37199929910636e-05,
+      "loss": 1.6945,
+      "step": 4144
+    },
+    {
+      "epoch": 1.0893684106808552,
+      "grad_norm": 0.7448277473449707,
+      "learning_rate": 6.370247065007885e-05,
+      "loss": 1.6953,
+      "step": 4146
+    },
+    {
+      "epoch": 1.089893914014517,
+      "grad_norm": 0.6629153490066528,
+      "learning_rate": 6.368494830909409e-05,
+      "loss": 1.703,
+      "step": 4148
+    },
+    {
+      "epoch": 1.0904194173481787,
+      "grad_norm": 0.7909244298934937,
+      "learning_rate": 6.366742596810934e-05,
+      "loss": 1.6915,
+      "step": 4150
+    },
+    {
+      "epoch": 1.0909449206818405,
+      "grad_norm": 0.5901594161987305,
+      "learning_rate": 6.364990362712458e-05,
+      "loss": 1.6747,
+      "step": 4152
+    },
+    {
+      "epoch": 1.0914704240155024,
+      "grad_norm": 0.6351743340492249,
+      "learning_rate": 6.363238128613983e-05,
+      "loss": 1.6776,
+      "step": 4154
+    },
+    {
+      "epoch": 1.0919959273491642,
+      "grad_norm": 0.7577309608459473,
+      "learning_rate": 6.361485894515508e-05,
+      "loss": 1.6941,
+      "step": 4156
+    },
+    {
+      "epoch": 1.092521430682826,
+      "grad_norm": 0.8337988257408142,
+      "learning_rate": 6.359733660417033e-05,
+      "loss": 1.6985,
+      "step": 4158
+    },
+    {
+      "epoch": 1.0930469340164877,
+      "grad_norm": 0.6406223177909851,
+      "learning_rate": 6.357981426318556e-05,
+      "loss": 1.6687,
+      "step": 4160
+    },
+    {
+      "epoch": 1.0935724373501494,
+      "grad_norm": 0.5720377564430237,
+      "learning_rate": 6.356229192220081e-05,
+      "loss": 1.6844,
+      "step": 4162
+    },
+    {
+      "epoch": 1.0940979406838112,
+      "grad_norm": 0.5843801498413086,
+      "learning_rate": 6.354476958121606e-05,
+      "loss": 1.7099,
+      "step": 4164
+    },
+    {
+      "epoch": 1.094623444017473,
+      "grad_norm": 0.7225820422172546,
+      "learning_rate": 6.35272472402313e-05,
+      "loss": 1.6941,
+      "step": 4166
+    },
+    {
+      "epoch": 1.0951489473511347,
+      "grad_norm": 0.6221319437026978,
+      "learning_rate": 6.350972489924655e-05,
+      "loss": 1.7102,
+      "step": 4168
+    },
+    {
+      "epoch": 1.0956744506847966,
+      "grad_norm": 0.6108186841011047,
+      "learning_rate": 6.349220255826178e-05,
+      "loss": 1.688,
+      "step": 4170
+    },
+    {
+      "epoch": 1.0961999540184584,
+      "grad_norm": 0.6956847906112671,
+      "learning_rate": 6.347468021727703e-05,
+      "loss": 1.67,
+      "step": 4172
+    },
+    {
+      "epoch": 1.0967254573521201,
+      "grad_norm": 0.616248369216919,
+      "learning_rate": 6.345715787629227e-05,
+      "loss": 1.7036,
+      "step": 4174
+    },
+    {
+      "epoch": 1.0972509606857819,
+      "grad_norm": 0.6128188371658325,
+      "learning_rate": 6.343963553530751e-05,
+      "loss": 1.6961,
+      "step": 4176
+    },
+    {
+      "epoch": 1.0977764640194436,
+      "grad_norm": 0.6509534120559692,
+      "learning_rate": 6.342211319432276e-05,
+      "loss": 1.7298,
+      "step": 4178
+    },
+    {
+      "epoch": 1.0983019673531054,
+      "grad_norm": 0.8290244340896606,
+      "learning_rate": 6.340459085333801e-05,
+      "loss": 1.7022,
+      "step": 4180
+    },
+    {
+      "epoch": 1.0988274706867671,
+      "grad_norm": 0.8365241289138794,
+      "learning_rate": 6.338706851235326e-05,
+      "loss": 1.7145,
+      "step": 4182
+    },
+    {
+      "epoch": 1.0993529740204289,
+      "grad_norm": 0.6763657927513123,
+      "learning_rate": 6.33695461713685e-05,
+      "loss": 1.7356,
+      "step": 4184
+    },
+    {
+      "epoch": 1.0998784773540906,
+      "grad_norm": 0.6510267853736877,
+      "learning_rate": 6.335202383038374e-05,
+      "loss": 1.7102,
+      "step": 4186
+    },
+    {
+      "epoch": 1.1004039806877526,
+      "grad_norm": 0.8549639582633972,
+      "learning_rate": 6.333450148939899e-05,
+      "loss": 1.7083,
+      "step": 4188
+    },
+    {
+      "epoch": 1.1009294840214143,
+      "grad_norm": 0.5909300446510315,
+      "learning_rate": 6.331697914841423e-05,
+      "loss": 1.7224,
+      "step": 4190
+    },
+    {
+      "epoch": 1.101454987355076,
+      "grad_norm": 0.5720754265785217,
+      "learning_rate": 6.329945680742948e-05,
+      "loss": 1.6936,
+      "step": 4192
+    },
+    {
+      "epoch": 1.1019804906887378,
+      "grad_norm": 0.8893141150474548,
+      "learning_rate": 6.328193446644473e-05,
+      "loss": 1.6726,
+      "step": 4194
+    },
+    {
+      "epoch": 1.1025059940223996,
+      "grad_norm": 0.5539205074310303,
+      "learning_rate": 6.326441212545996e-05,
+      "loss": 1.6461,
+      "step": 4196
+    },
+    {
+      "epoch": 1.1030314973560613,
+      "grad_norm": 0.8009784817695618,
+      "learning_rate": 6.324688978447521e-05,
+      "loss": 1.7247,
+      "step": 4198
+    },
+    {
+      "epoch": 1.103557000689723,
+      "grad_norm": 0.6527197957038879,
+      "learning_rate": 6.322936744349046e-05,
+      "loss": 1.7007,
+      "step": 4200
+    },
+    {
+      "epoch": 1.1040825040233848,
+      "grad_norm": 0.6934798955917358,
+      "learning_rate": 6.321184510250569e-05,
+      "loss": 1.6944,
+      "step": 4202
+    },
+    {
+      "epoch": 1.1046080073570468,
+      "grad_norm": 0.6967670917510986,
+      "learning_rate": 6.319432276152094e-05,
+      "loss": 1.6937,
+      "step": 4204
+    },
+    {
+      "epoch": 1.1051335106907085,
+      "grad_norm": 0.7582138776779175,
+      "learning_rate": 6.317680042053619e-05,
+      "loss": 1.703,
+      "step": 4206
+    },
+    {
+      "epoch": 1.1056590140243703,
+      "grad_norm": 0.6267343759536743,
+      "learning_rate": 6.315927807955143e-05,
+      "loss": 1.6888,
+      "step": 4208
+    },
+    {
+      "epoch": 1.106184517358032,
+      "grad_norm": 0.6019650101661682,
+      "learning_rate": 6.314175573856668e-05,
+      "loss": 1.7076,
+      "step": 4210
+    },
+    {
+      "epoch": 1.1067100206916938,
+      "grad_norm": 0.5411569476127625,
+      "learning_rate": 6.312423339758193e-05,
+      "loss": 1.6816,
+      "step": 4212
+    },
+    {
+      "epoch": 1.1072355240253555,
+      "grad_norm": 1.0304468870162964,
+      "learning_rate": 6.310671105659716e-05,
+      "loss": 1.7147,
+      "step": 4214
+    },
+    {
+      "epoch": 1.1077610273590173,
+      "grad_norm": 0.6413879990577698,
+      "learning_rate": 6.308918871561241e-05,
+      "loss": 1.6949,
+      "step": 4216
+    },
+    {
+      "epoch": 1.108286530692679,
+      "grad_norm": 0.6234250664710999,
+      "learning_rate": 6.307166637462766e-05,
+      "loss": 1.6998,
+      "step": 4218
+    },
+    {
+      "epoch": 1.1088120340263408,
+      "grad_norm": 0.5468757152557373,
+      "learning_rate": 6.305414403364289e-05,
+      "loss": 1.7209,
+      "step": 4220
+    },
+    {
+      "epoch": 1.1093375373600027,
+      "grad_norm": 0.7098813056945801,
+      "learning_rate": 6.303662169265814e-05,
+      "loss": 1.6913,
+      "step": 4222
+    },
+    {
+      "epoch": 1.1098630406936645,
+      "grad_norm": 0.7469329237937927,
+      "learning_rate": 6.301909935167339e-05,
+      "loss": 1.6945,
+      "step": 4224
+    },
+    {
+      "epoch": 1.1103885440273262,
+      "grad_norm": 0.634135901927948,
+      "learning_rate": 6.300157701068863e-05,
+      "loss": 1.6777,
+      "step": 4226
+    },
+    {
+      "epoch": 1.110914047360988,
+      "grad_norm": 0.670148491859436,
+      "learning_rate": 6.298405466970387e-05,
+      "loss": 1.6921,
+      "step": 4228
+    },
+    {
+      "epoch": 1.1114395506946497,
+      "grad_norm": 0.6443579792976379,
+      "learning_rate": 6.296653232871912e-05,
+      "loss": 1.7338,
+      "step": 4230
+    },
+    {
+      "epoch": 1.1119650540283115,
+      "grad_norm": 0.6890257000923157,
+      "learning_rate": 6.294900998773436e-05,
+      "loss": 1.7463,
+      "step": 4232
+    },
+    {
+      "epoch": 1.1124905573619732,
+      "grad_norm": 0.5529821515083313,
+      "learning_rate": 6.293148764674961e-05,
+      "loss": 1.6557,
+      "step": 4234
+    },
+    {
+      "epoch": 1.113016060695635,
+      "grad_norm": 0.7622890472412109,
+      "learning_rate": 6.291396530576486e-05,
+      "loss": 1.7009,
+      "step": 4236
+    },
+    {
+      "epoch": 1.1135415640292967,
+      "grad_norm": 0.6255015730857849,
+      "learning_rate": 6.28964429647801e-05,
+      "loss": 1.6922,
+      "step": 4238
+    },
+    {
+      "epoch": 1.1140670673629587,
+      "grad_norm": 0.5990403890609741,
+      "learning_rate": 6.287892062379534e-05,
+      "loss": 1.7081,
+      "step": 4240
+    },
+    {
+      "epoch": 1.1145925706966204,
+      "grad_norm": 0.6255038976669312,
+      "learning_rate": 6.286139828281059e-05,
+      "loss": 1.6878,
+      "step": 4242
+    },
+    {
+      "epoch": 1.1151180740302822,
+      "grad_norm": 0.6530934572219849,
+      "learning_rate": 6.284387594182584e-05,
+      "loss": 1.6948,
+      "step": 4244
+    },
+    {
+      "epoch": 1.115643577363944,
+      "grad_norm": 0.6111620664596558,
+      "learning_rate": 6.282635360084107e-05,
+      "loss": 1.6752,
+      "step": 4246
+    },
+    {
+      "epoch": 1.1161690806976057,
+      "grad_norm": 0.7106771469116211,
+      "learning_rate": 6.280883125985632e-05,
+      "loss": 1.6941,
+      "step": 4248
+    },
+    {
+      "epoch": 1.1166945840312674,
+      "grad_norm": 0.6513094902038574,
+      "learning_rate": 6.279130891887156e-05,
+      "loss": 1.6937,
+      "step": 4250
+    },
+    {
+      "epoch": 1.1172200873649292,
+      "grad_norm": 0.6989749670028687,
+      "learning_rate": 6.277378657788681e-05,
+      "loss": 1.7048,
+      "step": 4252
+    },
+    {
+      "epoch": 1.117745590698591,
+      "grad_norm": 0.6800375580787659,
+      "learning_rate": 6.275626423690205e-05,
+      "loss": 1.6896,
+      "step": 4254
+    },
+    {
+      "epoch": 1.1182710940322527,
+      "grad_norm": 0.5913258790969849,
+      "learning_rate": 6.27387418959173e-05,
+      "loss": 1.7218,
+      "step": 4256
+    },
+    {
+      "epoch": 1.1187965973659146,
+      "grad_norm": 0.6066367030143738,
+      "learning_rate": 6.272121955493254e-05,
+      "loss": 1.7018,
+      "step": 4258
+    },
+    {
+      "epoch": 1.1193221006995764,
+      "grad_norm": 0.6343132853507996,
+      "learning_rate": 6.270369721394779e-05,
+      "loss": 1.6942,
+      "step": 4260
+    },
+    {
+      "epoch": 1.1198476040332381,
+      "grad_norm": 0.72137451171875,
+      "learning_rate": 6.268617487296304e-05,
+      "loss": 1.6996,
+      "step": 4262
+    },
+    {
+      "epoch": 1.1203731073668999,
+      "grad_norm": 0.6029708981513977,
+      "learning_rate": 6.266865253197828e-05,
+      "loss": 1.7288,
+      "step": 4264
+    },
+    {
+      "epoch": 1.1208986107005616,
+      "grad_norm": 0.5431277751922607,
+      "learning_rate": 6.265113019099352e-05,
+      "loss": 1.677,
+      "step": 4266
+    },
+    {
+      "epoch": 1.1214241140342234,
+      "grad_norm": 0.5661784410476685,
+      "learning_rate": 6.263360785000877e-05,
+      "loss": 1.7043,
+      "step": 4268
+    },
+    {
+      "epoch": 1.1219496173678851,
+      "grad_norm": 0.611405074596405,
+      "learning_rate": 6.261608550902401e-05,
+      "loss": 1.6822,
+      "step": 4270
+    },
+    {
+      "epoch": 1.1224751207015469,
+      "grad_norm": 0.6432753205299377,
+      "learning_rate": 6.259856316803925e-05,
+      "loss": 1.7392,
+      "step": 4272
+    },
+    {
+      "epoch": 1.1230006240352086,
+      "grad_norm": 0.6208961606025696,
+      "learning_rate": 6.25810408270545e-05,
+      "loss": 1.649,
+      "step": 4274
+    },
+    {
+      "epoch": 1.1235261273688706,
+      "grad_norm": 0.628790020942688,
+      "learning_rate": 6.256351848606974e-05,
+      "loss": 1.6912,
+      "step": 4276
+    },
+    {
+      "epoch": 1.1240516307025323,
+      "grad_norm": 0.6716508269309998,
+      "learning_rate": 6.254599614508499e-05,
+      "loss": 1.7001,
+      "step": 4278
+    },
+    {
+      "epoch": 1.124577134036194,
+      "grad_norm": 0.7140381932258606,
+      "learning_rate": 6.252847380410022e-05,
+      "loss": 1.6913,
+      "step": 4280
+    },
+    {
+      "epoch": 1.1251026373698558,
+      "grad_norm": 0.6927066445350647,
+      "learning_rate": 6.251095146311547e-05,
+      "loss": 1.7051,
+      "step": 4282
+    },
+    {
+      "epoch": 1.1256281407035176,
+      "grad_norm": 0.6016809940338135,
+      "learning_rate": 6.249342912213072e-05,
+      "loss": 1.6734,
+      "step": 4284
+    },
+    {
+      "epoch": 1.1261536440371793,
+      "grad_norm": 0.6632773876190186,
+      "learning_rate": 6.247590678114597e-05,
+      "loss": 1.6911,
+      "step": 4286
+    },
+    {
+      "epoch": 1.126679147370841,
+      "grad_norm": 0.6406089663505554,
+      "learning_rate": 6.245838444016121e-05,
+      "loss": 1.6867,
+      "step": 4288
+    },
+    {
+      "epoch": 1.1272046507045028,
+      "grad_norm": 0.6394585371017456,
+      "learning_rate": 6.244086209917646e-05,
+      "loss": 1.7197,
+      "step": 4290
+    },
+    {
+      "epoch": 1.1277301540381646,
+      "grad_norm": 0.9273669719696045,
+      "learning_rate": 6.24233397581917e-05,
+      "loss": 1.6678,
+      "step": 4292
+    },
+    {
+      "epoch": 1.1282556573718265,
+      "grad_norm": 0.703338086605072,
+      "learning_rate": 6.240581741720694e-05,
+      "loss": 1.6839,
+      "step": 4294
+    },
+    {
+      "epoch": 1.1287811607054883,
+      "grad_norm": 0.788475751876831,
+      "learning_rate": 6.238829507622218e-05,
+      "loss": 1.6906,
+      "step": 4296
+    },
+    {
+      "epoch": 1.12930666403915,
+      "grad_norm": 0.9603837728500366,
+      "learning_rate": 6.237077273523742e-05,
+      "loss": 1.6735,
+      "step": 4298
+    },
+    {
+      "epoch": 1.1298321673728118,
+      "grad_norm": 0.7163801789283752,
+      "learning_rate": 6.235325039425267e-05,
+      "loss": 1.6773,
+      "step": 4300
+    },
+    {
+      "epoch": 1.1303576707064735,
+      "grad_norm": 0.8321335911750793,
+      "learning_rate": 6.233572805326792e-05,
+      "loss": 1.6836,
+      "step": 4302
+    },
+    {
+      "epoch": 1.1308831740401353,
+      "grad_norm": 0.6714113354682922,
+      "learning_rate": 6.231820571228317e-05,
+      "loss": 1.6949,
+      "step": 4304
+    },
+    {
+      "epoch": 1.131408677373797,
+      "grad_norm": 0.9358461499214172,
+      "learning_rate": 6.23006833712984e-05,
+      "loss": 1.6965,
+      "step": 4306
+    },
+    {
+      "epoch": 1.1319341807074588,
+      "grad_norm": 0.8378779292106628,
+      "learning_rate": 6.228316103031365e-05,
+      "loss": 1.697,
+      "step": 4308
+    },
+    {
+      "epoch": 1.1324596840411205,
+      "grad_norm": 0.5909398198127747,
+      "learning_rate": 6.22656386893289e-05,
+      "loss": 1.7349,
+      "step": 4310
+    },
+    {
+      "epoch": 1.1329851873747825,
+      "grad_norm": 0.5922428965568542,
+      "learning_rate": 6.224811634834414e-05,
+      "loss": 1.6598,
+      "step": 4312
+    },
+    {
+      "epoch": 1.1335106907084442,
+      "grad_norm": 0.7783740162849426,
+      "learning_rate": 6.223059400735939e-05,
+      "loss": 1.6822,
+      "step": 4314
+    },
+    {
+      "epoch": 1.134036194042106,
+      "grad_norm": 0.5940402746200562,
+      "learning_rate": 6.221307166637464e-05,
+      "loss": 1.7093,
+      "step": 4316
+    },
+    {
+      "epoch": 1.1345616973757677,
+      "grad_norm": 0.7679793834686279,
+      "learning_rate": 6.219554932538987e-05,
+      "loss": 1.6886,
+      "step": 4318
+    },
+    {
+      "epoch": 1.1350872007094295,
+      "grad_norm": 0.624814510345459,
+      "learning_rate": 6.217802698440512e-05,
+      "loss": 1.684,
+      "step": 4320
+    },
+    {
+      "epoch": 1.1356127040430912,
+      "grad_norm": 0.87510085105896,
+      "learning_rate": 6.216050464342036e-05,
+      "loss": 1.7228,
+      "step": 4322
+    },
+    {
+      "epoch": 1.136138207376753,
+      "grad_norm": 0.6720306277275085,
+      "learning_rate": 6.21429823024356e-05,
+      "loss": 1.6675,
+      "step": 4324
+    },
+    {
+      "epoch": 1.136663710710415,
+      "grad_norm": 0.6201871037483215,
+      "learning_rate": 6.212545996145085e-05,
+      "loss": 1.708,
+      "step": 4326
+    },
+    {
+      "epoch": 1.1371892140440765,
+      "grad_norm": 0.6067994832992554,
+      "learning_rate": 6.21079376204661e-05,
+      "loss": 1.6825,
+      "step": 4328
+    },
+    {
+      "epoch": 1.1377147173777384,
+      "grad_norm": 0.5989435315132141,
+      "learning_rate": 6.209041527948135e-05,
+      "loss": 1.6796,
+      "step": 4330
+    },
+    {
+      "epoch": 1.1382402207114002,
+      "grad_norm": 0.6668772101402283,
+      "learning_rate": 6.207289293849658e-05,
+      "loss": 1.6989,
+      "step": 4332
+    },
+    {
+      "epoch": 1.138765724045062,
+      "grad_norm": 0.6205788850784302,
+      "learning_rate": 6.205537059751183e-05,
+      "loss": 1.6787,
+      "step": 4334
+    },
+    {
+      "epoch": 1.1392912273787237,
+      "grad_norm": 0.7359316349029541,
+      "learning_rate": 6.203784825652707e-05,
+      "loss": 1.6806,
+      "step": 4336
+    },
+    {
+      "epoch": 1.1398167307123854,
+      "grad_norm": 0.5981895923614502,
+      "learning_rate": 6.202032591554232e-05,
+      "loss": 1.6804,
+      "step": 4338
+    },
+    {
+      "epoch": 1.1403422340460472,
+      "grad_norm": 0.6678347587585449,
+      "learning_rate": 6.200280357455757e-05,
+      "loss": 1.6966,
+      "step": 4340
+    },
+    {
+      "epoch": 1.140867737379709,
+      "grad_norm": 0.691186249256134,
+      "learning_rate": 6.198528123357282e-05,
+      "loss": 1.6746,
+      "step": 4342
+    },
+    {
+      "epoch": 1.141393240713371,
+      "grad_norm": 0.6158877611160278,
+      "learning_rate": 6.196775889258805e-05,
+      "loss": 1.6968,
+      "step": 4344
+    },
+    {
+      "epoch": 1.1419187440470326,
+      "grad_norm": 0.6359262466430664,
+      "learning_rate": 6.19502365516033e-05,
+      "loss": 1.6768,
+      "step": 4346
+    },
+    {
+      "epoch": 1.1424442473806944,
+      "grad_norm": 0.5825957655906677,
+      "learning_rate": 6.193271421061853e-05,
+      "loss": 1.6953,
+      "step": 4348
+    },
+    {
+      "epoch": 1.1429697507143561,
+      "grad_norm": 0.6011431217193604,
+      "learning_rate": 6.191519186963378e-05,
+      "loss": 1.6895,
+      "step": 4350
+    },
+    {
+      "epoch": 1.1434952540480179,
+      "grad_norm": 0.5862739086151123,
+      "learning_rate": 6.189766952864903e-05,
+      "loss": 1.6716,
+      "step": 4352
+    },
+    {
+      "epoch": 1.1440207573816796,
+      "grad_norm": 0.6541095972061157,
+      "learning_rate": 6.188014718766428e-05,
+      "loss": 1.6748,
+      "step": 4354
+    },
+    {
+      "epoch": 1.1445462607153414,
+      "grad_norm": 0.6046082973480225,
+      "learning_rate": 6.186262484667952e-05,
+      "loss": 1.6677,
+      "step": 4356
+    },
+    {
+      "epoch": 1.1450717640490031,
+      "grad_norm": 0.5532662868499756,
+      "learning_rate": 6.184510250569476e-05,
+      "loss": 1.6567,
+      "step": 4358
+    },
+    {
+      "epoch": 1.1455972673826649,
+      "grad_norm": 0.6711378693580627,
+      "learning_rate": 6.182758016471e-05,
+      "loss": 1.6731,
+      "step": 4360
+    },
+    {
+      "epoch": 1.1461227707163268,
+      "grad_norm": 0.6071760654449463,
+      "learning_rate": 6.181005782372525e-05,
+      "loss": 1.6749,
+      "step": 4362
+    },
+    {
+      "epoch": 1.1466482740499886,
+      "grad_norm": 0.6723074913024902,
+      "learning_rate": 6.17925354827405e-05,
+      "loss": 1.6757,
+      "step": 4364
+    },
+    {
+      "epoch": 1.1471737773836503,
+      "grad_norm": 0.649804949760437,
+      "learning_rate": 6.177501314175575e-05,
+      "loss": 1.6809,
+      "step": 4366
+    },
+    {
+      "epoch": 1.147699280717312,
+      "grad_norm": 0.6812226176261902,
+      "learning_rate": 6.1757490800771e-05,
+      "loss": 1.694,
+      "step": 4368
+    },
+    {
+      "epoch": 1.1482247840509738,
+      "grad_norm": 0.6040016412734985,
+      "learning_rate": 6.173996845978623e-05,
+      "loss": 1.7262,
+      "step": 4370
+    },
+    {
+      "epoch": 1.1487502873846356,
+      "grad_norm": 0.652336597442627,
+      "learning_rate": 6.172244611880148e-05,
+      "loss": 1.6702,
+      "step": 4372
+    },
+    {
+      "epoch": 1.1492757907182973,
+      "grad_norm": 0.5667666792869568,
+      "learning_rate": 6.170492377781671e-05,
+      "loss": 1.6873,
+      "step": 4374
+    },
+    {
+      "epoch": 1.149801294051959,
+      "grad_norm": 0.6907036304473877,
+      "learning_rate": 6.168740143683196e-05,
+      "loss": 1.7076,
+      "step": 4376
+    },
+    {
+      "epoch": 1.1503267973856208,
+      "grad_norm": 0.5834628939628601,
+      "learning_rate": 6.16698790958472e-05,
+      "loss": 1.6956,
+      "step": 4378
+    },
+    {
+      "epoch": 1.1508523007192828,
+      "grad_norm": 0.6255402565002441,
+      "learning_rate": 6.165235675486245e-05,
+      "loss": 1.7165,
+      "step": 4380
+    },
+    {
+      "epoch": 1.1513778040529445,
+      "grad_norm": 0.6630857586860657,
+      "learning_rate": 6.16348344138777e-05,
+      "loss": 1.6898,
+      "step": 4382
+    },
+    {
+      "epoch": 1.1519033073866063,
+      "grad_norm": 0.6802171468734741,
+      "learning_rate": 6.161731207289293e-05,
+      "loss": 1.6946,
+      "step": 4384
+    },
+    {
+      "epoch": 1.152428810720268,
+      "grad_norm": 0.5508185029029846,
+      "learning_rate": 6.159978973190818e-05,
+      "loss": 1.676,
+      "step": 4386
+    },
+    {
+      "epoch": 1.1529543140539298,
+      "grad_norm": 0.5373386740684509,
+      "learning_rate": 6.158226739092343e-05,
+      "loss": 1.6595,
+      "step": 4388
+    },
+    {
+      "epoch": 1.1534798173875915,
+      "grad_norm": 0.5505905747413635,
+      "learning_rate": 6.156474504993868e-05,
+      "loss": 1.679,
+      "step": 4390
+    },
+    {
+      "epoch": 1.1540053207212533,
+      "grad_norm": 0.5813027620315552,
+      "learning_rate": 6.154722270895392e-05,
+      "loss": 1.6824,
+      "step": 4392
+    },
+    {
+      "epoch": 1.154530824054915,
+      "grad_norm": 0.6369917392730713,
+      "learning_rate": 6.152970036796917e-05,
+      "loss": 1.6704,
+      "step": 4394
+    },
+    {
+      "epoch": 1.1550563273885768,
+      "grad_norm": 0.5745425224304199,
+      "learning_rate": 6.15121780269844e-05,
+      "loss": 1.6583,
+      "step": 4396
+    },
+    {
+      "epoch": 1.1555818307222387,
+      "grad_norm": 0.6166395545005798,
+      "learning_rate": 6.149465568599964e-05,
+      "loss": 1.6866,
+      "step": 4398
+    },
+    {
+      "epoch": 1.1561073340559005,
+      "grad_norm": 0.5606328248977661,
+      "learning_rate": 6.147713334501489e-05,
+      "loss": 1.6844,
+      "step": 4400
+    },
+    {
+      "epoch": 1.1561073340559005,
+      "eval_loss": 1.7070621252059937,
+      "eval_runtime": 487.2485,
+      "eval_samples_per_second": 249.953,
+      "eval_steps_per_second": 31.245,
+      "step": 4400
+    },
+    {
+      "epoch": 1.1566328373895622,
+      "grad_norm": 0.6324471235275269,
+      "learning_rate": 6.145961100403014e-05,
+      "loss": 1.6756,
+      "step": 4402
+    },
+    {
+      "epoch": 1.157158340723224,
+      "grad_norm": 0.66776043176651,
+      "learning_rate": 6.144208866304538e-05,
+      "loss": 1.7205,
+      "step": 4404
+    },
+    {
+      "epoch": 1.1576838440568857,
+      "grad_norm": 0.5929023623466492,
+      "learning_rate": 6.142456632206063e-05,
+      "loss": 1.6974,
+      "step": 4406
+    },
+    {
+      "epoch": 1.1582093473905475,
+      "grad_norm": 0.6727093458175659,
+      "learning_rate": 6.140704398107588e-05,
+      "loss": 1.6482,
+      "step": 4408
+    },
+    {
+      "epoch": 1.1587348507242092,
+      "grad_norm": 0.5299540758132935,
+      "learning_rate": 6.138952164009111e-05,
+      "loss": 1.712,
+      "step": 4410
+    },
+    {
+      "epoch": 1.159260354057871,
+      "grad_norm": 0.6730796098709106,
+      "learning_rate": 6.137199929910636e-05,
+      "loss": 1.661,
+      "step": 4412
+    },
+    {
+      "epoch": 1.1597858573915327,
+      "grad_norm": 0.6617345213890076,
+      "learning_rate": 6.135447695812161e-05,
+      "loss": 1.706,
+      "step": 4414
+    },
+    {
+      "epoch": 1.1603113607251947,
+      "grad_norm": 0.5500732660293579,
+      "learning_rate": 6.133695461713686e-05,
+      "loss": 1.7226,
+      "step": 4416
+    },
+    {
+      "epoch": 1.1608368640588564,
+      "grad_norm": 0.6082700490951538,
+      "learning_rate": 6.13194322761521e-05,
+      "loss": 1.6724,
+      "step": 4418
+    },
+    {
+      "epoch": 1.1613623673925182,
+      "grad_norm": 0.6681337356567383,
+      "learning_rate": 6.130190993516735e-05,
+      "loss": 1.6395,
+      "step": 4420
+    },
+    {
+      "epoch": 1.16188787072618,
+      "grad_norm": 0.6328011155128479,
+      "learning_rate": 6.128438759418258e-05,
+      "loss": 1.6802,
+      "step": 4422
+    },
+    {
+      "epoch": 1.1624133740598417,
+      "grad_norm": 0.6065677404403687,
+      "learning_rate": 6.126686525319782e-05,
+      "loss": 1.6797,
+      "step": 4424
+    },
+    {
+      "epoch": 1.1629388773935034,
+      "grad_norm": 0.5718259811401367,
+      "learning_rate": 6.124934291221307e-05,
+      "loss": 1.6686,
+      "step": 4426
+    },
+    {
+      "epoch": 1.1634643807271652,
+      "grad_norm": 0.5725675821304321,
+      "learning_rate": 6.123182057122831e-05,
+      "loss": 1.7017,
+      "step": 4428
+    },
+    {
+      "epoch": 1.163989884060827,
+      "grad_norm": 0.5983812808990479,
+      "learning_rate": 6.121429823024356e-05,
+      "loss": 1.6779,
+      "step": 4430
+    },
+    {
+      "epoch": 1.1645153873944887,
+      "grad_norm": 0.5576106309890747,
+      "learning_rate": 6.119677588925881e-05,
+      "loss": 1.6843,
+      "step": 4432
+    },
+    {
+      "epoch": 1.1650408907281506,
+      "grad_norm": 0.9009653329849243,
+      "learning_rate": 6.117925354827406e-05,
+      "loss": 1.6473,
+      "step": 4434
+    },
+    {
+      "epoch": 1.1655663940618124,
+      "grad_norm": 0.7119090557098389,
+      "learning_rate": 6.116173120728929e-05,
+      "loss": 1.6738,
+      "step": 4436
+    },
+    {
+      "epoch": 1.1660918973954741,
+      "grad_norm": 0.8985000848770142,
+      "learning_rate": 6.114420886630454e-05,
+      "loss": 1.6753,
+      "step": 4438
+    },
+    {
+      "epoch": 1.166617400729136,
+      "grad_norm": 0.6928833723068237,
+      "learning_rate": 6.112668652531979e-05,
+      "loss": 1.6922,
+      "step": 4440
+    },
+    {
+      "epoch": 1.1671429040627976,
+      "grad_norm": 0.5745196342468262,
+      "learning_rate": 6.110916418433503e-05,
+      "loss": 1.6845,
+      "step": 4442
+    },
+    {
+      "epoch": 1.1676684073964594,
+      "grad_norm": 0.6172153949737549,
+      "learning_rate": 6.109164184335028e-05,
+      "loss": 1.7065,
+      "step": 4444
+    },
+    {
+      "epoch": 1.1681939107301211,
+      "grad_norm": 0.606926441192627,
+      "learning_rate": 6.107411950236553e-05,
+      "loss": 1.6673,
+      "step": 4446
+    },
+    {
+      "epoch": 1.1687194140637829,
+      "grad_norm": 0.6731804609298706,
+      "learning_rate": 6.105659716138076e-05,
+      "loss": 1.7006,
+      "step": 4448
+    },
+    {
+      "epoch": 1.1692449173974446,
+      "grad_norm": 0.8230140805244446,
+      "learning_rate": 6.103907482039601e-05,
+      "loss": 1.694,
+      "step": 4450
+    },
+    {
+      "epoch": 1.1697704207311066,
+      "grad_norm": 0.809855580329895,
+      "learning_rate": 6.102155247941125e-05,
+      "loss": 1.699,
+      "step": 4452
+    },
+    {
+      "epoch": 1.1702959240647683,
+      "grad_norm": 0.6014560461044312,
+      "learning_rate": 6.100403013842649e-05,
+      "loss": 1.7296,
+      "step": 4454
+    },
+    {
+      "epoch": 1.17082142739843,
+      "grad_norm": 0.6129987835884094,
+      "learning_rate": 6.098650779744174e-05,
+      "loss": 1.6596,
+      "step": 4456
+    },
+    {
+      "epoch": 1.1713469307320918,
+      "grad_norm": 0.8771330118179321,
+      "learning_rate": 6.0968985456456986e-05,
+      "loss": 1.6907,
+      "step": 4458
+    },
+    {
+      "epoch": 1.1718724340657536,
+      "grad_norm": 0.5508044362068176,
+      "learning_rate": 6.095146311547223e-05,
+      "loss": 1.7086,
+      "step": 4460
+    },
+    {
+      "epoch": 1.1723979373994153,
+      "grad_norm": 0.7334650158882141,
+      "learning_rate": 6.0933940774487474e-05,
+      "loss": 1.6956,
+      "step": 4462
+    },
+    {
+      "epoch": 1.172923440733077,
+      "grad_norm": 0.7739654183387756,
+      "learning_rate": 6.091641843350272e-05,
+      "loss": 1.6785,
+      "step": 4464
+    },
+    {
+      "epoch": 1.1734489440667388,
+      "grad_norm": 0.6686828136444092,
+      "learning_rate": 6.089889609251796e-05,
+      "loss": 1.6571,
+      "step": 4466
+    },
+    {
+      "epoch": 1.1739744474004006,
+      "grad_norm": 0.6591079831123352,
+      "learning_rate": 6.088137375153321e-05,
+      "loss": 1.6661,
+      "step": 4468
+    },
+    {
+      "epoch": 1.1744999507340625,
+      "grad_norm": 0.520408570766449,
+      "learning_rate": 6.086385141054846e-05,
+      "loss": 1.7076,
+      "step": 4470
+    },
+    {
+      "epoch": 1.1750254540677243,
+      "grad_norm": 0.7765664458274841,
+      "learning_rate": 6.08463290695637e-05,
+      "loss": 1.6801,
+      "step": 4472
+    },
+    {
+      "epoch": 1.175550957401386,
+      "grad_norm": 0.6031415462493896,
+      "learning_rate": 6.0828806728578946e-05,
+      "loss": 1.6999,
+      "step": 4474
+    },
+    {
+      "epoch": 1.1760764607350478,
+      "grad_norm": 0.694837749004364,
+      "learning_rate": 6.081128438759418e-05,
+      "loss": 1.7186,
+      "step": 4476
+    },
+    {
+      "epoch": 1.1766019640687095,
+      "grad_norm": 0.5900973081588745,
+      "learning_rate": 6.079376204660943e-05,
+      "loss": 1.6832,
+      "step": 4478
+    },
+    {
+      "epoch": 1.1771274674023713,
+      "grad_norm": 0.5364487171173096,
+      "learning_rate": 6.077623970562467e-05,
+      "loss": 1.6756,
+      "step": 4480
+    },
+    {
+      "epoch": 1.177652970736033,
+      "grad_norm": 0.5482999086380005,
+      "learning_rate": 6.0758717364639916e-05,
+      "loss": 1.697,
+      "step": 4482
+    },
+    {
+      "epoch": 1.178178474069695,
+      "grad_norm": 0.7080986499786377,
+      "learning_rate": 6.0741195023655164e-05,
+      "loss": 1.7187,
+      "step": 4484
+    },
+    {
+      "epoch": 1.1787039774033565,
+      "grad_norm": 0.6446167230606079,
+      "learning_rate": 6.0723672682670405e-05,
+      "loss": 1.6753,
+      "step": 4486
+    },
+    {
+      "epoch": 1.1792294807370185,
+      "grad_norm": 0.6965610384941101,
+      "learning_rate": 6.070615034168565e-05,
+      "loss": 1.689,
+      "step": 4488
+    },
+    {
+      "epoch": 1.1797549840706802,
+      "grad_norm": 0.5724831819534302,
+      "learning_rate": 6.06886280007009e-05,
+      "loss": 1.6872,
+      "step": 4490
+    },
+    {
+      "epoch": 1.180280487404342,
+      "grad_norm": 0.7344499230384827,
+      "learning_rate": 6.067110565971614e-05,
+      "loss": 1.6421,
+      "step": 4492
+    },
+    {
+      "epoch": 1.1808059907380037,
+      "grad_norm": 0.663684070110321,
+      "learning_rate": 6.065358331873139e-05,
+      "loss": 1.6909,
+      "step": 4494
+    },
+    {
+      "epoch": 1.1813314940716655,
+      "grad_norm": 0.5467018485069275,
+      "learning_rate": 6.0636060977746636e-05,
+      "loss": 1.6707,
+      "step": 4496
+    },
+    {
+      "epoch": 1.1818569974053272,
+      "grad_norm": 0.5809999704360962,
+      "learning_rate": 6.0618538636761876e-05,
+      "loss": 1.692,
+      "step": 4498
+    },
+    {
+      "epoch": 1.182382500738989,
+      "grad_norm": 0.6746166944503784,
+      "learning_rate": 6.060101629577711e-05,
+      "loss": 1.6944,
+      "step": 4500
+    },
+    {
+      "epoch": 1.182908004072651,
+      "grad_norm": 0.5990844368934631,
+      "learning_rate": 6.058349395479236e-05,
+      "loss": 1.703,
+      "step": 4502
+    },
+    {
+      "epoch": 1.1834335074063127,
+      "grad_norm": 0.7285276651382446,
+      "learning_rate": 6.0565971613807606e-05,
+      "loss": 1.7037,
+      "step": 4504
+    },
+    {
+      "epoch": 1.1839590107399744,
+      "grad_norm": 0.5541601181030273,
+      "learning_rate": 6.0548449272822846e-05,
+      "loss": 1.712,
+      "step": 4506
+    },
+    {
+      "epoch": 1.1844845140736362,
+      "grad_norm": 0.5341848134994507,
+      "learning_rate": 6.0530926931838094e-05,
+      "loss": 1.6848,
+      "step": 4508
+    },
+    {
+      "epoch": 1.185010017407298,
+      "grad_norm": 0.5783855319023132,
+      "learning_rate": 6.051340459085334e-05,
+      "loss": 1.6941,
+      "step": 4510
+    },
+    {
+      "epoch": 1.1855355207409597,
+      "grad_norm": 0.7894533276557922,
+      "learning_rate": 6.049588224986858e-05,
+      "loss": 1.699,
+      "step": 4512
+    },
+    {
+      "epoch": 1.1860610240746214,
+      "grad_norm": 0.6631985306739807,
+      "learning_rate": 6.047835990888383e-05,
+      "loss": 1.7063,
+      "step": 4514
+    },
+    {
+      "epoch": 1.1865865274082832,
+      "grad_norm": 0.5982935428619385,
+      "learning_rate": 6.046083756789908e-05,
+      "loss": 1.6921,
+      "step": 4516
+    },
+    {
+      "epoch": 1.187112030741945,
+      "grad_norm": 0.5341594219207764,
+      "learning_rate": 6.044331522691432e-05,
+      "loss": 1.7252,
+      "step": 4518
+    },
+    {
+      "epoch": 1.187637534075607,
+      "grad_norm": 0.5512247681617737,
+      "learning_rate": 6.0425792885929566e-05,
+      "loss": 1.7033,
+      "step": 4520
+    },
+    {
+      "epoch": 1.1881630374092687,
+      "grad_norm": 0.5994440913200378,
+      "learning_rate": 6.040827054494481e-05,
+      "loss": 1.7003,
+      "step": 4522
+    },
+    {
+      "epoch": 1.1886885407429304,
+      "grad_norm": 0.6756665706634521,
+      "learning_rate": 6.0390748203960054e-05,
+      "loss": 1.6817,
+      "step": 4524
+    },
+    {
+      "epoch": 1.1892140440765921,
+      "grad_norm": 0.6200160980224609,
+      "learning_rate": 6.037322586297529e-05,
+      "loss": 1.7068,
+      "step": 4526
+    },
+    {
+      "epoch": 1.189739547410254,
+      "grad_norm": 0.5417330861091614,
+      "learning_rate": 6.0355703521990536e-05,
+      "loss": 1.7199,
+      "step": 4528
+    },
+    {
+      "epoch": 1.1902650507439156,
+      "grad_norm": 0.6741480231285095,
+      "learning_rate": 6.033818118100578e-05,
+      "loss": 1.6577,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1907905540775774,
+      "grad_norm": 0.667620062828064,
+      "learning_rate": 6.0320658840021024e-05,
+      "loss": 1.6835,
+      "step": 4532
+    },
+    {
+      "epoch": 1.1913160574112391,
+      "grad_norm": 0.783929705619812,
+      "learning_rate": 6.030313649903627e-05,
+      "loss": 1.7028,
+      "step": 4534
+    },
+    {
+      "epoch": 1.1918415607449009,
+      "grad_norm": 0.6510748267173767,
+      "learning_rate": 6.028561415805152e-05,
+      "loss": 1.6668,
+      "step": 4536
+    },
+    {
+      "epoch": 1.1923670640785629,
+      "grad_norm": 0.5531290769577026,
+      "learning_rate": 6.026809181706676e-05,
+      "loss": 1.6691,
+      "step": 4538
+    },
+    {
+      "epoch": 1.1928925674122246,
+      "grad_norm": 0.5906832218170166,
+      "learning_rate": 6.025056947608201e-05,
+      "loss": 1.6689,
+      "step": 4540
+    },
+    {
+      "epoch": 1.1934180707458864,
+      "grad_norm": 0.580287516117096,
+      "learning_rate": 6.0233047135097255e-05,
+      "loss": 1.7105,
+      "step": 4542
+    },
+    {
+      "epoch": 1.193943574079548,
+      "grad_norm": 0.6183664798736572,
+      "learning_rate": 6.0215524794112496e-05,
+      "loss": 1.7059,
+      "step": 4544
+    },
+    {
+      "epoch": 1.1944690774132098,
+      "grad_norm": 0.5283176898956299,
+      "learning_rate": 6.019800245312774e-05,
+      "loss": 1.701,
+      "step": 4546
+    },
+    {
+      "epoch": 1.1949945807468716,
+      "grad_norm": 0.5322774052619934,
+      "learning_rate": 6.018048011214299e-05,
+      "loss": 1.6725,
+      "step": 4548
+    },
+    {
+      "epoch": 1.1955200840805333,
+      "grad_norm": 0.6731706857681274,
+      "learning_rate": 6.016295777115823e-05,
+      "loss": 1.6892,
+      "step": 4550
+    },
+    {
+      "epoch": 1.196045587414195,
+      "grad_norm": 0.5544896125793457,
+      "learning_rate": 6.0145435430173466e-05,
+      "loss": 1.6925,
+      "step": 4552
+    },
+    {
+      "epoch": 1.1965710907478568,
+      "grad_norm": 0.6916193962097168,
+      "learning_rate": 6.012791308918871e-05,
+      "loss": 1.7058,
+      "step": 4554
+    },
+    {
+      "epoch": 1.1970965940815188,
+      "grad_norm": 0.6361691355705261,
+      "learning_rate": 6.011039074820396e-05,
+      "loss": 1.6505,
+      "step": 4556
+    },
+    {
+      "epoch": 1.1976220974151806,
+      "grad_norm": 0.5423670411109924,
+      "learning_rate": 6.00928684072192e-05,
+      "loss": 1.6935,
+      "step": 4558
+    },
+    {
+      "epoch": 1.1981476007488423,
+      "grad_norm": 0.737524688243866,
+      "learning_rate": 6.007534606623445e-05,
+      "loss": 1.7217,
+      "step": 4560
+    },
+    {
+      "epoch": 1.198673104082504,
+      "grad_norm": 0.5811054110527039,
+      "learning_rate": 6.00578237252497e-05,
+      "loss": 1.6936,
+      "step": 4562
+    },
+    {
+      "epoch": 1.1991986074161658,
+      "grad_norm": 0.554607093334198,
+      "learning_rate": 6.004030138426494e-05,
+      "loss": 1.6963,
+      "step": 4564
+    },
+    {
+      "epoch": 1.1997241107498275,
+      "grad_norm": 0.6123262643814087,
+      "learning_rate": 6.0022779043280185e-05,
+      "loss": 1.6998,
+      "step": 4566
+    },
+    {
+      "epoch": 1.2002496140834893,
+      "grad_norm": 0.7441146969795227,
+      "learning_rate": 6.000525670229543e-05,
+      "loss": 1.6764,
+      "step": 4568
+    },
+    {
+      "epoch": 1.200775117417151,
+      "grad_norm": 0.6230175495147705,
+      "learning_rate": 5.9987734361310673e-05,
+      "loss": 1.6832,
+      "step": 4570
+    },
+    {
+      "epoch": 1.2013006207508128,
+      "grad_norm": 0.5986462831497192,
+      "learning_rate": 5.997021202032592e-05,
+      "loss": 1.6863,
+      "step": 4572
+    },
+    {
+      "epoch": 1.2018261240844748,
+      "grad_norm": 0.7910165786743164,
+      "learning_rate": 5.995268967934117e-05,
+      "loss": 1.6787,
+      "step": 4574
+    },
+    {
+      "epoch": 1.2023516274181365,
+      "grad_norm": 0.5927668213844299,
+      "learning_rate": 5.993516733835641e-05,
+      "loss": 1.6826,
+      "step": 4576
+    },
+    {
+      "epoch": 1.2028771307517983,
+      "grad_norm": 0.6229565143585205,
+      "learning_rate": 5.991764499737165e-05,
+      "loss": 1.6938,
+      "step": 4578
+    },
+    {
+      "epoch": 1.20340263408546,
+      "grad_norm": 0.5543463826179504,
+      "learning_rate": 5.990012265638689e-05,
+      "loss": 1.6557,
+      "step": 4580
+    },
+    {
+      "epoch": 1.2039281374191217,
+      "grad_norm": 0.6359471678733826,
+      "learning_rate": 5.988260031540214e-05,
+      "loss": 1.6984,
+      "step": 4582
+    },
+    {
+      "epoch": 1.2044536407527835,
+      "grad_norm": 0.6143317222595215,
+      "learning_rate": 5.9865077974417386e-05,
+      "loss": 1.6618,
+      "step": 4584
+    },
+    {
+      "epoch": 1.2049791440864452,
+      "grad_norm": 0.5732699036598206,
+      "learning_rate": 5.984755563343263e-05,
+      "loss": 1.7093,
+      "step": 4586
+    },
+    {
+      "epoch": 1.205504647420107,
+      "grad_norm": 0.6090834140777588,
+      "learning_rate": 5.9830033292447874e-05,
+      "loss": 1.6769,
+      "step": 4588
+    },
+    {
+      "epoch": 1.2060301507537687,
+      "grad_norm": 0.5685290098190308,
+      "learning_rate": 5.981251095146312e-05,
+      "loss": 1.689,
+      "step": 4590
+    },
+    {
+      "epoch": 1.2065556540874307,
+      "grad_norm": 0.641834557056427,
+      "learning_rate": 5.979498861047836e-05,
+      "loss": 1.6734,
+      "step": 4592
+    },
+    {
+      "epoch": 1.2070811574210925,
+      "grad_norm": 0.6050872206687927,
+      "learning_rate": 5.977746626949361e-05,
+      "loss": 1.6887,
+      "step": 4594
+    },
+    {
+      "epoch": 1.2076066607547542,
+      "grad_norm": 0.8347049951553345,
+      "learning_rate": 5.975994392850886e-05,
+      "loss": 1.7161,
+      "step": 4596
+    },
+    {
+      "epoch": 1.208132164088416,
+      "grad_norm": 0.5846490859985352,
+      "learning_rate": 5.97424215875241e-05,
+      "loss": 1.7121,
+      "step": 4598
+    },
+    {
+      "epoch": 1.2086576674220777,
+      "grad_norm": 0.8007673621177673,
+      "learning_rate": 5.9724899246539346e-05,
+      "loss": 1.6825,
+      "step": 4600
+    },
+    {
+      "epoch": 1.2091831707557394,
+      "grad_norm": 0.7300782203674316,
+      "learning_rate": 5.970737690555458e-05,
+      "loss": 1.6842,
+      "step": 4602
+    },
+    {
+      "epoch": 1.2097086740894012,
+      "grad_norm": 0.5872364044189453,
+      "learning_rate": 5.968985456456983e-05,
+      "loss": 1.6854,
+      "step": 4604
+    },
+    {
+      "epoch": 1.210234177423063,
+      "grad_norm": 0.6040078997612,
+      "learning_rate": 5.967233222358507e-05,
+      "loss": 1.7254,
+      "step": 4606
+    },
+    {
+      "epoch": 1.2107596807567247,
+      "grad_norm": 0.6352037787437439,
+      "learning_rate": 5.9654809882600316e-05,
+      "loss": 1.6559,
+      "step": 4608
+    },
+    {
+      "epoch": 1.2112851840903867,
+      "grad_norm": 0.8045578002929688,
+      "learning_rate": 5.9637287541615564e-05,
+      "loss": 1.6984,
+      "step": 4610
+    },
+    {
+      "epoch": 1.2118106874240484,
+      "grad_norm": 0.5676342248916626,
+      "learning_rate": 5.9619765200630805e-05,
+      "loss": 1.6881,
+      "step": 4612
+    },
+    {
+      "epoch": 1.2123361907577102,
+      "grad_norm": 0.5795140862464905,
+      "learning_rate": 5.960224285964605e-05,
+      "loss": 1.6856,
+      "step": 4614
+    },
+    {
+      "epoch": 1.212861694091372,
+      "grad_norm": 1.0082522630691528,
+      "learning_rate": 5.95847205186613e-05,
+      "loss": 1.7146,
+      "step": 4616
+    },
+    {
+      "epoch": 1.2133871974250336,
+      "grad_norm": 0.5609422922134399,
+      "learning_rate": 5.956719817767654e-05,
+      "loss": 1.708,
+      "step": 4618
+    },
+    {
+      "epoch": 1.2139127007586954,
+      "grad_norm": 0.6195114850997925,
+      "learning_rate": 5.954967583669179e-05,
+      "loss": 1.6505,
+      "step": 4620
+    },
+    {
+      "epoch": 1.2144382040923571,
+      "grad_norm": 0.7452388405799866,
+      "learning_rate": 5.9532153495707036e-05,
+      "loss": 1.6737,
+      "step": 4622
+    },
+    {
+      "epoch": 1.214963707426019,
+      "grad_norm": 0.624179482460022,
+      "learning_rate": 5.9514631154722276e-05,
+      "loss": 1.6642,
+      "step": 4624
+    },
+    {
+      "epoch": 1.2154892107596806,
+      "grad_norm": 0.5954496264457703,
+      "learning_rate": 5.9497108813737524e-05,
+      "loss": 1.6683,
+      "step": 4626
+    },
+    {
+      "epoch": 1.2160147140933426,
+      "grad_norm": 0.6502910852432251,
+      "learning_rate": 5.947958647275276e-05,
+      "loss": 1.6515,
+      "step": 4628
+    },
+    {
+      "epoch": 1.2165402174270044,
+      "grad_norm": 0.6815327405929565,
+      "learning_rate": 5.9462064131768005e-05,
+      "loss": 1.7016,
+      "step": 4630
+    },
+    {
+      "epoch": 1.217065720760666,
+      "grad_norm": 0.6848923563957214,
+      "learning_rate": 5.9444541790783246e-05,
+      "loss": 1.6792,
+      "step": 4632
+    },
+    {
+      "epoch": 1.2175912240943279,
+      "grad_norm": 0.5819631218910217,
+      "learning_rate": 5.9427019449798494e-05,
+      "loss": 1.6822,
+      "step": 4634
+    },
+    {
+      "epoch": 1.2181167274279896,
+      "grad_norm": 0.5571796894073486,
+      "learning_rate": 5.940949710881374e-05,
+      "loss": 1.6894,
+      "step": 4636
+    },
+    {
+      "epoch": 1.2186422307616513,
+      "grad_norm": 0.7528451681137085,
+      "learning_rate": 5.939197476782898e-05,
+      "loss": 1.6863,
+      "step": 4638
+    },
+    {
+      "epoch": 1.219167734095313,
+      "grad_norm": 0.5486443042755127,
+      "learning_rate": 5.937445242684423e-05,
+      "loss": 1.7017,
+      "step": 4640
+    },
+    {
+      "epoch": 1.219693237428975,
+      "grad_norm": 0.7798599004745483,
+      "learning_rate": 5.935693008585948e-05,
+      "loss": 1.7108,
+      "step": 4642
+    },
+    {
+      "epoch": 1.2202187407626366,
+      "grad_norm": 0.6270305514335632,
+      "learning_rate": 5.933940774487472e-05,
+      "loss": 1.6979,
+      "step": 4644
+    },
+    {
+      "epoch": 1.2207442440962986,
+      "grad_norm": 0.5500916838645935,
+      "learning_rate": 5.9321885403889966e-05,
+      "loss": 1.6832,
+      "step": 4646
+    },
+    {
+      "epoch": 1.2212697474299603,
+      "grad_norm": 0.6205405592918396,
+      "learning_rate": 5.930436306290521e-05,
+      "loss": 1.7544,
+      "step": 4648
+    },
+    {
+      "epoch": 1.221795250763622,
+      "grad_norm": 0.6169667840003967,
+      "learning_rate": 5.9286840721920454e-05,
+      "loss": 1.7024,
+      "step": 4650
+    },
+    {
+      "epoch": 1.2223207540972838,
+      "grad_norm": 0.7890437841415405,
+      "learning_rate": 5.92693183809357e-05,
+      "loss": 1.6707,
+      "step": 4652
+    },
+    {
+      "epoch": 1.2228462574309455,
+      "grad_norm": 0.6405115723609924,
+      "learning_rate": 5.9251796039950936e-05,
+      "loss": 1.6503,
+      "step": 4654
+    },
+    {
+      "epoch": 1.2233717607646073,
+      "grad_norm": 0.6368988156318665,
+      "learning_rate": 5.923427369896618e-05,
+      "loss": 1.6963,
+      "step": 4656
+    },
+    {
+      "epoch": 1.223897264098269,
+      "grad_norm": 0.7032740712165833,
+      "learning_rate": 5.9216751357981424e-05,
+      "loss": 1.6675,
+      "step": 4658
+    },
+    {
+      "epoch": 1.224422767431931,
+      "grad_norm": 0.6669613718986511,
+      "learning_rate": 5.919922901699667e-05,
+      "loss": 1.7002,
+      "step": 4660
+    },
+    {
+      "epoch": 1.2249482707655928,
+      "grad_norm": 0.6690141558647156,
+      "learning_rate": 5.918170667601192e-05,
+      "loss": 1.6499,
+      "step": 4662
+    },
+    {
+      "epoch": 1.2254737740992545,
+      "grad_norm": 0.5619009733200073,
+      "learning_rate": 5.916418433502716e-05,
+      "loss": 1.7039,
+      "step": 4664
+    },
+    {
+      "epoch": 1.2259992774329163,
+      "grad_norm": 0.6794602274894714,
+      "learning_rate": 5.914666199404241e-05,
+      "loss": 1.6732,
+      "step": 4666
+    },
+    {
+      "epoch": 1.226524780766578,
+      "grad_norm": 0.8924415707588196,
+      "learning_rate": 5.9129139653057655e-05,
+      "loss": 1.6833,
+      "step": 4668
+    },
+    {
+      "epoch": 1.2270502841002398,
+      "grad_norm": 0.6746704578399658,
+      "learning_rate": 5.9111617312072896e-05,
+      "loss": 1.6816,
+      "step": 4670
+    },
+    {
+      "epoch": 1.2275757874339015,
+      "grad_norm": 0.5335586071014404,
+      "learning_rate": 5.909409497108814e-05,
+      "loss": 1.7214,
+      "step": 4672
+    },
+    {
+      "epoch": 1.2281012907675632,
+      "grad_norm": 0.6099651455879211,
+      "learning_rate": 5.907657263010339e-05,
+      "loss": 1.6652,
+      "step": 4674
+    },
+    {
+      "epoch": 1.228626794101225,
+      "grad_norm": 0.6068762540817261,
+      "learning_rate": 5.905905028911863e-05,
+      "loss": 1.6563,
+      "step": 4676
+    },
+    {
+      "epoch": 1.229152297434887,
+      "grad_norm": 0.561000645160675,
+      "learning_rate": 5.904152794813388e-05,
+      "loss": 1.6927,
+      "step": 4678
+    },
+    {
+      "epoch": 1.2296778007685487,
+      "grad_norm": 0.5757449269294739,
+      "learning_rate": 5.902400560714911e-05,
+      "loss": 1.7042,
+      "step": 4680
+    },
+    {
+      "epoch": 1.2302033041022105,
+      "grad_norm": 0.5956955552101135,
+      "learning_rate": 5.900648326616436e-05,
+      "loss": 1.6785,
+      "step": 4682
+    },
+    {
+      "epoch": 1.2307288074358722,
+      "grad_norm": 0.5795134902000427,
+      "learning_rate": 5.89889609251796e-05,
+      "loss": 1.6774,
+      "step": 4684
+    },
+    {
+      "epoch": 1.231254310769534,
+      "grad_norm": 0.8627724051475525,
+      "learning_rate": 5.897143858419485e-05,
+      "loss": 1.7131,
+      "step": 4686
+    },
+    {
+      "epoch": 1.2317798141031957,
+      "grad_norm": 0.5390210747718811,
+      "learning_rate": 5.89539162432101e-05,
+      "loss": 1.7153,
+      "step": 4688
+    },
+    {
+      "epoch": 1.2323053174368575,
+      "grad_norm": 0.8949792981147766,
+      "learning_rate": 5.893639390222534e-05,
+      "loss": 1.689,
+      "step": 4690
+    },
+    {
+      "epoch": 1.2328308207705192,
+      "grad_norm": 0.6706136465072632,
+      "learning_rate": 5.8918871561240585e-05,
+      "loss": 1.71,
+      "step": 4692
+    },
+    {
+      "epoch": 1.233356324104181,
+      "grad_norm": 0.5577556490898132,
+      "learning_rate": 5.890134922025583e-05,
+      "loss": 1.6462,
+      "step": 4694
+    },
+    {
+      "epoch": 1.233881827437843,
+      "grad_norm": 0.6497912406921387,
+      "learning_rate": 5.8883826879271073e-05,
+      "loss": 1.6806,
+      "step": 4696
+    },
+    {
+      "epoch": 1.2344073307715047,
+      "grad_norm": 0.5242089629173279,
+      "learning_rate": 5.886630453828632e-05,
+      "loss": 1.7029,
+      "step": 4698
+    },
+    {
+      "epoch": 1.2349328341051664,
+      "grad_norm": 0.9607106447219849,
+      "learning_rate": 5.884878219730157e-05,
+      "loss": 1.6864,
+      "step": 4700
+    },
+    {
+      "epoch": 1.2354583374388282,
+      "grad_norm": 0.9370645880699158,
+      "learning_rate": 5.883125985631681e-05,
+      "loss": 1.6973,
+      "step": 4702
+    },
+    {
+      "epoch": 1.23598384077249,
+      "grad_norm": 0.7093036770820618,
+      "learning_rate": 5.881373751533204e-05,
+      "loss": 1.7237,
+      "step": 4704
+    },
+    {
+      "epoch": 1.2365093441061517,
+      "grad_norm": 0.6128284931182861,
+      "learning_rate": 5.879621517434729e-05,
+      "loss": 1.6673,
+      "step": 4706
+    },
+    {
+      "epoch": 1.2370348474398134,
+      "grad_norm": 1.1151705980300903,
+      "learning_rate": 5.877869283336254e-05,
+      "loss": 1.6668,
+      "step": 4708
+    },
+    {
+      "epoch": 1.2375603507734751,
+      "grad_norm": 0.7388045191764832,
+      "learning_rate": 5.876117049237778e-05,
+      "loss": 1.6923,
+      "step": 4710
+    },
+    {
+      "epoch": 1.238085854107137,
+      "grad_norm": 0.539486289024353,
+      "learning_rate": 5.874364815139303e-05,
+      "loss": 1.6779,
+      "step": 4712
+    },
+    {
+      "epoch": 1.2386113574407989,
+      "grad_norm": 0.7663931250572205,
+      "learning_rate": 5.8726125810408274e-05,
+      "loss": 1.686,
+      "step": 4714
+    },
+    {
+      "epoch": 1.2391368607744606,
+      "grad_norm": 0.6578073501586914,
+      "learning_rate": 5.8708603469423515e-05,
+      "loss": 1.7054,
+      "step": 4716
+    },
+    {
+      "epoch": 1.2396623641081224,
+      "grad_norm": 0.5744708776473999,
+      "learning_rate": 5.869108112843876e-05,
+      "loss": 1.6785,
+      "step": 4718
+    },
+    {
+      "epoch": 1.240187867441784,
+      "grad_norm": 0.7532958388328552,
+      "learning_rate": 5.867355878745401e-05,
+      "loss": 1.6743,
+      "step": 4720
+    },
+    {
+      "epoch": 1.2407133707754459,
+      "grad_norm": 0.5337387323379517,
+      "learning_rate": 5.865603644646925e-05,
+      "loss": 1.7303,
+      "step": 4722
+    },
+    {
+      "epoch": 1.2412388741091076,
+      "grad_norm": 0.5469831228256226,
+      "learning_rate": 5.86385141054845e-05,
+      "loss": 1.6566,
+      "step": 4724
+    },
+    {
+      "epoch": 1.2417643774427694,
+      "grad_norm": 0.5634649395942688,
+      "learning_rate": 5.8620991764499746e-05,
+      "loss": 1.6868,
+      "step": 4726
+    },
+    {
+      "epoch": 1.242289880776431,
+      "grad_norm": 0.7761363983154297,
+      "learning_rate": 5.860346942351499e-05,
+      "loss": 1.6792,
+      "step": 4728
+    },
+    {
+      "epoch": 1.2428153841100928,
+      "grad_norm": 0.6965597867965698,
+      "learning_rate": 5.858594708253022e-05,
+      "loss": 1.715,
+      "step": 4730
+    },
+    {
+      "epoch": 1.2433408874437548,
+      "grad_norm": 0.6957322359085083,
+      "learning_rate": 5.856842474154547e-05,
+      "loss": 1.6739,
+      "step": 4732
+    },
+    {
+      "epoch": 1.2438663907774166,
+      "grad_norm": 0.5814897418022156,
+      "learning_rate": 5.8550902400560716e-05,
+      "loss": 1.6988,
+      "step": 4734
+    },
+    {
+      "epoch": 1.2443918941110783,
+      "grad_norm": 0.9491192698478699,
+      "learning_rate": 5.853338005957596e-05,
+      "loss": 1.6994,
+      "step": 4736
+    },
+    {
+      "epoch": 1.24491739744474,
+      "grad_norm": 0.7140454649925232,
+      "learning_rate": 5.8515857718591205e-05,
+      "loss": 1.6665,
+      "step": 4738
+    },
+    {
+      "epoch": 1.2454429007784018,
+      "grad_norm": 0.5758277773857117,
+      "learning_rate": 5.849833537760645e-05,
+      "loss": 1.6842,
+      "step": 4740
+    },
+    {
+      "epoch": 1.2459684041120636,
+      "grad_norm": 0.5660161972045898,
+      "learning_rate": 5.848081303662169e-05,
+      "loss": 1.6844,
+      "step": 4742
+    },
+    {
+      "epoch": 1.2464939074457253,
+      "grad_norm": 0.7871302962303162,
+      "learning_rate": 5.846329069563694e-05,
+      "loss": 1.6983,
+      "step": 4744
+    },
+    {
+      "epoch": 1.247019410779387,
+      "grad_norm": 0.5468810200691223,
+      "learning_rate": 5.844576835465219e-05,
+      "loss": 1.6561,
+      "step": 4746
+    },
+    {
+      "epoch": 1.2475449141130488,
+      "grad_norm": 0.545019268989563,
+      "learning_rate": 5.842824601366743e-05,
+      "loss": 1.6704,
+      "step": 4748
+    },
+    {
+      "epoch": 1.2480704174467108,
+      "grad_norm": 0.5376728177070618,
+      "learning_rate": 5.8410723672682676e-05,
+      "loss": 1.6373,
+      "step": 4750
+    },
+    {
+      "epoch": 1.2485959207803725,
+      "grad_norm": 0.6872471570968628,
+      "learning_rate": 5.8393201331697924e-05,
+      "loss": 1.6629,
+      "step": 4752
+    },
+    {
+      "epoch": 1.2491214241140343,
+      "grad_norm": 0.6468409299850464,
+      "learning_rate": 5.8375678990713165e-05,
+      "loss": 1.6785,
+      "step": 4754
+    },
+    {
+      "epoch": 1.249646927447696,
+      "grad_norm": 0.7774249315261841,
+      "learning_rate": 5.83581566497284e-05,
+      "loss": 1.642,
+      "step": 4756
+    },
+    {
+      "epoch": 1.2501724307813578,
+      "grad_norm": 0.6062743067741394,
+      "learning_rate": 5.8340634308743646e-05,
+      "loss": 1.7039,
+      "step": 4758
+    },
+    {
+      "epoch": 1.2506979341150195,
+      "grad_norm": 0.6170024871826172,
+      "learning_rate": 5.8323111967758894e-05,
+      "loss": 1.7296,
+      "step": 4760
+    },
+    {
+      "epoch": 1.2512234374486813,
+      "grad_norm": 0.5989388823509216,
+      "learning_rate": 5.8305589626774135e-05,
+      "loss": 1.6776,
+      "step": 4762
+    },
+    {
+      "epoch": 1.2517489407823432,
+      "grad_norm": 0.6566706895828247,
+      "learning_rate": 5.828806728578938e-05,
+      "loss": 1.7016,
+      "step": 4764
+    },
+    {
+      "epoch": 1.2522744441160047,
+      "grad_norm": 0.6715628504753113,
+      "learning_rate": 5.827054494480463e-05,
+      "loss": 1.6946,
+      "step": 4766
+    },
+    {
+      "epoch": 1.2527999474496667,
+      "grad_norm": 0.7730180621147156,
+      "learning_rate": 5.825302260381987e-05,
+      "loss": 1.7043,
+      "step": 4768
+    },
+    {
+      "epoch": 1.2533254507833285,
+      "grad_norm": 0.5837029814720154,
+      "learning_rate": 5.823550026283512e-05,
+      "loss": 1.7019,
+      "step": 4770
+    },
+    {
+      "epoch": 1.2538509541169902,
+      "grad_norm": 0.6363146901130676,
+      "learning_rate": 5.8217977921850366e-05,
+      "loss": 1.6965,
+      "step": 4772
+    },
+    {
+      "epoch": 1.254376457450652,
+      "grad_norm": 0.6856462359428406,
+      "learning_rate": 5.8200455580865606e-05,
+      "loss": 1.6899,
+      "step": 4774
+    },
+    {
+      "epoch": 1.2549019607843137,
+      "grad_norm": 0.7007546424865723,
+      "learning_rate": 5.8182933239880854e-05,
+      "loss": 1.6673,
+      "step": 4776
+    },
+    {
+      "epoch": 1.2554274641179755,
+      "grad_norm": 0.6396967172622681,
+      "learning_rate": 5.81654108988961e-05,
+      "loss": 1.6826,
+      "step": 4778
+    },
+    {
+      "epoch": 1.2559529674516372,
+      "grad_norm": 0.6191834211349487,
+      "learning_rate": 5.814788855791134e-05,
+      "loss": 1.6848,
+      "step": 4780
+    },
+    {
+      "epoch": 1.2564784707852992,
+      "grad_norm": 0.7049083709716797,
+      "learning_rate": 5.8130366216926576e-05,
+      "loss": 1.6785,
+      "step": 4782
+    },
+    {
+      "epoch": 1.2570039741189607,
+      "grad_norm": 0.5608694553375244,
+      "learning_rate": 5.8112843875941824e-05,
+      "loss": 1.7138,
+      "step": 4784
+    },
+    {
+      "epoch": 1.2575294774526227,
+      "grad_norm": 0.5731740593910217,
+      "learning_rate": 5.809532153495707e-05,
+      "loss": 1.6912,
+      "step": 4786
+    },
+    {
+      "epoch": 1.2580549807862844,
+      "grad_norm": 0.5545026063919067,
+      "learning_rate": 5.807779919397231e-05,
+      "loss": 1.6621,
+      "step": 4788
+    },
+    {
+      "epoch": 1.2585804841199462,
+      "grad_norm": 0.5389024019241333,
+      "learning_rate": 5.806027685298756e-05,
+      "loss": 1.6956,
+      "step": 4790
+    },
+    {
+      "epoch": 1.259105987453608,
+      "grad_norm": 0.5764096975326538,
+      "learning_rate": 5.804275451200281e-05,
+      "loss": 1.6556,
+      "step": 4792
+    },
+    {
+      "epoch": 1.2596314907872697,
+      "grad_norm": 0.5584897398948669,
+      "learning_rate": 5.802523217101805e-05,
+      "loss": 1.6837,
+      "step": 4794
+    },
+    {
+      "epoch": 1.2601569941209314,
+      "grad_norm": 0.6685154438018799,
+      "learning_rate": 5.8007709830033296e-05,
+      "loss": 1.6343,
+      "step": 4796
+    },
+    {
+      "epoch": 1.2606824974545932,
+      "grad_norm": 0.6008186936378479,
+      "learning_rate": 5.799018748904854e-05,
+      "loss": 1.6674,
+      "step": 4798
+    },
+    {
+      "epoch": 1.2612080007882551,
+      "grad_norm": 0.6519007086753845,
+      "learning_rate": 5.7972665148063784e-05,
+      "loss": 1.6898,
+      "step": 4800
+    },
+    {
+      "epoch": 1.2612080007882551,
+      "eval_loss": 1.700656533241272,
+      "eval_runtime": 487.1424,
+      "eval_samples_per_second": 250.007,
+      "eval_steps_per_second": 31.252,
+      "step": 4800
+    },
+    {
+      "epoch": 1.2617335041219166,
+      "grad_norm": 0.5863731503486633,
+      "learning_rate": 5.795514280707903e-05,
+      "loss": 1.6239,
+      "step": 4802
+    },
+    {
+      "epoch": 1.2622590074555786,
+      "grad_norm": 0.6118308901786804,
+      "learning_rate": 5.793762046609428e-05,
+      "loss": 1.6997,
+      "step": 4804
+    },
+    {
+      "epoch": 1.2627845107892404,
+      "grad_norm": 0.7632036209106445,
+      "learning_rate": 5.792009812510951e-05,
+      "loss": 1.6626,
+      "step": 4806
+    },
+    {
+      "epoch": 1.2633100141229021,
+      "grad_norm": 0.593360960483551,
+      "learning_rate": 5.7902575784124754e-05,
+      "loss": 1.6919,
+      "step": 4808
+    },
+    {
+      "epoch": 1.2638355174565639,
+      "grad_norm": 0.8928617238998413,
+      "learning_rate": 5.788505344314e-05,
+      "loss": 1.7171,
+      "step": 4810
+    },
+    {
+      "epoch": 1.2643610207902256,
+      "grad_norm": 0.6598185896873474,
+      "learning_rate": 5.786753110215525e-05,
+      "loss": 1.6581,
+      "step": 4812
+    },
+    {
+      "epoch": 1.2648865241238874,
+      "grad_norm": 0.6565474271774292,
+      "learning_rate": 5.785000876117049e-05,
+      "loss": 1.6763,
+      "step": 4814
+    },
+    {
+      "epoch": 1.265412027457549,
+      "grad_norm": 0.796642541885376,
+      "learning_rate": 5.783248642018574e-05,
+      "loss": 1.6925,
+      "step": 4816
+    },
+    {
+      "epoch": 1.265937530791211,
+      "grad_norm": 0.517112135887146,
+      "learning_rate": 5.7814964079200985e-05,
+      "loss": 1.7095,
+      "step": 4818
+    },
+    {
+      "epoch": 1.2664630341248726,
+      "grad_norm": 0.6834176182746887,
+      "learning_rate": 5.7797441738216226e-05,
+      "loss": 1.6919,
+      "step": 4820
+    },
+    {
+      "epoch": 1.2669885374585346,
+      "grad_norm": 0.5728265643119812,
+      "learning_rate": 5.7779919397231473e-05,
+      "loss": 1.6796,
+      "step": 4822
+    },
+    {
+      "epoch": 1.2675140407921963,
+      "grad_norm": 0.8350217342376709,
+      "learning_rate": 5.776239705624672e-05,
+      "loss": 1.6946,
+      "step": 4824
+    },
+    {
+      "epoch": 1.268039544125858,
+      "grad_norm": 0.7911032438278198,
+      "learning_rate": 5.774487471526196e-05,
+      "loss": 1.705,
+      "step": 4826
+    },
+    {
+      "epoch": 1.2685650474595198,
+      "grad_norm": 0.6383318305015564,
+      "learning_rate": 5.772735237427721e-05,
+      "loss": 1.6945,
+      "step": 4828
+    },
+    {
+      "epoch": 1.2690905507931816,
+      "grad_norm": 0.7389723062515259,
+      "learning_rate": 5.770983003329246e-05,
+      "loss": 1.6889,
+      "step": 4830
+    },
+    {
+      "epoch": 1.2696160541268433,
+      "grad_norm": 0.5114458203315735,
+      "learning_rate": 5.769230769230769e-05,
+      "loss": 1.69,
+      "step": 4832
+    },
+    {
+      "epoch": 1.270141557460505,
+      "grad_norm": 0.7528431415557861,
+      "learning_rate": 5.767478535132294e-05,
+      "loss": 1.6829,
+      "step": 4834
+    },
+    {
+      "epoch": 1.270667060794167,
+      "grad_norm": 0.6436894536018372,
+      "learning_rate": 5.765726301033818e-05,
+      "loss": 1.6968,
+      "step": 4836
+    },
+    {
+      "epoch": 1.2711925641278286,
+      "grad_norm": 0.601234495639801,
+      "learning_rate": 5.763974066935343e-05,
+      "loss": 1.6826,
+      "step": 4838
+    },
+    {
+      "epoch": 1.2717180674614905,
+      "grad_norm": 0.6918678283691406,
+      "learning_rate": 5.7622218328368674e-05,
+      "loss": 1.698,
+      "step": 4840
+    },
+    {
+      "epoch": 1.2722435707951523,
+      "grad_norm": 0.5951560139656067,
+      "learning_rate": 5.7604695987383915e-05,
+      "loss": 1.7462,
+      "step": 4842
+    },
+    {
+      "epoch": 1.272769074128814,
+      "grad_norm": 0.8118217587471008,
+      "learning_rate": 5.758717364639916e-05,
+      "loss": 1.6818,
+      "step": 4844
+    },
+    {
+      "epoch": 1.2732945774624758,
+      "grad_norm": 0.5995089411735535,
+      "learning_rate": 5.756965130541441e-05,
+      "loss": 1.7061,
+      "step": 4846
+    },
+    {
+      "epoch": 1.2738200807961375,
+      "grad_norm": 0.5712124705314636,
+      "learning_rate": 5.755212896442965e-05,
+      "loss": 1.6635,
+      "step": 4848
+    },
+    {
+      "epoch": 1.2743455841297993,
+      "grad_norm": 0.975643515586853,
+      "learning_rate": 5.75346066234449e-05,
+      "loss": 1.6871,
+      "step": 4850
+    },
+    {
+      "epoch": 1.274871087463461,
+      "grad_norm": 0.6375470161437988,
+      "learning_rate": 5.7517084282460146e-05,
+      "loss": 1.6946,
+      "step": 4852
+    },
+    {
+      "epoch": 1.275396590797123,
+      "grad_norm": 0.6874341368675232,
+      "learning_rate": 5.749956194147539e-05,
+      "loss": 1.6987,
+      "step": 4854
+    },
+    {
+      "epoch": 1.2759220941307847,
+      "grad_norm": 0.5625795125961304,
+      "learning_rate": 5.7482039600490635e-05,
+      "loss": 1.6946,
+      "step": 4856
+    },
+    {
+      "epoch": 1.2764475974644465,
+      "grad_norm": 0.7773122191429138,
+      "learning_rate": 5.746451725950587e-05,
+      "loss": 1.6922,
+      "step": 4858
+    },
+    {
+      "epoch": 1.2769731007981082,
+      "grad_norm": 0.6410396099090576,
+      "learning_rate": 5.7446994918521116e-05,
+      "loss": 1.676,
+      "step": 4860
+    },
+    {
+      "epoch": 1.27749860413177,
+      "grad_norm": 0.5476670861244202,
+      "learning_rate": 5.742947257753636e-05,
+      "loss": 1.7063,
+      "step": 4862
+    },
+    {
+      "epoch": 1.2780241074654317,
+      "grad_norm": 0.5646496415138245,
+      "learning_rate": 5.7411950236551604e-05,
+      "loss": 1.6657,
+      "step": 4864
+    },
+    {
+      "epoch": 1.2785496107990935,
+      "grad_norm": 0.7577764987945557,
+      "learning_rate": 5.739442789556685e-05,
+      "loss": 1.6869,
+      "step": 4866
+    },
+    {
+      "epoch": 1.2790751141327552,
+      "grad_norm": 0.5627464652061462,
+      "learning_rate": 5.737690555458209e-05,
+      "loss": 1.6608,
+      "step": 4868
+    },
+    {
+      "epoch": 1.279600617466417,
+      "grad_norm": 0.5386861562728882,
+      "learning_rate": 5.735938321359734e-05,
+      "loss": 1.6826,
+      "step": 4870
+    },
+    {
+      "epoch": 1.280126120800079,
+      "grad_norm": 0.8496121168136597,
+      "learning_rate": 5.734186087261259e-05,
+      "loss": 1.6771,
+      "step": 4872
+    },
+    {
+      "epoch": 1.2806516241337407,
+      "grad_norm": 0.6117552518844604,
+      "learning_rate": 5.732433853162783e-05,
+      "loss": 1.6957,
+      "step": 4874
+    },
+    {
+      "epoch": 1.2811771274674024,
+      "grad_norm": 0.7759193181991577,
+      "learning_rate": 5.7306816190643076e-05,
+      "loss": 1.6396,
+      "step": 4876
+    },
+    {
+      "epoch": 1.2817026308010642,
+      "grad_norm": 0.6560829281806946,
+      "learning_rate": 5.7289293849658324e-05,
+      "loss": 1.6983,
+      "step": 4878
+    },
+    {
+      "epoch": 1.282228134134726,
+      "grad_norm": 0.6807093024253845,
+      "learning_rate": 5.7271771508673565e-05,
+      "loss": 1.6545,
+      "step": 4880
+    },
+    {
+      "epoch": 1.2827536374683877,
+      "grad_norm": 0.6546826958656311,
+      "learning_rate": 5.72542491676888e-05,
+      "loss": 1.7138,
+      "step": 4882
+    },
+    {
+      "epoch": 1.2832791408020494,
+      "grad_norm": 0.5959972739219666,
+      "learning_rate": 5.7236726826704046e-05,
+      "loss": 1.6567,
+      "step": 4884
+    },
+    {
+      "epoch": 1.2838046441357112,
+      "grad_norm": 0.6360346078872681,
+      "learning_rate": 5.7219204485719294e-05,
+      "loss": 1.6764,
+      "step": 4886
+    },
+    {
+      "epoch": 1.284330147469373,
+      "grad_norm": 0.7171315550804138,
+      "learning_rate": 5.7201682144734535e-05,
+      "loss": 1.6489,
+      "step": 4888
+    },
+    {
+      "epoch": 1.2848556508030349,
+      "grad_norm": 0.6501976251602173,
+      "learning_rate": 5.718415980374978e-05,
+      "loss": 1.7136,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2853811541366966,
+      "grad_norm": 0.7134879231452942,
+      "learning_rate": 5.716663746276503e-05,
+      "loss": 1.6789,
+      "step": 4892
+    },
+    {
+      "epoch": 1.2859066574703584,
+      "grad_norm": 0.6386318206787109,
+      "learning_rate": 5.714911512178027e-05,
+      "loss": 1.6877,
+      "step": 4894
+    },
+    {
+      "epoch": 1.2864321608040201,
+      "grad_norm": 0.5816686749458313,
+      "learning_rate": 5.713159278079552e-05,
+      "loss": 1.7148,
+      "step": 4896
+    },
+    {
+      "epoch": 1.2869576641376819,
+      "grad_norm": 0.49736329913139343,
+      "learning_rate": 5.7114070439810766e-05,
+      "loss": 1.6591,
+      "step": 4898
+    },
+    {
+      "epoch": 1.2874831674713436,
+      "grad_norm": 0.5454695820808411,
+      "learning_rate": 5.7096548098826006e-05,
+      "loss": 1.7062,
+      "step": 4900
+    },
+    {
+      "epoch": 1.2880086708050054,
+      "grad_norm": 0.5578910708427429,
+      "learning_rate": 5.7079025757841254e-05,
+      "loss": 1.6773,
+      "step": 4902
+    },
+    {
+      "epoch": 1.288534174138667,
+      "grad_norm": 0.7455135583877563,
+      "learning_rate": 5.70615034168565e-05,
+      "loss": 1.6605,
+      "step": 4904
+    },
+    {
+      "epoch": 1.2890596774723289,
+      "grad_norm": 0.6031019687652588,
+      "learning_rate": 5.704398107587174e-05,
+      "loss": 1.706,
+      "step": 4906
+    },
+    {
+      "epoch": 1.2895851808059908,
+      "grad_norm": 0.6043726801872253,
+      "learning_rate": 5.7026458734886976e-05,
+      "loss": 1.6919,
+      "step": 4908
+    },
+    {
+      "epoch": 1.2901106841396526,
+      "grad_norm": 0.6083821654319763,
+      "learning_rate": 5.7008936393902224e-05,
+      "loss": 1.6752,
+      "step": 4910
+    },
+    {
+      "epoch": 1.2906361874733143,
+      "grad_norm": 0.5966817736625671,
+      "learning_rate": 5.699141405291747e-05,
+      "loss": 1.6838,
+      "step": 4912
+    },
+    {
+      "epoch": 1.291161690806976,
+      "grad_norm": 0.5941465497016907,
+      "learning_rate": 5.697389171193271e-05,
+      "loss": 1.6899,
+      "step": 4914
+    },
+    {
+      "epoch": 1.2916871941406378,
+      "grad_norm": 0.6362673044204712,
+      "learning_rate": 5.695636937094796e-05,
+      "loss": 1.684,
+      "step": 4916
+    },
+    {
+      "epoch": 1.2922126974742996,
+      "grad_norm": 0.6817765831947327,
+      "learning_rate": 5.693884702996321e-05,
+      "loss": 1.668,
+      "step": 4918
+    },
+    {
+      "epoch": 1.2927382008079613,
+      "grad_norm": 0.6063424944877625,
+      "learning_rate": 5.692132468897845e-05,
+      "loss": 1.6628,
+      "step": 4920
+    },
+    {
+      "epoch": 1.2932637041416233,
+      "grad_norm": 0.595859169960022,
+      "learning_rate": 5.6903802347993696e-05,
+      "loss": 1.6989,
+      "step": 4922
+    },
+    {
+      "epoch": 1.2937892074752848,
+      "grad_norm": 0.5616616606712341,
+      "learning_rate": 5.688628000700894e-05,
+      "loss": 1.7036,
+      "step": 4924
+    },
+    {
+      "epoch": 1.2943147108089468,
+      "grad_norm": 0.5995229482650757,
+      "learning_rate": 5.6868757666024184e-05,
+      "loss": 1.6894,
+      "step": 4926
+    },
+    {
+      "epoch": 1.2948402141426085,
+      "grad_norm": 0.5990728735923767,
+      "learning_rate": 5.685123532503943e-05,
+      "loss": 1.6612,
+      "step": 4928
+    },
+    {
+      "epoch": 1.2953657174762703,
+      "grad_norm": 0.5425363183021545,
+      "learning_rate": 5.683371298405468e-05,
+      "loss": 1.6523,
+      "step": 4930
+    },
+    {
+      "epoch": 1.295891220809932,
+      "grad_norm": 0.5370776653289795,
+      "learning_rate": 5.681619064306992e-05,
+      "loss": 1.6977,
+      "step": 4932
+    },
+    {
+      "epoch": 1.2964167241435938,
+      "grad_norm": 0.6286599636077881,
+      "learning_rate": 5.6798668302085154e-05,
+      "loss": 1.7288,
+      "step": 4934
+    },
+    {
+      "epoch": 1.2969422274772555,
+      "grad_norm": 0.575813889503479,
+      "learning_rate": 5.67811459611004e-05,
+      "loss": 1.6914,
+      "step": 4936
+    },
+    {
+      "epoch": 1.2974677308109173,
+      "grad_norm": 0.5949034690856934,
+      "learning_rate": 5.676362362011565e-05,
+      "loss": 1.6733,
+      "step": 4938
+    },
+    {
+      "epoch": 1.2979932341445792,
+      "grad_norm": 0.6935321688652039,
+      "learning_rate": 5.674610127913089e-05,
+      "loss": 1.6886,
+      "step": 4940
+    },
+    {
+      "epoch": 1.2985187374782408,
+      "grad_norm": 0.6900460124015808,
+      "learning_rate": 5.672857893814614e-05,
+      "loss": 1.6985,
+      "step": 4942
+    },
+    {
+      "epoch": 1.2990442408119027,
+      "grad_norm": 0.5892564654350281,
+      "learning_rate": 5.6711056597161385e-05,
+      "loss": 1.6837,
+      "step": 4944
+    },
+    {
+      "epoch": 1.2995697441455645,
+      "grad_norm": 0.6516983509063721,
+      "learning_rate": 5.6693534256176626e-05,
+      "loss": 1.6797,
+      "step": 4946
+    },
+    {
+      "epoch": 1.3000952474792262,
+      "grad_norm": 0.8544719219207764,
+      "learning_rate": 5.6676011915191873e-05,
+      "loss": 1.6911,
+      "step": 4948
+    },
+    {
+      "epoch": 1.300620750812888,
+      "grad_norm": 0.7540829181671143,
+      "learning_rate": 5.665848957420712e-05,
+      "loss": 1.7333,
+      "step": 4950
+    },
+    {
+      "epoch": 1.3011462541465497,
+      "grad_norm": 0.8198524117469788,
+      "learning_rate": 5.664096723322236e-05,
+      "loss": 1.7286,
+      "step": 4952
+    },
+    {
+      "epoch": 1.3016717574802115,
+      "grad_norm": 0.6572164297103882,
+      "learning_rate": 5.662344489223761e-05,
+      "loss": 1.7171,
+      "step": 4954
+    },
+    {
+      "epoch": 1.3021972608138732,
+      "grad_norm": 0.6047423481941223,
+      "learning_rate": 5.660592255125286e-05,
+      "loss": 1.6762,
+      "step": 4956
+    },
+    {
+      "epoch": 1.3027227641475352,
+      "grad_norm": 0.6690229177474976,
+      "learning_rate": 5.65884002102681e-05,
+      "loss": 1.6802,
+      "step": 4958
+    },
+    {
+      "epoch": 1.3032482674811967,
+      "grad_norm": 0.57923823595047,
+      "learning_rate": 5.657087786928333e-05,
+      "loss": 1.6864,
+      "step": 4960
+    },
+    {
+      "epoch": 1.3037737708148587,
+      "grad_norm": 0.6738536357879639,
+      "learning_rate": 5.655335552829858e-05,
+      "loss": 1.6883,
+      "step": 4962
+    },
+    {
+      "epoch": 1.3042992741485204,
+      "grad_norm": 0.5950745344161987,
+      "learning_rate": 5.653583318731383e-05,
+      "loss": 1.6688,
+      "step": 4964
+    },
+    {
+      "epoch": 1.3048247774821822,
+      "grad_norm": 0.5781100988388062,
+      "learning_rate": 5.651831084632907e-05,
+      "loss": 1.699,
+      "step": 4966
+    },
+    {
+      "epoch": 1.305350280815844,
+      "grad_norm": 0.535622775554657,
+      "learning_rate": 5.6500788505344315e-05,
+      "loss": 1.6881,
+      "step": 4968
+    },
+    {
+      "epoch": 1.3058757841495057,
+      "grad_norm": 0.6280853152275085,
+      "learning_rate": 5.648326616435956e-05,
+      "loss": 1.6802,
+      "step": 4970
+    },
+    {
+      "epoch": 1.3064012874831674,
+      "grad_norm": 0.538373589515686,
+      "learning_rate": 5.6465743823374803e-05,
+      "loss": 1.6446,
+      "step": 4972
+    },
+    {
+      "epoch": 1.3069267908168292,
+      "grad_norm": 0.5861368775367737,
+      "learning_rate": 5.644822148239005e-05,
+      "loss": 1.6667,
+      "step": 4974
+    },
+    {
+      "epoch": 1.3074522941504911,
+      "grad_norm": 0.6625964045524597,
+      "learning_rate": 5.64306991414053e-05,
+      "loss": 1.6879,
+      "step": 4976
+    },
+    {
+      "epoch": 1.3079777974841527,
+      "grad_norm": 0.6457346081733704,
+      "learning_rate": 5.641317680042054e-05,
+      "loss": 1.688,
+      "step": 4978
+    },
+    {
+      "epoch": 1.3085033008178146,
+      "grad_norm": 0.6461663246154785,
+      "learning_rate": 5.639565445943579e-05,
+      "loss": 1.6974,
+      "step": 4980
+    },
+    {
+      "epoch": 1.3090288041514764,
+      "grad_norm": 0.6430699825286865,
+      "learning_rate": 5.6378132118451035e-05,
+      "loss": 1.7059,
+      "step": 4982
+    },
+    {
+      "epoch": 1.3095543074851381,
+      "grad_norm": 0.6152409315109253,
+      "learning_rate": 5.636060977746627e-05,
+      "loss": 1.649,
+      "step": 4984
+    },
+    {
+      "epoch": 1.3100798108187999,
+      "grad_norm": 0.5380004644393921,
+      "learning_rate": 5.634308743648151e-05,
+      "loss": 1.6893,
+      "step": 4986
+    },
+    {
+      "epoch": 1.3106053141524616,
+      "grad_norm": 0.6044327020645142,
+      "learning_rate": 5.632556509549676e-05,
+      "loss": 1.714,
+      "step": 4988
+    },
+    {
+      "epoch": 1.3111308174861234,
+      "grad_norm": 0.5950967073440552,
+      "learning_rate": 5.6308042754512004e-05,
+      "loss": 1.695,
+      "step": 4990
+    },
+    {
+      "epoch": 1.3116563208197851,
+      "grad_norm": 0.5887439846992493,
+      "learning_rate": 5.6290520413527245e-05,
+      "loss": 1.6914,
+      "step": 4992
+    },
+    {
+      "epoch": 1.312181824153447,
+      "grad_norm": 0.6020603775978088,
+      "learning_rate": 5.627299807254249e-05,
+      "loss": 1.6847,
+      "step": 4994
+    },
+    {
+      "epoch": 1.3127073274871086,
+      "grad_norm": 0.6895347237586975,
+      "learning_rate": 5.625547573155774e-05,
+      "loss": 1.6888,
+      "step": 4996
+    },
+    {
+      "epoch": 1.3132328308207706,
+      "grad_norm": 0.7459006309509277,
+      "learning_rate": 5.623795339057298e-05,
+      "loss": 1.6828,
+      "step": 4998
+    },
+    {
+      "epoch": 1.3137583341544323,
+      "grad_norm": 0.528291642665863,
+      "learning_rate": 5.622043104958823e-05,
+      "loss": 1.6612,
+      "step": 5000
+    },
+    {
+      "epoch": 1.314283837488094,
+      "grad_norm": 0.5710020065307617,
+      "learning_rate": 5.6202908708603476e-05,
+      "loss": 1.7157,
+      "step": 5002
+    },
+    {
+      "epoch": 1.3148093408217558,
+      "grad_norm": 0.7354927062988281,
+      "learning_rate": 5.618538636761872e-05,
+      "loss": 1.6948,
+      "step": 5004
+    },
+    {
+      "epoch": 1.3153348441554176,
+      "grad_norm": 0.5487669110298157,
+      "learning_rate": 5.6167864026633965e-05,
+      "loss": 1.6546,
+      "step": 5006
+    },
+    {
+      "epoch": 1.3158603474890793,
+      "grad_norm": 0.5558237433433533,
+      "learning_rate": 5.615034168564921e-05,
+      "loss": 1.6729,
+      "step": 5008
+    },
+    {
+      "epoch": 1.316385850822741,
+      "grad_norm": 0.6180663704872131,
+      "learning_rate": 5.6132819344664446e-05,
+      "loss": 1.672,
+      "step": 5010
+    },
+    {
+      "epoch": 1.316911354156403,
+      "grad_norm": 0.5914913415908813,
+      "learning_rate": 5.611529700367969e-05,
+      "loss": 1.6727,
+      "step": 5012
+    },
+    {
+      "epoch": 1.3174368574900648,
+      "grad_norm": 0.6253445148468018,
+      "learning_rate": 5.6097774662694935e-05,
+      "loss": 1.6805,
+      "step": 5014
+    },
+    {
+      "epoch": 1.3179623608237265,
+      "grad_norm": 0.755477249622345,
+      "learning_rate": 5.608025232171018e-05,
+      "loss": 1.6646,
+      "step": 5016
+    },
+    {
+      "epoch": 1.3184878641573883,
+      "grad_norm": 0.5822807550430298,
+      "learning_rate": 5.606272998072542e-05,
+      "loss": 1.6779,
+      "step": 5018
+    },
+    {
+      "epoch": 1.31901336749105,
+      "grad_norm": 0.5383006930351257,
+      "learning_rate": 5.604520763974067e-05,
+      "loss": 1.6942,
+      "step": 5020
+    },
+    {
+      "epoch": 1.3195388708247118,
+      "grad_norm": 0.5826961398124695,
+      "learning_rate": 5.602768529875592e-05,
+      "loss": 1.6874,
+      "step": 5022
+    },
+    {
+      "epoch": 1.3200643741583735,
+      "grad_norm": 0.5648834705352783,
+      "learning_rate": 5.601016295777116e-05,
+      "loss": 1.6948,
+      "step": 5024
+    },
+    {
+      "epoch": 1.3205898774920353,
+      "grad_norm": 0.6096558570861816,
+      "learning_rate": 5.5992640616786406e-05,
+      "loss": 1.6924,
+      "step": 5026
+    },
+    {
+      "epoch": 1.321115380825697,
+      "grad_norm": 0.5826833844184875,
+      "learning_rate": 5.5975118275801654e-05,
+      "loss": 1.7116,
+      "step": 5028
+    },
+    {
+      "epoch": 1.321640884159359,
+      "grad_norm": 0.5931800007820129,
+      "learning_rate": 5.5957595934816895e-05,
+      "loss": 1.6996,
+      "step": 5030
+    },
+    {
+      "epoch": 1.3221663874930207,
+      "grad_norm": 0.6546478271484375,
+      "learning_rate": 5.594007359383214e-05,
+      "loss": 1.6639,
+      "step": 5032
+    },
+    {
+      "epoch": 1.3226918908266825,
+      "grad_norm": 0.6511355042457581,
+      "learning_rate": 5.592255125284739e-05,
+      "loss": 1.6912,
+      "step": 5034
+    },
+    {
+      "epoch": 1.3232173941603442,
+      "grad_norm": 0.57513028383255,
+      "learning_rate": 5.5905028911862624e-05,
+      "loss": 1.6479,
+      "step": 5036
+    },
+    {
+      "epoch": 1.323742897494006,
+      "grad_norm": 0.5553883910179138,
+      "learning_rate": 5.5887506570877865e-05,
+      "loss": 1.6877,
+      "step": 5038
+    },
+    {
+      "epoch": 1.3242684008276677,
+      "grad_norm": 0.6477178931236267,
+      "learning_rate": 5.586998422989311e-05,
+      "loss": 1.7209,
+      "step": 5040
+    },
+    {
+      "epoch": 1.3247939041613295,
+      "grad_norm": 0.6500737071037292,
+      "learning_rate": 5.585246188890836e-05,
+      "loss": 1.6838,
+      "step": 5042
+    },
+    {
+      "epoch": 1.3253194074949912,
+      "grad_norm": 0.499759703874588,
+      "learning_rate": 5.58349395479236e-05,
+      "loss": 1.6858,
+      "step": 5044
+    },
+    {
+      "epoch": 1.325844910828653,
+      "grad_norm": 0.6388137340545654,
+      "learning_rate": 5.581741720693885e-05,
+      "loss": 1.6836,
+      "step": 5046
+    },
+    {
+      "epoch": 1.326370414162315,
+      "grad_norm": 0.7066943049430847,
+      "learning_rate": 5.5799894865954096e-05,
+      "loss": 1.7018,
+      "step": 5048
+    },
+    {
+      "epoch": 1.3268959174959767,
+      "grad_norm": 0.5318599343299866,
+      "learning_rate": 5.5782372524969336e-05,
+      "loss": 1.676,
+      "step": 5050
+    },
+    {
+      "epoch": 1.3274214208296384,
+      "grad_norm": 0.6772581338882446,
+      "learning_rate": 5.5764850183984584e-05,
+      "loss": 1.6909,
+      "step": 5052
+    },
+    {
+      "epoch": 1.3279469241633002,
+      "grad_norm": 0.6018180847167969,
+      "learning_rate": 5.574732784299983e-05,
+      "loss": 1.6817,
+      "step": 5054
+    },
+    {
+      "epoch": 1.328472427496962,
+      "grad_norm": 0.5764731168746948,
+      "learning_rate": 5.572980550201507e-05,
+      "loss": 1.6866,
+      "step": 5056
+    },
+    {
+      "epoch": 1.3289979308306237,
+      "grad_norm": 0.613559365272522,
+      "learning_rate": 5.571228316103032e-05,
+      "loss": 1.6798,
+      "step": 5058
+    },
+    {
+      "epoch": 1.3295234341642854,
+      "grad_norm": 0.6392707228660583,
+      "learning_rate": 5.569476082004557e-05,
+      "loss": 1.6794,
+      "step": 5060
+    },
+    {
+      "epoch": 1.3300489374979472,
+      "grad_norm": 0.6719961762428284,
+      "learning_rate": 5.56772384790608e-05,
+      "loss": 1.6773,
+      "step": 5062
+    },
+    {
+      "epoch": 1.330574440831609,
+      "grad_norm": 0.6991645693778992,
+      "learning_rate": 5.565971613807604e-05,
+      "loss": 1.7275,
+      "step": 5064
+    },
+    {
+      "epoch": 1.3310999441652709,
+      "grad_norm": 0.54388827085495,
+      "learning_rate": 5.564219379709129e-05,
+      "loss": 1.6641,
+      "step": 5066
+    },
+    {
+      "epoch": 1.3316254474989326,
+      "grad_norm": 0.8866559267044067,
+      "learning_rate": 5.562467145610654e-05,
+      "loss": 1.666,
+      "step": 5068
+    },
+    {
+      "epoch": 1.3321509508325944,
+      "grad_norm": 0.5608734488487244,
+      "learning_rate": 5.560714911512178e-05,
+      "loss": 1.6793,
+      "step": 5070
+    },
+    {
+      "epoch": 1.3326764541662561,
+      "grad_norm": 0.5888034701347351,
+      "learning_rate": 5.5589626774137026e-05,
+      "loss": 1.6902,
+      "step": 5072
+    },
+    {
+      "epoch": 1.3332019574999179,
+      "grad_norm": 0.5383687019348145,
+      "learning_rate": 5.557210443315227e-05,
+      "loss": 1.6738,
+      "step": 5074
+    },
+    {
+      "epoch": 1.3337274608335796,
+      "grad_norm": 0.681666374206543,
+      "learning_rate": 5.5554582092167514e-05,
+      "loss": 1.6709,
+      "step": 5076
+    },
+    {
+      "epoch": 1.3342529641672414,
+      "grad_norm": 0.6175423264503479,
+      "learning_rate": 5.553705975118276e-05,
+      "loss": 1.6617,
+      "step": 5078
+    },
+    {
+      "epoch": 1.3347784675009033,
+      "grad_norm": 0.5964341163635254,
+      "learning_rate": 5.551953741019801e-05,
+      "loss": 1.6484,
+      "step": 5080
+    },
+    {
+      "epoch": 1.3353039708345649,
+      "grad_norm": 0.5749648809432983,
+      "learning_rate": 5.550201506921325e-05,
+      "loss": 1.6819,
+      "step": 5082
+    },
+    {
+      "epoch": 1.3358294741682268,
+      "grad_norm": 0.6597902774810791,
+      "learning_rate": 5.54844927282285e-05,
+      "loss": 1.7234,
+      "step": 5084
+    },
+    {
+      "epoch": 1.3363549775018886,
+      "grad_norm": 0.6115815043449402,
+      "learning_rate": 5.546697038724373e-05,
+      "loss": 1.683,
+      "step": 5086
+    },
+    {
+      "epoch": 1.3368804808355503,
+      "grad_norm": 0.63287353515625,
+      "learning_rate": 5.544944804625898e-05,
+      "loss": 1.6759,
+      "step": 5088
+    },
+    {
+      "epoch": 1.337405984169212,
+      "grad_norm": 0.6917816400527954,
+      "learning_rate": 5.543192570527423e-05,
+      "loss": 1.6938,
+      "step": 5090
+    },
+    {
+      "epoch": 1.3379314875028738,
+      "grad_norm": 0.5193148255348206,
+      "learning_rate": 5.541440336428947e-05,
+      "loss": 1.6918,
+      "step": 5092
+    },
+    {
+      "epoch": 1.3384569908365356,
+      "grad_norm": 0.5470016002655029,
+      "learning_rate": 5.5396881023304715e-05,
+      "loss": 1.6733,
+      "step": 5094
+    },
+    {
+      "epoch": 1.3389824941701973,
+      "grad_norm": 0.6966432332992554,
+      "learning_rate": 5.537935868231996e-05,
+      "loss": 1.6883,
+      "step": 5096
+    },
+    {
+      "epoch": 1.3395079975038593,
+      "grad_norm": 0.6289670467376709,
+      "learning_rate": 5.5361836341335203e-05,
+      "loss": 1.6666,
+      "step": 5098
+    },
+    {
+      "epoch": 1.3400335008375208,
+      "grad_norm": 0.5251250863075256,
+      "learning_rate": 5.534431400035045e-05,
+      "loss": 1.6477,
+      "step": 5100
+    },
+    {
+      "epoch": 1.3405590041711828,
+      "grad_norm": 0.663750171661377,
+      "learning_rate": 5.53267916593657e-05,
+      "loss": 1.6713,
+      "step": 5102
+    },
+    {
+      "epoch": 1.3410845075048445,
+      "grad_norm": 0.5892371535301208,
+      "learning_rate": 5.530926931838094e-05,
+      "loss": 1.6935,
+      "step": 5104
+    },
+    {
+      "epoch": 1.3416100108385063,
+      "grad_norm": 0.7290005683898926,
+      "learning_rate": 5.529174697739619e-05,
+      "loss": 1.6685,
+      "step": 5106
+    },
+    {
+      "epoch": 1.342135514172168,
+      "grad_norm": 0.6136685013771057,
+      "learning_rate": 5.5274224636411435e-05,
+      "loss": 1.6441,
+      "step": 5108
+    },
+    {
+      "epoch": 1.3426610175058298,
+      "grad_norm": 0.648530900478363,
+      "learning_rate": 5.5256702295426675e-05,
+      "loss": 1.6874,
+      "step": 5110
+    },
+    {
+      "epoch": 1.3431865208394915,
+      "grad_norm": 0.556920051574707,
+      "learning_rate": 5.523917995444191e-05,
+      "loss": 1.6945,
+      "step": 5112
+    },
+    {
+      "epoch": 1.3437120241731533,
+      "grad_norm": 0.6408731937408447,
+      "learning_rate": 5.522165761345716e-05,
+      "loss": 1.6719,
+      "step": 5114
+    },
+    {
+      "epoch": 1.3442375275068152,
+      "grad_norm": 0.6126033663749695,
+      "learning_rate": 5.5204135272472404e-05,
+      "loss": 1.6851,
+      "step": 5116
+    },
+    {
+      "epoch": 1.3447630308404768,
+      "grad_norm": 0.7299725413322449,
+      "learning_rate": 5.5186612931487645e-05,
+      "loss": 1.6442,
+      "step": 5118
+    },
+    {
+      "epoch": 1.3452885341741387,
+      "grad_norm": 0.5226702690124512,
+      "learning_rate": 5.516909059050289e-05,
+      "loss": 1.6779,
+      "step": 5120
+    },
+    {
+      "epoch": 1.3458140375078005,
+      "grad_norm": 0.5271426439285278,
+      "learning_rate": 5.515156824951814e-05,
+      "loss": 1.6756,
+      "step": 5122
+    },
+    {
+      "epoch": 1.3463395408414622,
+      "grad_norm": 0.5417149066925049,
+      "learning_rate": 5.513404590853338e-05,
+      "loss": 1.7041,
+      "step": 5124
+    },
+    {
+      "epoch": 1.346865044175124,
+      "grad_norm": 0.639298677444458,
+      "learning_rate": 5.511652356754863e-05,
+      "loss": 1.7372,
+      "step": 5126
+    },
+    {
+      "epoch": 1.3473905475087857,
+      "grad_norm": 0.7180479764938354,
+      "learning_rate": 5.5099001226563876e-05,
+      "loss": 1.7078,
+      "step": 5128
+    },
+    {
+      "epoch": 1.3479160508424475,
+      "grad_norm": 0.6249936819076538,
+      "learning_rate": 5.508147888557912e-05,
+      "loss": 1.672,
+      "step": 5130
+    },
+    {
+      "epoch": 1.3484415541761092,
+      "grad_norm": 0.6733811497688293,
+      "learning_rate": 5.5063956544594365e-05,
+      "loss": 1.6648,
+      "step": 5132
+    },
+    {
+      "epoch": 1.3489670575097712,
+      "grad_norm": 0.628979504108429,
+      "learning_rate": 5.504643420360961e-05,
+      "loss": 1.6724,
+      "step": 5134
+    },
+    {
+      "epoch": 1.3494925608434327,
+      "grad_norm": 0.6797083020210266,
+      "learning_rate": 5.502891186262485e-05,
+      "loss": 1.6681,
+      "step": 5136
+    },
+    {
+      "epoch": 1.3500180641770947,
+      "grad_norm": 0.5499573945999146,
+      "learning_rate": 5.501138952164009e-05,
+      "loss": 1.6877,
+      "step": 5138
+    },
+    {
+      "epoch": 1.3505435675107564,
+      "grad_norm": 0.6924422383308411,
+      "learning_rate": 5.4993867180655335e-05,
+      "loss": 1.6836,
+      "step": 5140
+    },
+    {
+      "epoch": 1.3510690708444182,
+      "grad_norm": 0.5735469460487366,
+      "learning_rate": 5.497634483967058e-05,
+      "loss": 1.6953,
+      "step": 5142
+    },
+    {
+      "epoch": 1.35159457417808,
+      "grad_norm": 0.6161128282546997,
+      "learning_rate": 5.495882249868582e-05,
+      "loss": 1.6663,
+      "step": 5144
+    },
+    {
+      "epoch": 1.3521200775117417,
+      "grad_norm": 0.5718648433685303,
+      "learning_rate": 5.494130015770107e-05,
+      "loss": 1.6785,
+      "step": 5146
+    },
+    {
+      "epoch": 1.3526455808454034,
+      "grad_norm": 0.6534953713417053,
+      "learning_rate": 5.492377781671632e-05,
+      "loss": 1.6859,
+      "step": 5148
+    },
+    {
+      "epoch": 1.3531710841790652,
+      "grad_norm": 0.7436297535896301,
+      "learning_rate": 5.490625547573156e-05,
+      "loss": 1.6833,
+      "step": 5150
+    },
+    {
+      "epoch": 1.3536965875127271,
+      "grad_norm": 0.5223855972290039,
+      "learning_rate": 5.4888733134746806e-05,
+      "loss": 1.7009,
+      "step": 5152
+    },
+    {
+      "epoch": 1.3542220908463887,
+      "grad_norm": 0.6421153545379639,
+      "learning_rate": 5.4871210793762054e-05,
+      "loss": 1.6884,
+      "step": 5154
+    },
+    {
+      "epoch": 1.3547475941800506,
+      "grad_norm": 0.5627347826957703,
+      "learning_rate": 5.4853688452777295e-05,
+      "loss": 1.6761,
+      "step": 5156
+    },
+    {
+      "epoch": 1.3552730975137124,
+      "grad_norm": 0.5863370299339294,
+      "learning_rate": 5.483616611179254e-05,
+      "loss": 1.7144,
+      "step": 5158
+    },
+    {
+      "epoch": 1.3557986008473741,
+      "grad_norm": 0.6079491376876831,
+      "learning_rate": 5.481864377080779e-05,
+      "loss": 1.6679,
+      "step": 5160
+    },
+    {
+      "epoch": 1.3563241041810359,
+      "grad_norm": 0.603889524936676,
+      "learning_rate": 5.480112142982303e-05,
+      "loss": 1.6858,
+      "step": 5162
+    },
+    {
+      "epoch": 1.3568496075146976,
+      "grad_norm": 0.7292947173118591,
+      "learning_rate": 5.4783599088838265e-05,
+      "loss": 1.6597,
+      "step": 5164
+    },
+    {
+      "epoch": 1.3573751108483594,
+      "grad_norm": 0.6873249411582947,
+      "learning_rate": 5.476607674785351e-05,
+      "loss": 1.6614,
+      "step": 5166
+    },
+    {
+      "epoch": 1.3579006141820211,
+      "grad_norm": 0.6474335789680481,
+      "learning_rate": 5.474855440686876e-05,
+      "loss": 1.6651,
+      "step": 5168
+    },
+    {
+      "epoch": 1.358426117515683,
+      "grad_norm": 0.6388329863548279,
+      "learning_rate": 5.4731032065884e-05,
+      "loss": 1.7278,
+      "step": 5170
+    },
+    {
+      "epoch": 1.3589516208493448,
+      "grad_norm": 0.6011612415313721,
+      "learning_rate": 5.471350972489925e-05,
+      "loss": 1.6685,
+      "step": 5172
+    },
+    {
+      "epoch": 1.3594771241830066,
+      "grad_norm": 0.5934230089187622,
+      "learning_rate": 5.4695987383914496e-05,
+      "loss": 1.6956,
+      "step": 5174
+    },
+    {
+      "epoch": 1.3600026275166683,
+      "grad_norm": 0.5996755361557007,
+      "learning_rate": 5.4678465042929736e-05,
+      "loss": 1.6717,
+      "step": 5176
+    },
+    {
+      "epoch": 1.36052813085033,
+      "grad_norm": 0.5604281425476074,
+      "learning_rate": 5.4660942701944984e-05,
+      "loss": 1.6535,
+      "step": 5178
+    },
+    {
+      "epoch": 1.3610536341839918,
+      "grad_norm": 0.7021388411521912,
+      "learning_rate": 5.464342036096023e-05,
+      "loss": 1.6387,
+      "step": 5180
+    },
+    {
+      "epoch": 1.3615791375176536,
+      "grad_norm": 0.599388599395752,
+      "learning_rate": 5.462589801997547e-05,
+      "loss": 1.7095,
+      "step": 5182
+    },
+    {
+      "epoch": 1.3621046408513153,
+      "grad_norm": 0.6312151551246643,
+      "learning_rate": 5.460837567899072e-05,
+      "loss": 1.7024,
+      "step": 5184
+    },
+    {
+      "epoch": 1.362630144184977,
+      "grad_norm": 0.6004204154014587,
+      "learning_rate": 5.459085333800597e-05,
+      "loss": 1.6731,
+      "step": 5186
+    },
+    {
+      "epoch": 1.363155647518639,
+      "grad_norm": 0.5637236833572388,
+      "learning_rate": 5.45733309970212e-05,
+      "loss": 1.6846,
+      "step": 5188
+    },
+    {
+      "epoch": 1.3636811508523008,
+      "grad_norm": 0.5435346961021423,
+      "learning_rate": 5.455580865603644e-05,
+      "loss": 1.697,
+      "step": 5190
+    },
+    {
+      "epoch": 1.3642066541859625,
+      "grad_norm": 0.7027901411056519,
+      "learning_rate": 5.453828631505169e-05,
+      "loss": 1.6658,
+      "step": 5192
+    },
+    {
+      "epoch": 1.3647321575196243,
+      "grad_norm": 0.5427135229110718,
+      "learning_rate": 5.452076397406694e-05,
+      "loss": 1.6802,
+      "step": 5194
+    },
+    {
+      "epoch": 1.365257660853286,
+      "grad_norm": 0.9551213979721069,
+      "learning_rate": 5.450324163308218e-05,
+      "loss": 1.6675,
+      "step": 5196
+    },
+    {
+      "epoch": 1.3657831641869478,
+      "grad_norm": 0.6727270483970642,
+      "learning_rate": 5.4485719292097426e-05,
+      "loss": 1.6909,
+      "step": 5198
+    },
+    {
+      "epoch": 1.3663086675206095,
+      "grad_norm": 0.5825324058532715,
+      "learning_rate": 5.446819695111267e-05,
+      "loss": 1.6678,
+      "step": 5200
+    },
+    {
+      "epoch": 1.3663086675206095,
+      "eval_loss": 1.6925063133239746,
+      "eval_runtime": 487.2699,
+      "eval_samples_per_second": 249.942,
+      "eval_steps_per_second": 31.243,
+      "step": 5200
+    },
+    {
+      "epoch": 1.3668341708542713,
+      "grad_norm": 0.5803776979446411,
+      "learning_rate": 5.4450674610127914e-05,
+      "loss": 1.687,
+      "step": 5202
+    },
+    {
+      "epoch": 1.367359674187933,
+      "grad_norm": 0.5400435328483582,
+      "learning_rate": 5.443315226914316e-05,
+      "loss": 1.7248,
+      "step": 5204
+    },
+    {
+      "epoch": 1.367885177521595,
+      "grad_norm": 0.5866122245788574,
+      "learning_rate": 5.441562992815841e-05,
+      "loss": 1.6612,
+      "step": 5206
+    },
+    {
+      "epoch": 1.3684106808552567,
+      "grad_norm": 0.6416072249412537,
+      "learning_rate": 5.439810758717365e-05,
+      "loss": 1.6948,
+      "step": 5208
+    },
+    {
+      "epoch": 1.3689361841889185,
+      "grad_norm": 0.5607174634933472,
+      "learning_rate": 5.43805852461889e-05,
+      "loss": 1.7079,
+      "step": 5210
+    },
+    {
+      "epoch": 1.3694616875225802,
+      "grad_norm": 0.5060856938362122,
+      "learning_rate": 5.4363062905204145e-05,
+      "loss": 1.6811,
+      "step": 5212
+    },
+    {
+      "epoch": 1.369987190856242,
+      "grad_norm": 0.5724950432777405,
+      "learning_rate": 5.434554056421938e-05,
+      "loss": 1.6251,
+      "step": 5214
+    },
+    {
+      "epoch": 1.3705126941899037,
+      "grad_norm": 0.595393717288971,
+      "learning_rate": 5.432801822323462e-05,
+      "loss": 1.6867,
+      "step": 5216
+    },
+    {
+      "epoch": 1.3710381975235655,
+      "grad_norm": 0.5957738161087036,
+      "learning_rate": 5.431049588224987e-05,
+      "loss": 1.6971,
+      "step": 5218
+    },
+    {
+      "epoch": 1.3715637008572272,
+      "grad_norm": 0.6889095306396484,
+      "learning_rate": 5.4292973541265115e-05,
+      "loss": 1.6999,
+      "step": 5220
+    },
+    {
+      "epoch": 1.372089204190889,
+      "grad_norm": 0.6807567477226257,
+      "learning_rate": 5.4275451200280356e-05,
+      "loss": 1.6674,
+      "step": 5222
+    },
+    {
+      "epoch": 1.372614707524551,
+      "grad_norm": 0.6046966314315796,
+      "learning_rate": 5.4257928859295603e-05,
+      "loss": 1.6622,
+      "step": 5224
+    },
+    {
+      "epoch": 1.3731402108582127,
+      "grad_norm": 0.6781222224235535,
+      "learning_rate": 5.424040651831085e-05,
+      "loss": 1.6974,
+      "step": 5226
+    },
+    {
+      "epoch": 1.3736657141918744,
+      "grad_norm": 0.5710211992263794,
+      "learning_rate": 5.422288417732609e-05,
+      "loss": 1.6769,
+      "step": 5228
+    },
+    {
+      "epoch": 1.3741912175255362,
+      "grad_norm": 0.6032019257545471,
+      "learning_rate": 5.420536183634134e-05,
+      "loss": 1.6615,
+      "step": 5230
+    },
+    {
+      "epoch": 1.374716720859198,
+      "grad_norm": 0.6502026915550232,
+      "learning_rate": 5.418783949535659e-05,
+      "loss": 1.6965,
+      "step": 5232
+    },
+    {
+      "epoch": 1.3752422241928597,
+      "grad_norm": 0.7333769798278809,
+      "learning_rate": 5.417031715437183e-05,
+      "loss": 1.689,
+      "step": 5234
+    },
+    {
+      "epoch": 1.3757677275265214,
+      "grad_norm": 0.5849090218544006,
+      "learning_rate": 5.4152794813387075e-05,
+      "loss": 1.6696,
+      "step": 5236
+    },
+    {
+      "epoch": 1.3762932308601834,
+      "grad_norm": 0.7076939940452576,
+      "learning_rate": 5.413527247240232e-05,
+      "loss": 1.6674,
+      "step": 5238
+    },
+    {
+      "epoch": 1.376818734193845,
+      "grad_norm": 0.686165988445282,
+      "learning_rate": 5.411775013141756e-05,
+      "loss": 1.6729,
+      "step": 5240
+    },
+    {
+      "epoch": 1.377344237527507,
+      "grad_norm": 0.7064648270606995,
+      "learning_rate": 5.41002277904328e-05,
+      "loss": 1.6532,
+      "step": 5242
+    },
+    {
+      "epoch": 1.3778697408611686,
+      "grad_norm": 0.707788348197937,
+      "learning_rate": 5.4082705449448045e-05,
+      "loss": 1.635,
+      "step": 5244
+    },
+    {
+      "epoch": 1.3783952441948304,
+      "grad_norm": 0.711150586605072,
+      "learning_rate": 5.406518310846329e-05,
+      "loss": 1.6671,
+      "step": 5246
+    },
+    {
+      "epoch": 1.3789207475284921,
+      "grad_norm": 0.627653181552887,
+      "learning_rate": 5.4047660767478534e-05,
+      "loss": 1.6887,
+      "step": 5248
+    },
+    {
+      "epoch": 1.3794462508621539,
+      "grad_norm": 0.5618575215339661,
+      "learning_rate": 5.403013842649378e-05,
+      "loss": 1.6773,
+      "step": 5250
+    },
+    {
+      "epoch": 1.3799717541958156,
+      "grad_norm": 0.6045578718185425,
+      "learning_rate": 5.401261608550903e-05,
+      "loss": 1.7151,
+      "step": 5252
+    },
+    {
+      "epoch": 1.3804972575294774,
+      "grad_norm": 0.6123912930488586,
+      "learning_rate": 5.399509374452427e-05,
+      "loss": 1.6428,
+      "step": 5254
+    },
+    {
+      "epoch": 1.3810227608631394,
+      "grad_norm": 0.7001442909240723,
+      "learning_rate": 5.397757140353952e-05,
+      "loss": 1.7063,
+      "step": 5256
+    },
+    {
+      "epoch": 1.3815482641968009,
+      "grad_norm": 0.5598956346511841,
+      "learning_rate": 5.3960049062554765e-05,
+      "loss": 1.671,
+      "step": 5258
+    },
+    {
+      "epoch": 1.3820737675304628,
+      "grad_norm": 0.5916640758514404,
+      "learning_rate": 5.3942526721570005e-05,
+      "loss": 1.6685,
+      "step": 5260
+    },
+    {
+      "epoch": 1.3825992708641246,
+      "grad_norm": 0.5688573718070984,
+      "learning_rate": 5.392500438058525e-05,
+      "loss": 1.6357,
+      "step": 5262
+    },
+    {
+      "epoch": 1.3831247741977863,
+      "grad_norm": 0.6486282348632812,
+      "learning_rate": 5.39074820396005e-05,
+      "loss": 1.6803,
+      "step": 5264
+    },
+    {
+      "epoch": 1.383650277531448,
+      "grad_norm": 0.6464768052101135,
+      "learning_rate": 5.3889959698615735e-05,
+      "loss": 1.6781,
+      "step": 5266
+    },
+    {
+      "epoch": 1.3841757808651098,
+      "grad_norm": 0.5832222104072571,
+      "learning_rate": 5.3872437357630975e-05,
+      "loss": 1.7009,
+      "step": 5268
+    },
+    {
+      "epoch": 1.3847012841987716,
+      "grad_norm": 0.6113573908805847,
+      "learning_rate": 5.385491501664622e-05,
+      "loss": 1.6855,
+      "step": 5270
+    },
+    {
+      "epoch": 1.3852267875324333,
+      "grad_norm": 0.6312457323074341,
+      "learning_rate": 5.383739267566147e-05,
+      "loss": 1.7076,
+      "step": 5272
+    },
+    {
+      "epoch": 1.3857522908660953,
+      "grad_norm": 0.6757798790931702,
+      "learning_rate": 5.381987033467671e-05,
+      "loss": 1.6796,
+      "step": 5274
+    },
+    {
+      "epoch": 1.3862777941997568,
+      "grad_norm": 0.7711691856384277,
+      "learning_rate": 5.380234799369196e-05,
+      "loss": 1.656,
+      "step": 5276
+    },
+    {
+      "epoch": 1.3868032975334188,
+      "grad_norm": 0.6918832659721375,
+      "learning_rate": 5.3784825652707206e-05,
+      "loss": 1.6723,
+      "step": 5278
+    },
+    {
+      "epoch": 1.3873288008670805,
+      "grad_norm": 0.5812153220176697,
+      "learning_rate": 5.376730331172245e-05,
+      "loss": 1.6769,
+      "step": 5280
+    },
+    {
+      "epoch": 1.3878543042007423,
+      "grad_norm": 0.642285168170929,
+      "learning_rate": 5.3749780970737695e-05,
+      "loss": 1.6962,
+      "step": 5282
+    },
+    {
+      "epoch": 1.388379807534404,
+      "grad_norm": 0.5948686003684998,
+      "learning_rate": 5.373225862975294e-05,
+      "loss": 1.6874,
+      "step": 5284
+    },
+    {
+      "epoch": 1.3889053108680658,
+      "grad_norm": 0.8630838394165039,
+      "learning_rate": 5.371473628876818e-05,
+      "loss": 1.6602,
+      "step": 5286
+    },
+    {
+      "epoch": 1.3894308142017275,
+      "grad_norm": 0.624987006187439,
+      "learning_rate": 5.369721394778343e-05,
+      "loss": 1.6674,
+      "step": 5288
+    },
+    {
+      "epoch": 1.3899563175353893,
+      "grad_norm": 0.5534707307815552,
+      "learning_rate": 5.3679691606798665e-05,
+      "loss": 1.7101,
+      "step": 5290
+    },
+    {
+      "epoch": 1.3904818208690513,
+      "grad_norm": 0.547791063785553,
+      "learning_rate": 5.366216926581391e-05,
+      "loss": 1.7103,
+      "step": 5292
+    },
+    {
+      "epoch": 1.3910073242027128,
+      "grad_norm": 0.7196782827377319,
+      "learning_rate": 5.364464692482915e-05,
+      "loss": 1.7016,
+      "step": 5294
+    },
+    {
+      "epoch": 1.3915328275363748,
+      "grad_norm": 0.5943964123725891,
+      "learning_rate": 5.36271245838444e-05,
+      "loss": 1.6937,
+      "step": 5296
+    },
+    {
+      "epoch": 1.3920583308700365,
+      "grad_norm": 0.5824291706085205,
+      "learning_rate": 5.360960224285965e-05,
+      "loss": 1.6606,
+      "step": 5298
+    },
+    {
+      "epoch": 1.3925838342036982,
+      "grad_norm": 0.6096850633621216,
+      "learning_rate": 5.359207990187489e-05,
+      "loss": 1.6732,
+      "step": 5300
+    },
+    {
+      "epoch": 1.39310933753736,
+      "grad_norm": 0.5567105412483215,
+      "learning_rate": 5.3574557560890136e-05,
+      "loss": 1.6917,
+      "step": 5302
+    },
+    {
+      "epoch": 1.3936348408710217,
+      "grad_norm": 0.808890700340271,
+      "learning_rate": 5.3557035219905384e-05,
+      "loss": 1.6916,
+      "step": 5304
+    },
+    {
+      "epoch": 1.3941603442046835,
+      "grad_norm": 0.602245032787323,
+      "learning_rate": 5.3539512878920625e-05,
+      "loss": 1.6565,
+      "step": 5306
+    },
+    {
+      "epoch": 1.3946858475383452,
+      "grad_norm": 0.5515073537826538,
+      "learning_rate": 5.352199053793587e-05,
+      "loss": 1.6927,
+      "step": 5308
+    },
+    {
+      "epoch": 1.3952113508720072,
+      "grad_norm": 0.6020869016647339,
+      "learning_rate": 5.350446819695112e-05,
+      "loss": 1.6813,
+      "step": 5310
+    },
+    {
+      "epoch": 1.3957368542056687,
+      "grad_norm": 0.6068871021270752,
+      "learning_rate": 5.348694585596636e-05,
+      "loss": 1.699,
+      "step": 5312
+    },
+    {
+      "epoch": 1.3962623575393307,
+      "grad_norm": 0.8839384913444519,
+      "learning_rate": 5.346942351498161e-05,
+      "loss": 1.7296,
+      "step": 5314
+    },
+    {
+      "epoch": 1.3967878608729924,
+      "grad_norm": 0.6876543164253235,
+      "learning_rate": 5.345190117399684e-05,
+      "loss": 1.6827,
+      "step": 5316
+    },
+    {
+      "epoch": 1.3973133642066542,
+      "grad_norm": 0.9712225198745728,
+      "learning_rate": 5.343437883301209e-05,
+      "loss": 1.6821,
+      "step": 5318
+    },
+    {
+      "epoch": 1.397838867540316,
+      "grad_norm": 0.644990861415863,
+      "learning_rate": 5.341685649202733e-05,
+      "loss": 1.671,
+      "step": 5320
+    },
+    {
+      "epoch": 1.3983643708739777,
+      "grad_norm": 0.7299576997756958,
+      "learning_rate": 5.339933415104258e-05,
+      "loss": 1.7054,
+      "step": 5322
+    },
+    {
+      "epoch": 1.3988898742076394,
+      "grad_norm": 0.5608115196228027,
+      "learning_rate": 5.3381811810057826e-05,
+      "loss": 1.6713,
+      "step": 5324
+    },
+    {
+      "epoch": 1.3994153775413012,
+      "grad_norm": 0.6044563055038452,
+      "learning_rate": 5.3364289469073067e-05,
+      "loss": 1.6838,
+      "step": 5326
+    },
+    {
+      "epoch": 1.3999408808749632,
+      "grad_norm": 0.5304017663002014,
+      "learning_rate": 5.3346767128088314e-05,
+      "loss": 1.6351,
+      "step": 5328
+    },
+    {
+      "epoch": 1.400466384208625,
+      "grad_norm": 0.5934461355209351,
+      "learning_rate": 5.332924478710356e-05,
+      "loss": 1.6485,
+      "step": 5330
+    },
+    {
+      "epoch": 1.4009918875422867,
+      "grad_norm": 0.5502288937568665,
+      "learning_rate": 5.33117224461188e-05,
+      "loss": 1.683,
+      "step": 5332
+    },
+    {
+      "epoch": 1.4015173908759484,
+      "grad_norm": 0.6545527577400208,
+      "learning_rate": 5.329420010513405e-05,
+      "loss": 1.6823,
+      "step": 5334
+    },
+    {
+      "epoch": 1.4020428942096101,
+      "grad_norm": 0.5551027655601501,
+      "learning_rate": 5.32766777641493e-05,
+      "loss": 1.6867,
+      "step": 5336
+    },
+    {
+      "epoch": 1.402568397543272,
+      "grad_norm": 0.6732555627822876,
+      "learning_rate": 5.325915542316454e-05,
+      "loss": 1.7031,
+      "step": 5338
+    },
+    {
+      "epoch": 1.4030939008769336,
+      "grad_norm": 0.5857707858085632,
+      "learning_rate": 5.3241633082179786e-05,
+      "loss": 1.6607,
+      "step": 5340
+    },
+    {
+      "epoch": 1.4036194042105954,
+      "grad_norm": 0.6332495212554932,
+      "learning_rate": 5.322411074119502e-05,
+      "loss": 1.6979,
+      "step": 5342
+    },
+    {
+      "epoch": 1.4041449075442571,
+      "grad_norm": 0.5486958026885986,
+      "learning_rate": 5.320658840021027e-05,
+      "loss": 1.6582,
+      "step": 5344
+    },
+    {
+      "epoch": 1.404670410877919,
+      "grad_norm": 0.5463687181472778,
+      "learning_rate": 5.3189066059225515e-05,
+      "loss": 1.7192,
+      "step": 5346
+    },
+    {
+      "epoch": 1.4051959142115809,
+      "grad_norm": 0.4964401125907898,
+      "learning_rate": 5.3171543718240756e-05,
+      "loss": 1.6486,
+      "step": 5348
+    },
+    {
+      "epoch": 1.4057214175452426,
+      "grad_norm": 0.5375365018844604,
+      "learning_rate": 5.3154021377256003e-05,
+      "loss": 1.714,
+      "step": 5350
+    },
+    {
+      "epoch": 1.4062469208789043,
+      "grad_norm": 0.5706174373626709,
+      "learning_rate": 5.313649903627125e-05,
+      "loss": 1.702,
+      "step": 5352
+    },
+    {
+      "epoch": 1.406772424212566,
+      "grad_norm": 0.6177610158920288,
+      "learning_rate": 5.311897669528649e-05,
+      "loss": 1.6439,
+      "step": 5354
+    },
+    {
+      "epoch": 1.4072979275462278,
+      "grad_norm": 0.5553119778633118,
+      "learning_rate": 5.310145435430174e-05,
+      "loss": 1.69,
+      "step": 5356
+    },
+    {
+      "epoch": 1.4078234308798896,
+      "grad_norm": 0.6558260917663574,
+      "learning_rate": 5.308393201331699e-05,
+      "loss": 1.7068,
+      "step": 5358
+    },
+    {
+      "epoch": 1.4083489342135513,
+      "grad_norm": 0.7940452098846436,
+      "learning_rate": 5.306640967233223e-05,
+      "loss": 1.6912,
+      "step": 5360
+    },
+    {
+      "epoch": 1.408874437547213,
+      "grad_norm": 0.577286958694458,
+      "learning_rate": 5.3048887331347475e-05,
+      "loss": 1.6755,
+      "step": 5362
+    },
+    {
+      "epoch": 1.409399940880875,
+      "grad_norm": 0.6552362442016602,
+      "learning_rate": 5.303136499036272e-05,
+      "loss": 1.6952,
+      "step": 5364
+    },
+    {
+      "epoch": 1.4099254442145368,
+      "grad_norm": 0.6791645884513855,
+      "learning_rate": 5.3013842649377964e-05,
+      "loss": 1.6979,
+      "step": 5366
+    },
+    {
+      "epoch": 1.4104509475481986,
+      "grad_norm": 0.5956308841705322,
+      "learning_rate": 5.29963203083932e-05,
+      "loss": 1.6742,
+      "step": 5368
+    },
+    {
+      "epoch": 1.4109764508818603,
+      "grad_norm": 0.6051899790763855,
+      "learning_rate": 5.2978797967408445e-05,
+      "loss": 1.6704,
+      "step": 5370
+    },
+    {
+      "epoch": 1.411501954215522,
+      "grad_norm": 0.6038895845413208,
+      "learning_rate": 5.296127562642369e-05,
+      "loss": 1.6894,
+      "step": 5372
+    },
+    {
+      "epoch": 1.4120274575491838,
+      "grad_norm": 0.5314784049987793,
+      "learning_rate": 5.2943753285438934e-05,
+      "loss": 1.6656,
+      "step": 5374
+    },
+    {
+      "epoch": 1.4125529608828455,
+      "grad_norm": 0.5296756029129028,
+      "learning_rate": 5.292623094445418e-05,
+      "loss": 1.6957,
+      "step": 5376
+    },
+    {
+      "epoch": 1.4130784642165073,
+      "grad_norm": 0.5364149808883667,
+      "learning_rate": 5.290870860346943e-05,
+      "loss": 1.6585,
+      "step": 5378
+    },
+    {
+      "epoch": 1.413603967550169,
+      "grad_norm": 0.5761370062828064,
+      "learning_rate": 5.289118626248467e-05,
+      "loss": 1.6692,
+      "step": 5380
+    },
+    {
+      "epoch": 1.414129470883831,
+      "grad_norm": 0.6723489165306091,
+      "learning_rate": 5.287366392149992e-05,
+      "loss": 1.6826,
+      "step": 5382
+    },
+    {
+      "epoch": 1.4146549742174928,
+      "grad_norm": 0.5427255034446716,
+      "learning_rate": 5.2856141580515165e-05,
+      "loss": 1.6612,
+      "step": 5384
+    },
+    {
+      "epoch": 1.4151804775511545,
+      "grad_norm": 0.5773686170578003,
+      "learning_rate": 5.2838619239530405e-05,
+      "loss": 1.6751,
+      "step": 5386
+    },
+    {
+      "epoch": 1.4157059808848163,
+      "grad_norm": 0.9336940050125122,
+      "learning_rate": 5.282109689854565e-05,
+      "loss": 1.6976,
+      "step": 5388
+    },
+    {
+      "epoch": 1.416231484218478,
+      "grad_norm": 0.5543003082275391,
+      "learning_rate": 5.28035745575609e-05,
+      "loss": 1.6991,
+      "step": 5390
+    },
+    {
+      "epoch": 1.4167569875521397,
+      "grad_norm": 0.5574566125869751,
+      "learning_rate": 5.2786052216576134e-05,
+      "loss": 1.6575,
+      "step": 5392
+    },
+    {
+      "epoch": 1.4172824908858015,
+      "grad_norm": 0.6602552533149719,
+      "learning_rate": 5.2768529875591375e-05,
+      "loss": 1.6748,
+      "step": 5394
+    },
+    {
+      "epoch": 1.4178079942194635,
+      "grad_norm": 0.644187331199646,
+      "learning_rate": 5.275100753460662e-05,
+      "loss": 1.6565,
+      "step": 5396
+    },
+    {
+      "epoch": 1.418333497553125,
+      "grad_norm": 0.5814189314842224,
+      "learning_rate": 5.273348519362187e-05,
+      "loss": 1.6965,
+      "step": 5398
+    },
+    {
+      "epoch": 1.418859000886787,
+      "grad_norm": 0.5712095499038696,
+      "learning_rate": 5.271596285263711e-05,
+      "loss": 1.6593,
+      "step": 5400
+    },
+    {
+      "epoch": 1.4193845042204487,
+      "grad_norm": 0.6919686198234558,
+      "learning_rate": 5.269844051165236e-05,
+      "loss": 1.692,
+      "step": 5402
+    },
+    {
+      "epoch": 1.4199100075541105,
+      "grad_norm": 0.6162757277488708,
+      "learning_rate": 5.2680918170667606e-05,
+      "loss": 1.6726,
+      "step": 5404
+    },
+    {
+      "epoch": 1.4204355108877722,
+      "grad_norm": 0.5998090505599976,
+      "learning_rate": 5.266339582968285e-05,
+      "loss": 1.7206,
+      "step": 5406
+    },
+    {
+      "epoch": 1.420961014221434,
+      "grad_norm": 0.6242619156837463,
+      "learning_rate": 5.2645873488698095e-05,
+      "loss": 1.6647,
+      "step": 5408
+    },
+    {
+      "epoch": 1.4214865175550957,
+      "grad_norm": 0.5882295966148376,
+      "learning_rate": 5.262835114771334e-05,
+      "loss": 1.6562,
+      "step": 5410
+    },
+    {
+      "epoch": 1.4220120208887574,
+      "grad_norm": 0.5960384607315063,
+      "learning_rate": 5.261082880672858e-05,
+      "loss": 1.6877,
+      "step": 5412
+    },
+    {
+      "epoch": 1.4225375242224194,
+      "grad_norm": 0.6366286873817444,
+      "learning_rate": 5.259330646574383e-05,
+      "loss": 1.7009,
+      "step": 5414
+    },
+    {
+      "epoch": 1.423063027556081,
+      "grad_norm": 0.581294059753418,
+      "learning_rate": 5.257578412475908e-05,
+      "loss": 1.6918,
+      "step": 5416
+    },
+    {
+      "epoch": 1.423588530889743,
+      "grad_norm": 0.6820783615112305,
+      "learning_rate": 5.255826178377431e-05,
+      "loss": 1.6602,
+      "step": 5418
+    },
+    {
+      "epoch": 1.4241140342234047,
+      "grad_norm": 0.7352914810180664,
+      "learning_rate": 5.254073944278955e-05,
+      "loss": 1.6837,
+      "step": 5420
+    },
+    {
+      "epoch": 1.4246395375570664,
+      "grad_norm": 0.5740265250205994,
+      "learning_rate": 5.25232171018048e-05,
+      "loss": 1.6741,
+      "step": 5422
+    },
+    {
+      "epoch": 1.4251650408907282,
+      "grad_norm": 0.5686757564544678,
+      "learning_rate": 5.250569476082005e-05,
+      "loss": 1.6908,
+      "step": 5424
+    },
+    {
+      "epoch": 1.42569054422439,
+      "grad_norm": 0.765792191028595,
+      "learning_rate": 5.248817241983529e-05,
+      "loss": 1.6928,
+      "step": 5426
+    },
+    {
+      "epoch": 1.4262160475580516,
+      "grad_norm": 0.605812668800354,
+      "learning_rate": 5.2470650078850536e-05,
+      "loss": 1.7009,
+      "step": 5428
+    },
+    {
+      "epoch": 1.4267415508917134,
+      "grad_norm": 0.5373610258102417,
+      "learning_rate": 5.2453127737865784e-05,
+      "loss": 1.6906,
+      "step": 5430
+    },
+    {
+      "epoch": 1.4272670542253754,
+      "grad_norm": 0.6103860139846802,
+      "learning_rate": 5.2435605396881025e-05,
+      "loss": 1.683,
+      "step": 5432
+    },
+    {
+      "epoch": 1.427792557559037,
+      "grad_norm": 0.5279322862625122,
+      "learning_rate": 5.241808305589627e-05,
+      "loss": 1.6679,
+      "step": 5434
+    },
+    {
+      "epoch": 1.4283180608926989,
+      "grad_norm": 0.6445699334144592,
+      "learning_rate": 5.240056071491152e-05,
+      "loss": 1.6908,
+      "step": 5436
+    },
+    {
+      "epoch": 1.4288435642263606,
+      "grad_norm": 0.5819953680038452,
+      "learning_rate": 5.238303837392676e-05,
+      "loss": 1.6849,
+      "step": 5438
+    },
+    {
+      "epoch": 1.4293690675600224,
+      "grad_norm": 0.564022421836853,
+      "learning_rate": 5.236551603294201e-05,
+      "loss": 1.6799,
+      "step": 5440
+    },
+    {
+      "epoch": 1.429894570893684,
+      "grad_norm": 0.5687031149864197,
+      "learning_rate": 5.2347993691957256e-05,
+      "loss": 1.6905,
+      "step": 5442
+    },
+    {
+      "epoch": 1.4304200742273459,
+      "grad_norm": 0.6546675562858582,
+      "learning_rate": 5.233047135097249e-05,
+      "loss": 1.6752,
+      "step": 5444
+    },
+    {
+      "epoch": 1.4309455775610076,
+      "grad_norm": 0.6557585597038269,
+      "learning_rate": 5.231294900998773e-05,
+      "loss": 1.6959,
+      "step": 5446
+    },
+    {
+      "epoch": 1.4314710808946693,
+      "grad_norm": 0.6117187142372131,
+      "learning_rate": 5.229542666900298e-05,
+      "loss": 1.6983,
+      "step": 5448
+    },
+    {
+      "epoch": 1.4319965842283313,
+      "grad_norm": 0.6230733394622803,
+      "learning_rate": 5.2277904328018226e-05,
+      "loss": 1.7063,
+      "step": 5450
+    },
+    {
+      "epoch": 1.4325220875619928,
+      "grad_norm": 0.5881343483924866,
+      "learning_rate": 5.2260381987033467e-05,
+      "loss": 1.6718,
+      "step": 5452
+    },
+    {
+      "epoch": 1.4330475908956548,
+      "grad_norm": 0.606519341468811,
+      "learning_rate": 5.2242859646048714e-05,
+      "loss": 1.6876,
+      "step": 5454
+    },
+    {
+      "epoch": 1.4335730942293166,
+      "grad_norm": 0.6653869152069092,
+      "learning_rate": 5.222533730506396e-05,
+      "loss": 1.6707,
+      "step": 5456
+    },
+    {
+      "epoch": 1.4340985975629783,
+      "grad_norm": 0.5549430251121521,
+      "learning_rate": 5.22078149640792e-05,
+      "loss": 1.6994,
+      "step": 5458
+    },
+    {
+      "epoch": 1.43462410089664,
+      "grad_norm": 0.7070857882499695,
+      "learning_rate": 5.219029262309445e-05,
+      "loss": 1.7043,
+      "step": 5460
+    },
+    {
+      "epoch": 1.4351496042303018,
+      "grad_norm": 0.4982629716396332,
+      "learning_rate": 5.21727702821097e-05,
+      "loss": 1.6567,
+      "step": 5462
+    },
+    {
+      "epoch": 1.4356751075639635,
+      "grad_norm": 0.6242676973342896,
+      "learning_rate": 5.215524794112494e-05,
+      "loss": 1.6898,
+      "step": 5464
+    },
+    {
+      "epoch": 1.4362006108976253,
+      "grad_norm": 0.5926311016082764,
+      "learning_rate": 5.2137725600140186e-05,
+      "loss": 1.7012,
+      "step": 5466
+    },
+    {
+      "epoch": 1.4367261142312873,
+      "grad_norm": 0.7169223427772522,
+      "learning_rate": 5.212020325915542e-05,
+      "loss": 1.6782,
+      "step": 5468
+    },
+    {
+      "epoch": 1.4372516175649488,
+      "grad_norm": 0.7072371244430542,
+      "learning_rate": 5.210268091817067e-05,
+      "loss": 1.6442,
+      "step": 5470
+    },
+    {
+      "epoch": 1.4377771208986108,
+      "grad_norm": 0.5502269268035889,
+      "learning_rate": 5.208515857718591e-05,
+      "loss": 1.6743,
+      "step": 5472
+    },
+    {
+      "epoch": 1.4383026242322725,
+      "grad_norm": 0.7821040749549866,
+      "learning_rate": 5.2067636236201156e-05,
+      "loss": 1.7106,
+      "step": 5474
+    },
+    {
+      "epoch": 1.4388281275659343,
+      "grad_norm": 0.6043164730072021,
+      "learning_rate": 5.2050113895216403e-05,
+      "loss": 1.708,
+      "step": 5476
+    },
+    {
+      "epoch": 1.439353630899596,
+      "grad_norm": 0.5633584856987,
+      "learning_rate": 5.2032591554231644e-05,
+      "loss": 1.6696,
+      "step": 5478
+    },
+    {
+      "epoch": 1.4398791342332578,
+      "grad_norm": 0.5389137864112854,
+      "learning_rate": 5.201506921324689e-05,
+      "loss": 1.669,
+      "step": 5480
+    },
+    {
+      "epoch": 1.4404046375669195,
+      "grad_norm": 0.6789732575416565,
+      "learning_rate": 5.199754687226214e-05,
+      "loss": 1.6899,
+      "step": 5482
+    },
+    {
+      "epoch": 1.4409301409005812,
+      "grad_norm": 0.6296239495277405,
+      "learning_rate": 5.198002453127738e-05,
+      "loss": 1.6631,
+      "step": 5484
+    },
+    {
+      "epoch": 1.4414556442342432,
+      "grad_norm": 0.700206458568573,
+      "learning_rate": 5.196250219029263e-05,
+      "loss": 1.6616,
+      "step": 5486
+    },
+    {
+      "epoch": 1.441981147567905,
+      "grad_norm": 0.6670316457748413,
+      "learning_rate": 5.1944979849307875e-05,
+      "loss": 1.6699,
+      "step": 5488
+    },
+    {
+      "epoch": 1.4425066509015667,
+      "grad_norm": 0.6687952876091003,
+      "learning_rate": 5.1927457508323116e-05,
+      "loss": 1.7015,
+      "step": 5490
+    },
+    {
+      "epoch": 1.4430321542352285,
+      "grad_norm": 0.6341938972473145,
+      "learning_rate": 5.1909935167338364e-05,
+      "loss": 1.6713,
+      "step": 5492
+    },
+    {
+      "epoch": 1.4435576575688902,
+      "grad_norm": 0.5659570693969727,
+      "learning_rate": 5.18924128263536e-05,
+      "loss": 1.6662,
+      "step": 5494
+    },
+    {
+      "epoch": 1.444083160902552,
+      "grad_norm": 0.559827983379364,
+      "learning_rate": 5.1874890485368845e-05,
+      "loss": 1.6901,
+      "step": 5496
+    },
+    {
+      "epoch": 1.4446086642362137,
+      "grad_norm": 0.5669519901275635,
+      "learning_rate": 5.1857368144384086e-05,
+      "loss": 1.6966,
+      "step": 5498
+    },
+    {
+      "epoch": 1.4451341675698754,
+      "grad_norm": 0.544907808303833,
+      "learning_rate": 5.1839845803399334e-05,
+      "loss": 1.6837,
+      "step": 5500
+    },
+    {
+      "epoch": 1.4456596709035372,
+      "grad_norm": 0.510985255241394,
+      "learning_rate": 5.182232346241458e-05,
+      "loss": 1.7191,
+      "step": 5502
+    },
+    {
+      "epoch": 1.4461851742371992,
+      "grad_norm": 0.5953521728515625,
+      "learning_rate": 5.180480112142982e-05,
+      "loss": 1.6899,
+      "step": 5504
+    },
+    {
+      "epoch": 1.446710677570861,
+      "grad_norm": 0.5401633381843567,
+      "learning_rate": 5.178727878044507e-05,
+      "loss": 1.6602,
+      "step": 5506
+    },
+    {
+      "epoch": 1.4472361809045227,
+      "grad_norm": 0.6041406393051147,
+      "learning_rate": 5.176975643946032e-05,
+      "loss": 1.6573,
+      "step": 5508
+    },
+    {
+      "epoch": 1.4477616842381844,
+      "grad_norm": 0.5650386214256287,
+      "learning_rate": 5.175223409847556e-05,
+      "loss": 1.6747,
+      "step": 5510
+    },
+    {
+      "epoch": 1.4482871875718462,
+      "grad_norm": 0.6061777472496033,
+      "learning_rate": 5.1734711757490805e-05,
+      "loss": 1.677,
+      "step": 5512
+    },
+    {
+      "epoch": 1.448812690905508,
+      "grad_norm": 0.5969398617744446,
+      "learning_rate": 5.171718941650605e-05,
+      "loss": 1.6579,
+      "step": 5514
+    },
+    {
+      "epoch": 1.4493381942391697,
+      "grad_norm": 0.5931181311607361,
+      "learning_rate": 5.1699667075521294e-05,
+      "loss": 1.6772,
+      "step": 5516
+    },
+    {
+      "epoch": 1.4498636975728314,
+      "grad_norm": 0.5793447494506836,
+      "learning_rate": 5.168214473453654e-05,
+      "loss": 1.6302,
+      "step": 5518
+    },
+    {
+      "epoch": 1.4503892009064931,
+      "grad_norm": 0.5775421857833862,
+      "learning_rate": 5.1664622393551775e-05,
+      "loss": 1.6691,
+      "step": 5520
+    },
+    {
+      "epoch": 1.4509147042401551,
+      "grad_norm": 0.606317937374115,
+      "learning_rate": 5.164710005256702e-05,
+      "loss": 1.6876,
+      "step": 5522
+    },
+    {
+      "epoch": 1.4514402075738169,
+      "grad_norm": 0.5722896456718445,
+      "learning_rate": 5.1629577711582264e-05,
+      "loss": 1.643,
+      "step": 5524
+    },
+    {
+      "epoch": 1.4519657109074786,
+      "grad_norm": 0.7023299932479858,
+      "learning_rate": 5.161205537059751e-05,
+      "loss": 1.6527,
+      "step": 5526
+    },
+    {
+      "epoch": 1.4524912142411404,
+      "grad_norm": 0.5290958285331726,
+      "learning_rate": 5.159453302961276e-05,
+      "loss": 1.6556,
+      "step": 5528
+    },
+    {
+      "epoch": 1.453016717574802,
+      "grad_norm": 0.6962506771087646,
+      "learning_rate": 5.1577010688628e-05,
+      "loss": 1.6781,
+      "step": 5530
+    },
+    {
+      "epoch": 1.4535422209084639,
+      "grad_norm": 0.5442536473274231,
+      "learning_rate": 5.155948834764325e-05,
+      "loss": 1.6712,
+      "step": 5532
+    },
+    {
+      "epoch": 1.4540677242421256,
+      "grad_norm": 0.6162612438201904,
+      "learning_rate": 5.1541966006658495e-05,
+      "loss": 1.7003,
+      "step": 5534
+    },
+    {
+      "epoch": 1.4545932275757874,
+      "grad_norm": 0.5466321110725403,
+      "learning_rate": 5.1524443665673735e-05,
+      "loss": 1.6908,
+      "step": 5536
+    },
+    {
+      "epoch": 1.455118730909449,
+      "grad_norm": 0.5761128067970276,
+      "learning_rate": 5.150692132468898e-05,
+      "loss": 1.6745,
+      "step": 5538
+    },
+    {
+      "epoch": 1.455644234243111,
+      "grad_norm": 0.5764265656471252,
+      "learning_rate": 5.148939898370423e-05,
+      "loss": 1.6308,
+      "step": 5540
+    },
+    {
+      "epoch": 1.4561697375767728,
+      "grad_norm": 0.5794236063957214,
+      "learning_rate": 5.147187664271947e-05,
+      "loss": 1.7178,
+      "step": 5542
+    },
+    {
+      "epoch": 1.4566952409104346,
+      "grad_norm": 0.8042659759521484,
+      "learning_rate": 5.145435430173472e-05,
+      "loss": 1.6589,
+      "step": 5544
+    },
+    {
+      "epoch": 1.4572207442440963,
+      "grad_norm": 0.63565993309021,
+      "learning_rate": 5.143683196074995e-05,
+      "loss": 1.6601,
+      "step": 5546
+    },
+    {
+      "epoch": 1.457746247577758,
+      "grad_norm": 0.5769701600074768,
+      "learning_rate": 5.14193096197652e-05,
+      "loss": 1.6565,
+      "step": 5548
+    },
+    {
+      "epoch": 1.4582717509114198,
+      "grad_norm": 0.6259598135948181,
+      "learning_rate": 5.140178727878044e-05,
+      "loss": 1.6664,
+      "step": 5550
+    },
+    {
+      "epoch": 1.4587972542450816,
+      "grad_norm": 1.0551050901412964,
+      "learning_rate": 5.138426493779569e-05,
+      "loss": 1.7123,
+      "step": 5552
+    },
+    {
+      "epoch": 1.4593227575787435,
+      "grad_norm": 0.5595609545707703,
+      "learning_rate": 5.1366742596810936e-05,
+      "loss": 1.6494,
+      "step": 5554
+    },
+    {
+      "epoch": 1.459848260912405,
+      "grad_norm": 0.607704222202301,
+      "learning_rate": 5.134922025582618e-05,
+      "loss": 1.6585,
+      "step": 5556
+    },
+    {
+      "epoch": 1.460373764246067,
+      "grad_norm": 0.7237640023231506,
+      "learning_rate": 5.1331697914841425e-05,
+      "loss": 1.6797,
+      "step": 5558
+    },
+    {
+      "epoch": 1.4608992675797288,
+      "grad_norm": 0.7079179286956787,
+      "learning_rate": 5.131417557385667e-05,
+      "loss": 1.6882,
+      "step": 5560
+    },
+    {
+      "epoch": 1.4614247709133905,
+      "grad_norm": 0.5847229361534119,
+      "learning_rate": 5.129665323287191e-05,
+      "loss": 1.6694,
+      "step": 5562
+    },
+    {
+      "epoch": 1.4619502742470523,
+      "grad_norm": 0.5582857728004456,
+      "learning_rate": 5.127913089188716e-05,
+      "loss": 1.6723,
+      "step": 5564
+    },
+    {
+      "epoch": 1.462475777580714,
+      "grad_norm": 0.5761253833770752,
+      "learning_rate": 5.126160855090241e-05,
+      "loss": 1.6953,
+      "step": 5566
+    },
+    {
+      "epoch": 1.4630012809143758,
+      "grad_norm": 0.6386553645133972,
+      "learning_rate": 5.124408620991765e-05,
+      "loss": 1.6842,
+      "step": 5568
+    },
+    {
+      "epoch": 1.4635267842480375,
+      "grad_norm": 0.6150875687599182,
+      "learning_rate": 5.122656386893288e-05,
+      "loss": 1.6365,
+      "step": 5570
+    },
+    {
+      "epoch": 1.4640522875816995,
+      "grad_norm": 0.632836639881134,
+      "learning_rate": 5.120904152794813e-05,
+      "loss": 1.6497,
+      "step": 5572
+    },
+    {
+      "epoch": 1.464577790915361,
+      "grad_norm": 0.713758111000061,
+      "learning_rate": 5.119151918696338e-05,
+      "loss": 1.6505,
+      "step": 5574
+    },
+    {
+      "epoch": 1.465103294249023,
+      "grad_norm": 0.5495973825454712,
+      "learning_rate": 5.117399684597862e-05,
+      "loss": 1.6852,
+      "step": 5576
+    },
+    {
+      "epoch": 1.4656287975826847,
+      "grad_norm": 0.6123538613319397,
+      "learning_rate": 5.1156474504993867e-05,
+      "loss": 1.6334,
+      "step": 5578
+    },
+    {
+      "epoch": 1.4661543009163465,
+      "grad_norm": 0.6316397786140442,
+      "learning_rate": 5.1138952164009114e-05,
+      "loss": 1.6877,
+      "step": 5580
+    },
+    {
+      "epoch": 1.4666798042500082,
+      "grad_norm": 0.6072596311569214,
+      "learning_rate": 5.1121429823024355e-05,
+      "loss": 1.6586,
+      "step": 5582
+    },
+    {
+      "epoch": 1.46720530758367,
+      "grad_norm": 0.6176102757453918,
+      "learning_rate": 5.11039074820396e-05,
+      "loss": 1.6752,
+      "step": 5584
+    },
+    {
+      "epoch": 1.4677308109173317,
+      "grad_norm": 0.6328986883163452,
+      "learning_rate": 5.108638514105485e-05,
+      "loss": 1.6886,
+      "step": 5586
+    },
+    {
+      "epoch": 1.4682563142509935,
+      "grad_norm": 0.6120469570159912,
+      "learning_rate": 5.10688628000701e-05,
+      "loss": 1.6834,
+      "step": 5588
+    },
+    {
+      "epoch": 1.4687818175846554,
+      "grad_norm": 0.6649283170700073,
+      "learning_rate": 5.105134045908534e-05,
+      "loss": 1.6399,
+      "step": 5590
+    },
+    {
+      "epoch": 1.469307320918317,
+      "grad_norm": 0.6802614331245422,
+      "learning_rate": 5.1033818118100586e-05,
+      "loss": 1.7099,
+      "step": 5592
+    },
+    {
+      "epoch": 1.469832824251979,
+      "grad_norm": 0.5907096862792969,
+      "learning_rate": 5.1016295777115833e-05,
+      "loss": 1.6647,
+      "step": 5594
+    },
+    {
+      "epoch": 1.4703583275856407,
+      "grad_norm": 0.5924010276794434,
+      "learning_rate": 5.099877343613107e-05,
+      "loss": 1.6655,
+      "step": 5596
+    },
+    {
+      "epoch": 1.4708838309193024,
+      "grad_norm": 0.5676859617233276,
+      "learning_rate": 5.098125109514631e-05,
+      "loss": 1.6651,
+      "step": 5598
+    },
+    {
+      "epoch": 1.4714093342529642,
+      "grad_norm": 0.5564827919006348,
+      "learning_rate": 5.0963728754161556e-05,
+      "loss": 1.7036,
+      "step": 5600
+    },
+    {
+      "epoch": 1.4714093342529642,
+      "eval_loss": 1.6887091398239136,
+      "eval_runtime": 487.1901,
+      "eval_samples_per_second": 249.982,
+      "eval_steps_per_second": 31.249,
+      "step": 5600
+    },
+    {
+      "epoch": 1.471934837586626,
+      "grad_norm": 0.6894435286521912,
+      "learning_rate": 5.09462064131768e-05,
+      "loss": 1.6565,
+      "step": 5602
+    },
+    {
+      "epoch": 1.4724603409202877,
+      "grad_norm": 0.5259155035018921,
+      "learning_rate": 5.0928684072192044e-05,
+      "loss": 1.6428,
+      "step": 5604
+    },
+    {
+      "epoch": 1.4729858442539494,
+      "grad_norm": 0.6473843455314636,
+      "learning_rate": 5.091116173120729e-05,
+      "loss": 1.6773,
+      "step": 5606
+    },
+    {
+      "epoch": 1.4735113475876114,
+      "grad_norm": 0.5902252793312073,
+      "learning_rate": 5.089363939022254e-05,
+      "loss": 1.683,
+      "step": 5608
+    },
+    {
+      "epoch": 1.474036850921273,
+      "grad_norm": 0.5241686105728149,
+      "learning_rate": 5.087611704923778e-05,
+      "loss": 1.7068,
+      "step": 5610
+    },
+    {
+      "epoch": 1.4745623542549349,
+      "grad_norm": 0.5719613432884216,
+      "learning_rate": 5.085859470825303e-05,
+      "loss": 1.6997,
+      "step": 5612
+    },
+    {
+      "epoch": 1.4750878575885966,
+      "grad_norm": 0.7695503234863281,
+      "learning_rate": 5.0841072367268275e-05,
+      "loss": 1.7057,
+      "step": 5614
+    },
+    {
+      "epoch": 1.4756133609222584,
+      "grad_norm": 0.5506905913352966,
+      "learning_rate": 5.0823550026283516e-05,
+      "loss": 1.6853,
+      "step": 5616
+    },
+    {
+      "epoch": 1.4761388642559201,
+      "grad_norm": 0.5574632883071899,
+      "learning_rate": 5.0806027685298764e-05,
+      "loss": 1.7207,
+      "step": 5618
+    },
+    {
+      "epoch": 1.4766643675895819,
+      "grad_norm": 0.5256906747817993,
+      "learning_rate": 5.078850534431401e-05,
+      "loss": 1.6743,
+      "step": 5620
+    },
+    {
+      "epoch": 1.4771898709232436,
+      "grad_norm": 1.0742486715316772,
+      "learning_rate": 5.0770983003329245e-05,
+      "loss": 1.6613,
+      "step": 5622
+    },
+    {
+      "epoch": 1.4777153742569054,
+      "grad_norm": 0.6463772654533386,
+      "learning_rate": 5.0753460662344486e-05,
+      "loss": 1.6872,
+      "step": 5624
+    },
+    {
+      "epoch": 1.4782408775905673,
+      "grad_norm": 0.5381261706352234,
+      "learning_rate": 5.0735938321359733e-05,
+      "loss": 1.6418,
+      "step": 5626
+    },
+    {
+      "epoch": 1.4787663809242289,
+      "grad_norm": 0.7155745029449463,
+      "learning_rate": 5.071841598037498e-05,
+      "loss": 1.7034,
+      "step": 5628
+    },
+    {
+      "epoch": 1.4792918842578908,
+      "grad_norm": 0.8163059949874878,
+      "learning_rate": 5.070089363939022e-05,
+      "loss": 1.6595,
+      "step": 5630
+    },
+    {
+      "epoch": 1.4798173875915526,
+      "grad_norm": 0.6401390433311462,
+      "learning_rate": 5.068337129840547e-05,
+      "loss": 1.6658,
+      "step": 5632
+    },
+    {
+      "epoch": 1.4803428909252143,
+      "grad_norm": 0.5915647149085999,
+      "learning_rate": 5.066584895742072e-05,
+      "loss": 1.6495,
+      "step": 5634
+    },
+    {
+      "epoch": 1.480868394258876,
+      "grad_norm": 0.5638352632522583,
+      "learning_rate": 5.064832661643596e-05,
+      "loss": 1.6619,
+      "step": 5636
+    },
+    {
+      "epoch": 1.4813938975925378,
+      "grad_norm": 0.6860741376876831,
+      "learning_rate": 5.0630804275451205e-05,
+      "loss": 1.6827,
+      "step": 5638
+    },
+    {
+      "epoch": 1.4819194009261996,
+      "grad_norm": 0.5409403443336487,
+      "learning_rate": 5.061328193446645e-05,
+      "loss": 1.6391,
+      "step": 5640
+    },
+    {
+      "epoch": 1.4824449042598613,
+      "grad_norm": 0.6866961717605591,
+      "learning_rate": 5.0595759593481694e-05,
+      "loss": 1.6657,
+      "step": 5642
+    },
+    {
+      "epoch": 1.4829704075935233,
+      "grad_norm": 0.6035512089729309,
+      "learning_rate": 5.057823725249694e-05,
+      "loss": 1.6962,
+      "step": 5644
+    },
+    {
+      "epoch": 1.483495910927185,
+      "grad_norm": 0.539889931678772,
+      "learning_rate": 5.056071491151219e-05,
+      "loss": 1.676,
+      "step": 5646
+    },
+    {
+      "epoch": 1.4840214142608468,
+      "grad_norm": 0.5658326148986816,
+      "learning_rate": 5.054319257052742e-05,
+      "loss": 1.6876,
+      "step": 5648
+    },
+    {
+      "epoch": 1.4845469175945085,
+      "grad_norm": 0.9115592241287231,
+      "learning_rate": 5.0525670229542664e-05,
+      "loss": 1.6691,
+      "step": 5650
+    },
+    {
+      "epoch": 1.4850724209281703,
+      "grad_norm": 0.6758273839950562,
+      "learning_rate": 5.050814788855791e-05,
+      "loss": 1.6405,
+      "step": 5652
+    },
+    {
+      "epoch": 1.485597924261832,
+      "grad_norm": 0.6089572310447693,
+      "learning_rate": 5.049062554757316e-05,
+      "loss": 1.6954,
+      "step": 5654
+    },
+    {
+      "epoch": 1.4861234275954938,
+      "grad_norm": 0.5882745981216431,
+      "learning_rate": 5.04731032065884e-05,
+      "loss": 1.6768,
+      "step": 5656
+    },
+    {
+      "epoch": 1.4866489309291555,
+      "grad_norm": 0.6646392345428467,
+      "learning_rate": 5.045558086560365e-05,
+      "loss": 1.6832,
+      "step": 5658
+    },
+    {
+      "epoch": 1.4871744342628173,
+      "grad_norm": 0.7076845765113831,
+      "learning_rate": 5.0438058524618895e-05,
+      "loss": 1.6734,
+      "step": 5660
+    },
+    {
+      "epoch": 1.4876999375964792,
+      "grad_norm": 0.6591199636459351,
+      "learning_rate": 5.0420536183634135e-05,
+      "loss": 1.7321,
+      "step": 5662
+    },
+    {
+      "epoch": 1.488225440930141,
+      "grad_norm": 0.5403279066085815,
+      "learning_rate": 5.040301384264938e-05,
+      "loss": 1.6665,
+      "step": 5664
+    },
+    {
+      "epoch": 1.4887509442638027,
+      "grad_norm": 0.5921054482460022,
+      "learning_rate": 5.038549150166463e-05,
+      "loss": 1.6738,
+      "step": 5666
+    },
+    {
+      "epoch": 1.4892764475974645,
+      "grad_norm": 0.6299567818641663,
+      "learning_rate": 5.036796916067987e-05,
+      "loss": 1.6627,
+      "step": 5668
+    },
+    {
+      "epoch": 1.4898019509311262,
+      "grad_norm": 0.6079999208450317,
+      "learning_rate": 5.035044681969512e-05,
+      "loss": 1.6921,
+      "step": 5670
+    },
+    {
+      "epoch": 1.490327454264788,
+      "grad_norm": 0.5412185788154602,
+      "learning_rate": 5.033292447871035e-05,
+      "loss": 1.6779,
+      "step": 5672
+    },
+    {
+      "epoch": 1.4908529575984497,
+      "grad_norm": 0.603542685508728,
+      "learning_rate": 5.03154021377256e-05,
+      "loss": 1.6948,
+      "step": 5674
+    },
+    {
+      "epoch": 1.4913784609321115,
+      "grad_norm": 0.818874716758728,
+      "learning_rate": 5.029787979674084e-05,
+      "loss": 1.6824,
+      "step": 5676
+    },
+    {
+      "epoch": 1.4919039642657732,
+      "grad_norm": 0.5792141556739807,
+      "learning_rate": 5.028035745575609e-05,
+      "loss": 1.679,
+      "step": 5678
+    },
+    {
+      "epoch": 1.4924294675994352,
+      "grad_norm": 0.576353907585144,
+      "learning_rate": 5.0262835114771336e-05,
+      "loss": 1.6835,
+      "step": 5680
+    },
+    {
+      "epoch": 1.492954970933097,
+      "grad_norm": 0.5527434349060059,
+      "learning_rate": 5.024531277378658e-05,
+      "loss": 1.7095,
+      "step": 5682
+    },
+    {
+      "epoch": 1.4934804742667587,
+      "grad_norm": 0.5520890355110168,
+      "learning_rate": 5.0227790432801825e-05,
+      "loss": 1.6672,
+      "step": 5684
+    },
+    {
+      "epoch": 1.4940059776004204,
+      "grad_norm": 0.5043492317199707,
+      "learning_rate": 5.021026809181707e-05,
+      "loss": 1.6717,
+      "step": 5686
+    },
+    {
+      "epoch": 1.4945314809340822,
+      "grad_norm": 0.6247971653938293,
+      "learning_rate": 5.019274575083231e-05,
+      "loss": 1.6792,
+      "step": 5688
+    },
+    {
+      "epoch": 1.495056984267744,
+      "grad_norm": 0.6311819553375244,
+      "learning_rate": 5.017522340984756e-05,
+      "loss": 1.6628,
+      "step": 5690
+    },
+    {
+      "epoch": 1.4955824876014057,
+      "grad_norm": 0.5443778038024902,
+      "learning_rate": 5.015770106886281e-05,
+      "loss": 1.6362,
+      "step": 5692
+    },
+    {
+      "epoch": 1.4961079909350674,
+      "grad_norm": 0.6705971360206604,
+      "learning_rate": 5.014017872787805e-05,
+      "loss": 1.6975,
+      "step": 5694
+    },
+    {
+      "epoch": 1.4966334942687292,
+      "grad_norm": 0.6342272758483887,
+      "learning_rate": 5.0122656386893297e-05,
+      "loss": 1.6652,
+      "step": 5696
+    },
+    {
+      "epoch": 1.4971589976023911,
+      "grad_norm": 0.615489661693573,
+      "learning_rate": 5.010513404590853e-05,
+      "loss": 1.6503,
+      "step": 5698
+    },
+    {
+      "epoch": 1.4976845009360529,
+      "grad_norm": 0.5811960697174072,
+      "learning_rate": 5.008761170492378e-05,
+      "loss": 1.6369,
+      "step": 5700
+    },
+    {
+      "epoch": 1.4982100042697146,
+      "grad_norm": 0.5534473061561584,
+      "learning_rate": 5.007008936393902e-05,
+      "loss": 1.6311,
+      "step": 5702
+    },
+    {
+      "epoch": 1.4987355076033764,
+      "grad_norm": 0.6678853034973145,
+      "learning_rate": 5.0052567022954266e-05,
+      "loss": 1.6513,
+      "step": 5704
+    },
+    {
+      "epoch": 1.4992610109370381,
+      "grad_norm": 0.5876911878585815,
+      "learning_rate": 5.0035044681969514e-05,
+      "loss": 1.6847,
+      "step": 5706
+    },
+    {
+      "epoch": 1.4997865142706999,
+      "grad_norm": 0.6282883286476135,
+      "learning_rate": 5.0017522340984755e-05,
+      "loss": 1.6909,
+      "step": 5708
+    },
+    {
+      "epoch": 1.5003120176043616,
+      "grad_norm": 0.5211758613586426,
+      "learning_rate": 5e-05,
+      "loss": 1.6478,
+      "step": 5710
+    },
+    {
+      "epoch": 1.5008375209380236,
+      "grad_norm": 0.5276235938072205,
+      "learning_rate": 4.998247765901525e-05,
+      "loss": 1.6706,
+      "step": 5712
+    },
+    {
+      "epoch": 1.501363024271685,
+      "grad_norm": 0.6381866335868835,
+      "learning_rate": 4.996495531803049e-05,
+      "loss": 1.672,
+      "step": 5714
+    },
+    {
+      "epoch": 1.501888527605347,
+      "grad_norm": 0.5875132083892822,
+      "learning_rate": 4.994743297704574e-05,
+      "loss": 1.6611,
+      "step": 5716
+    },
+    {
+      "epoch": 1.5024140309390086,
+      "grad_norm": 0.5340953469276428,
+      "learning_rate": 4.992991063606098e-05,
+      "loss": 1.6612,
+      "step": 5718
+    },
+    {
+      "epoch": 1.5029395342726706,
+      "grad_norm": 0.6308703422546387,
+      "learning_rate": 4.991238829507623e-05,
+      "loss": 1.6988,
+      "step": 5720
+    },
+    {
+      "epoch": 1.5034650376063323,
+      "grad_norm": 0.6118327975273132,
+      "learning_rate": 4.989486595409147e-05,
+      "loss": 1.6677,
+      "step": 5722
+    },
+    {
+      "epoch": 1.503990540939994,
+      "grad_norm": 0.6005641222000122,
+      "learning_rate": 4.9877343613106715e-05,
+      "loss": 1.6629,
+      "step": 5724
+    },
+    {
+      "epoch": 1.5045160442736558,
+      "grad_norm": 0.6117488145828247,
+      "learning_rate": 4.985982127212196e-05,
+      "loss": 1.6429,
+      "step": 5726
+    },
+    {
+      "epoch": 1.5050415476073176,
+      "grad_norm": 0.7005740404129028,
+      "learning_rate": 4.98422989311372e-05,
+      "loss": 1.6896,
+      "step": 5728
+    },
+    {
+      "epoch": 1.5055670509409795,
+      "grad_norm": 0.5488820672035217,
+      "learning_rate": 4.9824776590152444e-05,
+      "loss": 1.6603,
+      "step": 5730
+    },
+    {
+      "epoch": 1.506092554274641,
+      "grad_norm": 0.6293522119522095,
+      "learning_rate": 4.980725424916769e-05,
+      "loss": 1.6465,
+      "step": 5732
+    },
+    {
+      "epoch": 1.506618057608303,
+      "grad_norm": 0.5625261664390564,
+      "learning_rate": 4.978973190818293e-05,
+      "loss": 1.6572,
+      "step": 5734
+    },
+    {
+      "epoch": 1.5071435609419648,
+      "grad_norm": 0.5663225054740906,
+      "learning_rate": 4.977220956719818e-05,
+      "loss": 1.6674,
+      "step": 5736
+    },
+    {
+      "epoch": 1.5076690642756265,
+      "grad_norm": 0.5049892663955688,
+      "learning_rate": 4.975468722621343e-05,
+      "loss": 1.6529,
+      "step": 5738
+    },
+    {
+      "epoch": 1.5081945676092883,
+      "grad_norm": 0.5977827906608582,
+      "learning_rate": 4.973716488522867e-05,
+      "loss": 1.6644,
+      "step": 5740
+    },
+    {
+      "epoch": 1.50872007094295,
+      "grad_norm": 0.5831950306892395,
+      "learning_rate": 4.971964254424391e-05,
+      "loss": 1.6659,
+      "step": 5742
+    },
+    {
+      "epoch": 1.5092455742766118,
+      "grad_norm": 0.6863638162612915,
+      "learning_rate": 4.970212020325916e-05,
+      "loss": 1.6686,
+      "step": 5744
+    },
+    {
+      "epoch": 1.5097710776102735,
+      "grad_norm": 0.5951880812644958,
+      "learning_rate": 4.9684597862274404e-05,
+      "loss": 1.6547,
+      "step": 5746
+    },
+    {
+      "epoch": 1.5102965809439355,
+      "grad_norm": 0.6561357975006104,
+      "learning_rate": 4.9667075521289645e-05,
+      "loss": 1.6984,
+      "step": 5748
+    },
+    {
+      "epoch": 1.510822084277597,
+      "grad_norm": 0.5892223715782166,
+      "learning_rate": 4.964955318030489e-05,
+      "loss": 1.6927,
+      "step": 5750
+    },
+    {
+      "epoch": 1.511347587611259,
+      "grad_norm": 0.5931475162506104,
+      "learning_rate": 4.963203083932014e-05,
+      "loss": 1.6643,
+      "step": 5752
+    },
+    {
+      "epoch": 1.5118730909449207,
+      "grad_norm": 0.5545600652694702,
+      "learning_rate": 4.961450849833538e-05,
+      "loss": 1.6818,
+      "step": 5754
+    },
+    {
+      "epoch": 1.5123985942785825,
+      "grad_norm": 0.606566309928894,
+      "learning_rate": 4.959698615735062e-05,
+      "loss": 1.6676,
+      "step": 5756
+    },
+    {
+      "epoch": 1.5129240976122442,
+      "grad_norm": 0.5394220352172852,
+      "learning_rate": 4.957946381636587e-05,
+      "loss": 1.6563,
+      "step": 5758
+    },
+    {
+      "epoch": 1.513449600945906,
+      "grad_norm": 0.5019289255142212,
+      "learning_rate": 4.956194147538111e-05,
+      "loss": 1.6536,
+      "step": 5760
+    },
+    {
+      "epoch": 1.5139751042795677,
+      "grad_norm": 0.5706535577774048,
+      "learning_rate": 4.954441913439636e-05,
+      "loss": 1.6781,
+      "step": 5762
+    },
+    {
+      "epoch": 1.5145006076132295,
+      "grad_norm": 0.7017900347709656,
+      "learning_rate": 4.9526896793411605e-05,
+      "loss": 1.7036,
+      "step": 5764
+    },
+    {
+      "epoch": 1.5150261109468914,
+      "grad_norm": 0.627980649471283,
+      "learning_rate": 4.9509374452426846e-05,
+      "loss": 1.7071,
+      "step": 5766
+    },
+    {
+      "epoch": 1.515551614280553,
+      "grad_norm": 0.5035885572433472,
+      "learning_rate": 4.949185211144209e-05,
+      "loss": 1.6837,
+      "step": 5768
+    },
+    {
+      "epoch": 1.516077117614215,
+      "grad_norm": 0.5663711428642273,
+      "learning_rate": 4.9474329770457334e-05,
+      "loss": 1.6622,
+      "step": 5770
+    },
+    {
+      "epoch": 1.5166026209478767,
+      "grad_norm": 0.5828589797019958,
+      "learning_rate": 4.945680742947258e-05,
+      "loss": 1.7211,
+      "step": 5772
+    },
+    {
+      "epoch": 1.5171281242815384,
+      "grad_norm": 0.644883930683136,
+      "learning_rate": 4.943928508848782e-05,
+      "loss": 1.6592,
+      "step": 5774
+    },
+    {
+      "epoch": 1.5176536276152002,
+      "grad_norm": 0.5231825113296509,
+      "learning_rate": 4.942176274750307e-05,
+      "loss": 1.6896,
+      "step": 5776
+    },
+    {
+      "epoch": 1.518179130948862,
+      "grad_norm": 0.6508246064186096,
+      "learning_rate": 4.940424040651832e-05,
+      "loss": 1.709,
+      "step": 5778
+    },
+    {
+      "epoch": 1.518704634282524,
+      "grad_norm": 0.7368707656860352,
+      "learning_rate": 4.938671806553356e-05,
+      "loss": 1.6449,
+      "step": 5780
+    },
+    {
+      "epoch": 1.5192301376161854,
+      "grad_norm": 0.5632103085517883,
+      "learning_rate": 4.93691957245488e-05,
+      "loss": 1.6599,
+      "step": 5782
+    },
+    {
+      "epoch": 1.5197556409498474,
+      "grad_norm": 0.5499205589294434,
+      "learning_rate": 4.935167338356405e-05,
+      "loss": 1.6798,
+      "step": 5784
+    },
+    {
+      "epoch": 1.520281144283509,
+      "grad_norm": 0.6031380295753479,
+      "learning_rate": 4.933415104257929e-05,
+      "loss": 1.7011,
+      "step": 5786
+    },
+    {
+      "epoch": 1.5208066476171709,
+      "grad_norm": 0.6366339325904846,
+      "learning_rate": 4.9316628701594535e-05,
+      "loss": 1.6722,
+      "step": 5788
+    },
+    {
+      "epoch": 1.5213321509508326,
+      "grad_norm": 0.6513427495956421,
+      "learning_rate": 4.929910636060978e-05,
+      "loss": 1.6998,
+      "step": 5790
+    },
+    {
+      "epoch": 1.5218576542844944,
+      "grad_norm": 0.5762157440185547,
+      "learning_rate": 4.9281584019625024e-05,
+      "loss": 1.7176,
+      "step": 5792
+    },
+    {
+      "epoch": 1.5223831576181561,
+      "grad_norm": 0.5817427635192871,
+      "learning_rate": 4.9264061678640265e-05,
+      "loss": 1.6686,
+      "step": 5794
+    },
+    {
+      "epoch": 1.5229086609518179,
+      "grad_norm": 0.6498470902442932,
+      "learning_rate": 4.924653933765551e-05,
+      "loss": 1.6536,
+      "step": 5796
+    },
+    {
+      "epoch": 1.5234341642854798,
+      "grad_norm": 0.5209649205207825,
+      "learning_rate": 4.922901699667076e-05,
+      "loss": 1.6711,
+      "step": 5798
+    },
+    {
+      "epoch": 1.5239596676191414,
+      "grad_norm": 0.5542787909507751,
+      "learning_rate": 4.9211494655686e-05,
+      "loss": 1.6931,
+      "step": 5800
+    },
+    {
+      "epoch": 1.5244851709528033,
+      "grad_norm": 0.5448528528213501,
+      "learning_rate": 4.919397231470125e-05,
+      "loss": 1.6952,
+      "step": 5802
+    },
+    {
+      "epoch": 1.5250106742864649,
+      "grad_norm": 0.6163997054100037,
+      "learning_rate": 4.9176449973716496e-05,
+      "loss": 1.638,
+      "step": 5804
+    },
+    {
+      "epoch": 1.5255361776201268,
+      "grad_norm": 0.5389247536659241,
+      "learning_rate": 4.915892763273173e-05,
+      "loss": 1.7243,
+      "step": 5806
+    },
+    {
+      "epoch": 1.5260616809537886,
+      "grad_norm": 0.5348747372627258,
+      "learning_rate": 4.914140529174698e-05,
+      "loss": 1.699,
+      "step": 5808
+    },
+    {
+      "epoch": 1.5265871842874503,
+      "grad_norm": 0.5715484023094177,
+      "learning_rate": 4.9123882950762225e-05,
+      "loss": 1.6672,
+      "step": 5810
+    },
+    {
+      "epoch": 1.527112687621112,
+      "grad_norm": 0.64728844165802,
+      "learning_rate": 4.9106360609777465e-05,
+      "loss": 1.66,
+      "step": 5812
+    },
+    {
+      "epoch": 1.5276381909547738,
+      "grad_norm": 0.5770621299743652,
+      "learning_rate": 4.908883826879271e-05,
+      "loss": 1.6614,
+      "step": 5814
+    },
+    {
+      "epoch": 1.5281636942884358,
+      "grad_norm": 0.9658989310264587,
+      "learning_rate": 4.907131592780796e-05,
+      "loss": 1.6941,
+      "step": 5816
+    },
+    {
+      "epoch": 1.5286891976220973,
+      "grad_norm": 0.6207730174064636,
+      "learning_rate": 4.90537935868232e-05,
+      "loss": 1.6654,
+      "step": 5818
+    },
+    {
+      "epoch": 1.5292147009557593,
+      "grad_norm": 0.5695872902870178,
+      "learning_rate": 4.903627124583844e-05,
+      "loss": 1.6503,
+      "step": 5820
+    },
+    {
+      "epoch": 1.5297402042894208,
+      "grad_norm": 0.7392182946205139,
+      "learning_rate": 4.901874890485369e-05,
+      "loss": 1.6731,
+      "step": 5822
+    },
+    {
+      "epoch": 1.5302657076230828,
+      "grad_norm": 0.5596084594726562,
+      "learning_rate": 4.900122656386894e-05,
+      "loss": 1.6816,
+      "step": 5824
+    },
+    {
+      "epoch": 1.5307912109567445,
+      "grad_norm": 0.5417515635490417,
+      "learning_rate": 4.898370422288418e-05,
+      "loss": 1.6556,
+      "step": 5826
+    },
+    {
+      "epoch": 1.5313167142904063,
+      "grad_norm": 0.6508921980857849,
+      "learning_rate": 4.8966181881899426e-05,
+      "loss": 1.684,
+      "step": 5828
+    },
+    {
+      "epoch": 1.531842217624068,
+      "grad_norm": 0.562555193901062,
+      "learning_rate": 4.894865954091467e-05,
+      "loss": 1.6706,
+      "step": 5830
+    },
+    {
+      "epoch": 1.5323677209577298,
+      "grad_norm": 0.6852477788925171,
+      "learning_rate": 4.893113719992991e-05,
+      "loss": 1.6604,
+      "step": 5832
+    },
+    {
+      "epoch": 1.5328932242913917,
+      "grad_norm": 0.8392126560211182,
+      "learning_rate": 4.8913614858945155e-05,
+      "loss": 1.6496,
+      "step": 5834
+    },
+    {
+      "epoch": 1.5334187276250533,
+      "grad_norm": 0.635260283946991,
+      "learning_rate": 4.88960925179604e-05,
+      "loss": 1.6715,
+      "step": 5836
+    },
+    {
+      "epoch": 1.5339442309587152,
+      "grad_norm": 0.5033748149871826,
+      "learning_rate": 4.887857017697565e-05,
+      "loss": 1.6391,
+      "step": 5838
+    },
+    {
+      "epoch": 1.5344697342923768,
+      "grad_norm": 0.6123384237289429,
+      "learning_rate": 4.886104783599089e-05,
+      "loss": 1.6689,
+      "step": 5840
+    },
+    {
+      "epoch": 1.5349952376260387,
+      "grad_norm": 0.8817549347877502,
+      "learning_rate": 4.884352549500614e-05,
+      "loss": 1.7004,
+      "step": 5842
+    },
+    {
+      "epoch": 1.5355207409597005,
+      "grad_norm": 0.5635519027709961,
+      "learning_rate": 4.882600315402138e-05,
+      "loss": 1.6854,
+      "step": 5844
+    },
+    {
+      "epoch": 1.5360462442933622,
+      "grad_norm": 0.5673317313194275,
+      "learning_rate": 4.880848081303662e-05,
+      "loss": 1.6538,
+      "step": 5846
+    },
+    {
+      "epoch": 1.536571747627024,
+      "grad_norm": 0.5565547347068787,
+      "learning_rate": 4.879095847205187e-05,
+      "loss": 1.6524,
+      "step": 5848
+    },
+    {
+      "epoch": 1.5370972509606857,
+      "grad_norm": 0.5782384276390076,
+      "learning_rate": 4.8773436131067115e-05,
+      "loss": 1.6929,
+      "step": 5850
+    },
+    {
+      "epoch": 1.5376227542943477,
+      "grad_norm": 0.6354172229766846,
+      "learning_rate": 4.8755913790082356e-05,
+      "loss": 1.6959,
+      "step": 5852
+    },
+    {
+      "epoch": 1.5381482576280092,
+      "grad_norm": 0.8500383496284485,
+      "learning_rate": 4.87383914490976e-05,
+      "loss": 1.6716,
+      "step": 5854
+    },
+    {
+      "epoch": 1.5386737609616712,
+      "grad_norm": 0.6829351782798767,
+      "learning_rate": 4.872086910811285e-05,
+      "loss": 1.619,
+      "step": 5856
+    },
+    {
+      "epoch": 1.5391992642953327,
+      "grad_norm": 0.5141264200210571,
+      "learning_rate": 4.870334676712809e-05,
+      "loss": 1.6637,
+      "step": 5858
+    },
+    {
+      "epoch": 1.5397247676289947,
+      "grad_norm": 0.5253787636756897,
+      "learning_rate": 4.868582442614333e-05,
+      "loss": 1.6297,
+      "step": 5860
+    },
+    {
+      "epoch": 1.5402502709626564,
+      "grad_norm": 0.5017191171646118,
+      "learning_rate": 4.866830208515858e-05,
+      "loss": 1.6502,
+      "step": 5862
+    },
+    {
+      "epoch": 1.5407757742963182,
+      "grad_norm": 0.5245093703269958,
+      "learning_rate": 4.865077974417383e-05,
+      "loss": 1.6876,
+      "step": 5864
+    },
+    {
+      "epoch": 1.54130127762998,
+      "grad_norm": 0.6714510321617126,
+      "learning_rate": 4.863325740318907e-05,
+      "loss": 1.6635,
+      "step": 5866
+    },
+    {
+      "epoch": 1.5418267809636417,
+      "grad_norm": 0.47801029682159424,
+      "learning_rate": 4.8615735062204316e-05,
+      "loss": 1.673,
+      "step": 5868
+    },
+    {
+      "epoch": 1.5423522842973036,
+      "grad_norm": 0.5477654337882996,
+      "learning_rate": 4.859821272121956e-05,
+      "loss": 1.7127,
+      "step": 5870
+    },
+    {
+      "epoch": 1.5428777876309652,
+      "grad_norm": 0.6024754643440247,
+      "learning_rate": 4.85806903802348e-05,
+      "loss": 1.6402,
+      "step": 5872
+    },
+    {
+      "epoch": 1.5434032909646271,
+      "grad_norm": 0.5506448149681091,
+      "learning_rate": 4.8563168039250045e-05,
+      "loss": 1.6795,
+      "step": 5874
+    },
+    {
+      "epoch": 1.5439287942982887,
+      "grad_norm": 0.6219335794448853,
+      "learning_rate": 4.854564569826529e-05,
+      "loss": 1.7043,
+      "step": 5876
+    },
+    {
+      "epoch": 1.5444542976319506,
+      "grad_norm": 0.5376720428466797,
+      "learning_rate": 4.8528123357280533e-05,
+      "loss": 1.6416,
+      "step": 5878
+    },
+    {
+      "epoch": 1.5449798009656124,
+      "grad_norm": 0.5363356471061707,
+      "learning_rate": 4.851060101629578e-05,
+      "loss": 1.6787,
+      "step": 5880
+    },
+    {
+      "epoch": 1.5455053042992741,
+      "grad_norm": 0.6602938175201416,
+      "learning_rate": 4.849307867531102e-05,
+      "loss": 1.6648,
+      "step": 5882
+    },
+    {
+      "epoch": 1.5460308076329359,
+      "grad_norm": 0.5136982798576355,
+      "learning_rate": 4.847555633432627e-05,
+      "loss": 1.6636,
+      "step": 5884
+    },
+    {
+      "epoch": 1.5465563109665976,
+      "grad_norm": 0.5922480225563049,
+      "learning_rate": 4.845803399334151e-05,
+      "loss": 1.645,
+      "step": 5886
+    },
+    {
+      "epoch": 1.5470818143002596,
+      "grad_norm": 0.6566057205200195,
+      "learning_rate": 4.844051165235676e-05,
+      "loss": 1.7023,
+      "step": 5888
+    },
+    {
+      "epoch": 1.5476073176339211,
+      "grad_norm": 0.5478838682174683,
+      "learning_rate": 4.8422989311372005e-05,
+      "loss": 1.6359,
+      "step": 5890
+    },
+    {
+      "epoch": 1.548132820967583,
+      "grad_norm": 0.5223366618156433,
+      "learning_rate": 4.8405466970387246e-05,
+      "loss": 1.6827,
+      "step": 5892
+    },
+    {
+      "epoch": 1.5486583243012448,
+      "grad_norm": 0.5406147837638855,
+      "learning_rate": 4.8387944629402494e-05,
+      "loss": 1.6776,
+      "step": 5894
+    },
+    {
+      "epoch": 1.5491838276349066,
+      "grad_norm": 0.5778117179870605,
+      "learning_rate": 4.8370422288417734e-05,
+      "loss": 1.6862,
+      "step": 5896
+    },
+    {
+      "epoch": 1.5497093309685683,
+      "grad_norm": 0.7566986083984375,
+      "learning_rate": 4.8352899947432975e-05,
+      "loss": 1.6777,
+      "step": 5898
+    },
+    {
+      "epoch": 1.55023483430223,
+      "grad_norm": 0.5816596150398254,
+      "learning_rate": 4.833537760644822e-05,
+      "loss": 1.6782,
+      "step": 5900
+    },
+    {
+      "epoch": 1.5507603376358918,
+      "grad_norm": 0.6017654538154602,
+      "learning_rate": 4.831785526546347e-05,
+      "loss": 1.6683,
+      "step": 5902
+    },
+    {
+      "epoch": 1.5512858409695536,
+      "grad_norm": 0.6033945679664612,
+      "learning_rate": 4.830033292447871e-05,
+      "loss": 1.6734,
+      "step": 5904
+    },
+    {
+      "epoch": 1.5518113443032155,
+      "grad_norm": 0.6159767508506775,
+      "learning_rate": 4.828281058349396e-05,
+      "loss": 1.6499,
+      "step": 5906
+    },
+    {
+      "epoch": 1.552336847636877,
+      "grad_norm": 0.5407187938690186,
+      "learning_rate": 4.82652882425092e-05,
+      "loss": 1.6808,
+      "step": 5908
+    },
+    {
+      "epoch": 1.552862350970539,
+      "grad_norm": 0.6374317407608032,
+      "learning_rate": 4.824776590152445e-05,
+      "loss": 1.6735,
+      "step": 5910
+    },
+    {
+      "epoch": 1.5533878543042008,
+      "grad_norm": 0.6518111228942871,
+      "learning_rate": 4.823024356053969e-05,
+      "loss": 1.6818,
+      "step": 5912
+    },
+    {
+      "epoch": 1.5539133576378625,
+      "grad_norm": 0.5397405028343201,
+      "learning_rate": 4.8212721219554935e-05,
+      "loss": 1.6484,
+      "step": 5914
+    },
+    {
+      "epoch": 1.5544388609715243,
+      "grad_norm": 0.5633911490440369,
+      "learning_rate": 4.819519887857018e-05,
+      "loss": 1.6438,
+      "step": 5916
+    },
+    {
+      "epoch": 1.554964364305186,
+      "grad_norm": 0.5151371359825134,
+      "learning_rate": 4.8177676537585424e-05,
+      "loss": 1.6492,
+      "step": 5918
+    },
+    {
+      "epoch": 1.5554898676388478,
+      "grad_norm": 0.5781606435775757,
+      "learning_rate": 4.816015419660067e-05,
+      "loss": 1.6881,
+      "step": 5920
+    },
+    {
+      "epoch": 1.5560153709725095,
+      "grad_norm": 0.6180148720741272,
+      "learning_rate": 4.814263185561591e-05,
+      "loss": 1.6847,
+      "step": 5922
+    },
+    {
+      "epoch": 1.5565408743061715,
+      "grad_norm": 0.5826863646507263,
+      "learning_rate": 4.812510951463115e-05,
+      "loss": 1.6514,
+      "step": 5924
+    },
+    {
+      "epoch": 1.557066377639833,
+      "grad_norm": 0.5639335513114929,
+      "learning_rate": 4.81075871736464e-05,
+      "loss": 1.6146,
+      "step": 5926
+    },
+    {
+      "epoch": 1.557591880973495,
+      "grad_norm": 0.5149716734886169,
+      "learning_rate": 4.809006483266165e-05,
+      "loss": 1.7064,
+      "step": 5928
+    },
+    {
+      "epoch": 1.5581173843071567,
+      "grad_norm": 0.5117688179016113,
+      "learning_rate": 4.807254249167689e-05,
+      "loss": 1.6757,
+      "step": 5930
+    },
+    {
+      "epoch": 1.5586428876408185,
+      "grad_norm": 0.6055058836936951,
+      "learning_rate": 4.8055020150692136e-05,
+      "loss": 1.6777,
+      "step": 5932
+    },
+    {
+      "epoch": 1.5591683909744802,
+      "grad_norm": 0.5205301642417908,
+      "learning_rate": 4.803749780970738e-05,
+      "loss": 1.6599,
+      "step": 5934
+    },
+    {
+      "epoch": 1.559693894308142,
+      "grad_norm": 0.5878111124038696,
+      "learning_rate": 4.8019975468722625e-05,
+      "loss": 1.6491,
+      "step": 5936
+    },
+    {
+      "epoch": 1.560219397641804,
+      "grad_norm": 0.6508564949035645,
+      "learning_rate": 4.8002453127737865e-05,
+      "loss": 1.7085,
+      "step": 5938
+    },
+    {
+      "epoch": 1.5607449009754655,
+      "grad_norm": 0.7473523020744324,
+      "learning_rate": 4.798493078675311e-05,
+      "loss": 1.6765,
+      "step": 5940
+    },
+    {
+      "epoch": 1.5612704043091274,
+      "grad_norm": 0.5754930377006531,
+      "learning_rate": 4.796740844576836e-05,
+      "loss": 1.6825,
+      "step": 5942
+    },
+    {
+      "epoch": 1.561795907642789,
+      "grad_norm": 0.5699328780174255,
+      "learning_rate": 4.79498861047836e-05,
+      "loss": 1.6851,
+      "step": 5944
+    },
+    {
+      "epoch": 1.562321410976451,
+      "grad_norm": 0.5697868466377258,
+      "learning_rate": 4.793236376379884e-05,
+      "loss": 1.644,
+      "step": 5946
+    },
+    {
+      "epoch": 1.5628469143101127,
+      "grad_norm": 0.5635419487953186,
+      "learning_rate": 4.791484142281409e-05,
+      "loss": 1.6765,
+      "step": 5948
+    },
+    {
+      "epoch": 1.5633724176437744,
+      "grad_norm": 0.7371823191642761,
+      "learning_rate": 4.789731908182933e-05,
+      "loss": 1.6876,
+      "step": 5950
+    },
+    {
+      "epoch": 1.5638979209774362,
+      "grad_norm": 0.4995562434196472,
+      "learning_rate": 4.787979674084458e-05,
+      "loss": 1.6593,
+      "step": 5952
+    },
+    {
+      "epoch": 1.564423424311098,
+      "grad_norm": 0.5490265488624573,
+      "learning_rate": 4.7862274399859826e-05,
+      "loss": 1.6424,
+      "step": 5954
+    },
+    {
+      "epoch": 1.56494892764476,
+      "grad_norm": 0.5627312064170837,
+      "learning_rate": 4.7844752058875066e-05,
+      "loss": 1.6849,
+      "step": 5956
+    },
+    {
+      "epoch": 1.5654744309784214,
+      "grad_norm": 0.5825894474983215,
+      "learning_rate": 4.7827229717890314e-05,
+      "loss": 1.6714,
+      "step": 5958
+    },
+    {
+      "epoch": 1.5659999343120834,
+      "grad_norm": 0.6863036751747131,
+      "learning_rate": 4.7809707376905555e-05,
+      "loss": 1.69,
+      "step": 5960
+    },
+    {
+      "epoch": 1.566525437645745,
+      "grad_norm": 0.6272795796394348,
+      "learning_rate": 4.77921850359208e-05,
+      "loss": 1.6536,
+      "step": 5962
+    },
+    {
+      "epoch": 1.567050940979407,
+      "grad_norm": 0.6180011630058289,
+      "learning_rate": 4.777466269493604e-05,
+      "loss": 1.6975,
+      "step": 5964
+    },
+    {
+      "epoch": 1.5675764443130686,
+      "grad_norm": 0.5767374038696289,
+      "learning_rate": 4.775714035395129e-05,
+      "loss": 1.6504,
+      "step": 5966
+    },
+    {
+      "epoch": 1.5681019476467304,
+      "grad_norm": 0.5190562009811401,
+      "learning_rate": 4.773961801296654e-05,
+      "loss": 1.6131,
+      "step": 5968
+    },
+    {
+      "epoch": 1.5686274509803921,
+      "grad_norm": 0.6157041788101196,
+      "learning_rate": 4.772209567198178e-05,
+      "loss": 1.6916,
+      "step": 5970
+    },
+    {
+      "epoch": 1.5691529543140539,
+      "grad_norm": 0.5237650275230408,
+      "learning_rate": 4.770457333099702e-05,
+      "loss": 1.6746,
+      "step": 5972
+    },
+    {
+      "epoch": 1.5696784576477159,
+      "grad_norm": 0.5459091663360596,
+      "learning_rate": 4.768705099001227e-05,
+      "loss": 1.6643,
+      "step": 5974
+    },
+    {
+      "epoch": 1.5702039609813774,
+      "grad_norm": 0.5838450193405151,
+      "learning_rate": 4.7669528649027515e-05,
+      "loss": 1.6745,
+      "step": 5976
+    },
+    {
+      "epoch": 1.5707294643150393,
+      "grad_norm": 0.5081990361213684,
+      "learning_rate": 4.7652006308042756e-05,
+      "loss": 1.6876,
+      "step": 5978
+    },
+    {
+      "epoch": 1.5712549676487009,
+      "grad_norm": 0.6588307619094849,
+      "learning_rate": 4.7634483967058e-05,
+      "loss": 1.7017,
+      "step": 5980
+    },
+    {
+      "epoch": 1.5717804709823628,
+      "grad_norm": 0.8295912146568298,
+      "learning_rate": 4.761696162607325e-05,
+      "loss": 1.6709,
+      "step": 5982
+    },
+    {
+      "epoch": 1.5723059743160246,
+      "grad_norm": 0.6136393547058105,
+      "learning_rate": 4.7599439285088485e-05,
+      "loss": 1.6989,
+      "step": 5984
+    },
+    {
+      "epoch": 1.5728314776496863,
+      "grad_norm": 0.5983635187149048,
+      "learning_rate": 4.758191694410373e-05,
+      "loss": 1.6786,
+      "step": 5986
+    },
+    {
+      "epoch": 1.573356980983348,
+      "grad_norm": 0.6584506630897522,
+      "learning_rate": 4.756439460311898e-05,
+      "loss": 1.6549,
+      "step": 5988
+    },
+    {
+      "epoch": 1.5738824843170098,
+      "grad_norm": 0.513239860534668,
+      "learning_rate": 4.754687226213422e-05,
+      "loss": 1.683,
+      "step": 5990
+    },
+    {
+      "epoch": 1.5744079876506718,
+      "grad_norm": 0.5512750744819641,
+      "learning_rate": 4.752934992114947e-05,
+      "loss": 1.703,
+      "step": 5992
+    },
+    {
+      "epoch": 1.5749334909843333,
+      "grad_norm": 0.7697680592536926,
+      "learning_rate": 4.7511827580164716e-05,
+      "loss": 1.6361,
+      "step": 5994
+    },
+    {
+      "epoch": 1.5754589943179953,
+      "grad_norm": 0.5221796035766602,
+      "learning_rate": 4.749430523917996e-05,
+      "loss": 1.6525,
+      "step": 5996
+    },
+    {
+      "epoch": 1.5759844976516568,
+      "grad_norm": 0.5528659820556641,
+      "learning_rate": 4.74767828981952e-05,
+      "loss": 1.6904,
+      "step": 5998
+    },
+    {
+      "epoch": 1.5765100009853188,
+      "grad_norm": 0.6376633644104004,
+      "learning_rate": 4.7459260557210445e-05,
+      "loss": 1.6849,
+      "step": 6000
+    },
+    {
+      "epoch": 1.5765100009853188,
+      "eval_loss": 1.6817389726638794,
+      "eval_runtime": 487.2585,
+      "eval_samples_per_second": 249.947,
+      "eval_steps_per_second": 31.244,
+      "step": 6000
+    },
+    {
+      "epoch": 1.5770355043189805,
+      "grad_norm": 0.5638622641563416,
+      "learning_rate": 4.744173821622569e-05,
+      "loss": 1.6566,
+      "step": 6002
+    },
+    {
+      "epoch": 1.5775610076526423,
+      "grad_norm": 0.6327787637710571,
+      "learning_rate": 4.7424215875240933e-05,
+      "loss": 1.6806,
+      "step": 6004
+    },
+    {
+      "epoch": 1.578086510986304,
+      "grad_norm": 0.6774953007698059,
+      "learning_rate": 4.740669353425618e-05,
+      "loss": 1.6806,
+      "step": 6006
+    },
+    {
+      "epoch": 1.5786120143199658,
+      "grad_norm": 0.6806586980819702,
+      "learning_rate": 4.738917119327143e-05,
+      "loss": 1.6435,
+      "step": 6008
+    },
+    {
+      "epoch": 1.5791375176536278,
+      "grad_norm": 0.5648463368415833,
+      "learning_rate": 4.737164885228666e-05,
+      "loss": 1.7014,
+      "step": 6010
+    },
+    {
+      "epoch": 1.5796630209872893,
+      "grad_norm": 0.5950681567192078,
+      "learning_rate": 4.735412651130191e-05,
+      "loss": 1.6408,
+      "step": 6012
+    },
+    {
+      "epoch": 1.5801885243209512,
+      "grad_norm": 0.6674718856811523,
+      "learning_rate": 4.733660417031716e-05,
+      "loss": 1.6451,
+      "step": 6014
+    },
+    {
+      "epoch": 1.5807140276546128,
+      "grad_norm": 0.8182973265647888,
+      "learning_rate": 4.73190818293324e-05,
+      "loss": 1.6913,
+      "step": 6016
+    },
+    {
+      "epoch": 1.5812395309882747,
+      "grad_norm": 0.5322865843772888,
+      "learning_rate": 4.7301559488347646e-05,
+      "loss": 1.7063,
+      "step": 6018
+    },
+    {
+      "epoch": 1.5817650343219365,
+      "grad_norm": 0.5284510850906372,
+      "learning_rate": 4.7284037147362894e-05,
+      "loss": 1.6556,
+      "step": 6020
+    },
+    {
+      "epoch": 1.5822905376555982,
+      "grad_norm": 0.6167967915534973,
+      "learning_rate": 4.7266514806378134e-05,
+      "loss": 1.6455,
+      "step": 6022
+    },
+    {
+      "epoch": 1.58281604098926,
+      "grad_norm": 0.5635978579521179,
+      "learning_rate": 4.7248992465393375e-05,
+      "loss": 1.6768,
+      "step": 6024
+    },
+    {
+      "epoch": 1.5833415443229217,
+      "grad_norm": 0.6442639231681824,
+      "learning_rate": 4.723147012440862e-05,
+      "loss": 1.6707,
+      "step": 6026
+    },
+    {
+      "epoch": 1.5838670476565837,
+      "grad_norm": 0.5623191595077515,
+      "learning_rate": 4.721394778342387e-05,
+      "loss": 1.711,
+      "step": 6028
+    },
+    {
+      "epoch": 1.5843925509902452,
+      "grad_norm": 0.5593612194061279,
+      "learning_rate": 4.719642544243911e-05,
+      "loss": 1.6713,
+      "step": 6030
+    },
+    {
+      "epoch": 1.5849180543239072,
+      "grad_norm": 0.7203925251960754,
+      "learning_rate": 4.717890310145436e-05,
+      "loss": 1.6842,
+      "step": 6032
+    },
+    {
+      "epoch": 1.5854435576575687,
+      "grad_norm": 0.5466020703315735,
+      "learning_rate": 4.7161380760469606e-05,
+      "loss": 1.6921,
+      "step": 6034
+    },
+    {
+      "epoch": 1.5859690609912307,
+      "grad_norm": 0.5434938669204712,
+      "learning_rate": 4.714385841948484e-05,
+      "loss": 1.6605,
+      "step": 6036
+    },
+    {
+      "epoch": 1.5864945643248924,
+      "grad_norm": 0.5552278757095337,
+      "learning_rate": 4.712633607850009e-05,
+      "loss": 1.707,
+      "step": 6038
+    },
+    {
+      "epoch": 1.5870200676585542,
+      "grad_norm": 0.6083989143371582,
+      "learning_rate": 4.7108813737515335e-05,
+      "loss": 1.6888,
+      "step": 6040
+    },
+    {
+      "epoch": 1.587545570992216,
+      "grad_norm": 0.5356237292289734,
+      "learning_rate": 4.7091291396530576e-05,
+      "loss": 1.6478,
+      "step": 6042
+    },
+    {
+      "epoch": 1.5880710743258777,
+      "grad_norm": 0.4878697097301483,
+      "learning_rate": 4.7073769055545824e-05,
+      "loss": 1.6558,
+      "step": 6044
+    },
+    {
+      "epoch": 1.5885965776595397,
+      "grad_norm": 0.5412904620170593,
+      "learning_rate": 4.705624671456107e-05,
+      "loss": 1.6763,
+      "step": 6046
+    },
+    {
+      "epoch": 1.5891220809932012,
+      "grad_norm": 0.5351582765579224,
+      "learning_rate": 4.703872437357631e-05,
+      "loss": 1.6891,
+      "step": 6048
+    },
+    {
+      "epoch": 1.5896475843268632,
+      "grad_norm": 0.5699527859687805,
+      "learning_rate": 4.702120203259155e-05,
+      "loss": 1.6841,
+      "step": 6050
+    },
+    {
+      "epoch": 1.590173087660525,
+      "grad_norm": 0.5373657941818237,
+      "learning_rate": 4.70036796916068e-05,
+      "loss": 1.6589,
+      "step": 6052
+    },
+    {
+      "epoch": 1.5906985909941866,
+      "grad_norm": 0.5621985197067261,
+      "learning_rate": 4.698615735062205e-05,
+      "loss": 1.6845,
+      "step": 6054
+    },
+    {
+      "epoch": 1.5912240943278484,
+      "grad_norm": 0.5501397848129272,
+      "learning_rate": 4.696863500963729e-05,
+      "loss": 1.6752,
+      "step": 6056
+    },
+    {
+      "epoch": 1.5917495976615101,
+      "grad_norm": 0.6588435173034668,
+      "learning_rate": 4.6951112668652536e-05,
+      "loss": 1.6727,
+      "step": 6058
+    },
+    {
+      "epoch": 1.5922751009951719,
+      "grad_norm": 0.5301753282546997,
+      "learning_rate": 4.6933590327667784e-05,
+      "loss": 1.6811,
+      "step": 6060
+    },
+    {
+      "epoch": 1.5928006043288336,
+      "grad_norm": 0.5546782612800598,
+      "learning_rate": 4.691606798668302e-05,
+      "loss": 1.6298,
+      "step": 6062
+    },
+    {
+      "epoch": 1.5933261076624956,
+      "grad_norm": 0.6335771679878235,
+      "learning_rate": 4.6898545645698265e-05,
+      "loss": 1.6606,
+      "step": 6064
+    },
+    {
+      "epoch": 1.5938516109961571,
+      "grad_norm": 0.5431744456291199,
+      "learning_rate": 4.688102330471351e-05,
+      "loss": 1.6444,
+      "step": 6066
+    },
+    {
+      "epoch": 1.594377114329819,
+      "grad_norm": 0.6626846790313721,
+      "learning_rate": 4.6863500963728754e-05,
+      "loss": 1.6857,
+      "step": 6068
+    },
+    {
+      "epoch": 1.5949026176634808,
+      "grad_norm": 0.5748207569122314,
+      "learning_rate": 4.6845978622744e-05,
+      "loss": 1.6811,
+      "step": 6070
+    },
+    {
+      "epoch": 1.5954281209971426,
+      "grad_norm": 0.6388061046600342,
+      "learning_rate": 4.682845628175925e-05,
+      "loss": 1.6696,
+      "step": 6072
+    },
+    {
+      "epoch": 1.5959536243308043,
+      "grad_norm": 0.5904421210289001,
+      "learning_rate": 4.681093394077449e-05,
+      "loss": 1.6872,
+      "step": 6074
+    },
+    {
+      "epoch": 1.596479127664466,
+      "grad_norm": 0.5612810254096985,
+      "learning_rate": 4.679341159978973e-05,
+      "loss": 1.7049,
+      "step": 6076
+    },
+    {
+      "epoch": 1.597004630998128,
+      "grad_norm": 0.5887944102287292,
+      "learning_rate": 4.677588925880498e-05,
+      "loss": 1.6556,
+      "step": 6078
+    },
+    {
+      "epoch": 1.5975301343317896,
+      "grad_norm": 0.5331344604492188,
+      "learning_rate": 4.6758366917820226e-05,
+      "loss": 1.6706,
+      "step": 6080
+    },
+    {
+      "epoch": 1.5980556376654516,
+      "grad_norm": 0.6311489939689636,
+      "learning_rate": 4.6740844576835466e-05,
+      "loss": 1.6893,
+      "step": 6082
+    },
+    {
+      "epoch": 1.598581140999113,
+      "grad_norm": 0.5310975313186646,
+      "learning_rate": 4.6723322235850714e-05,
+      "loss": 1.6057,
+      "step": 6084
+    },
+    {
+      "epoch": 1.599106644332775,
+      "grad_norm": 0.5268464684486389,
+      "learning_rate": 4.6705799894865955e-05,
+      "loss": 1.6649,
+      "step": 6086
+    },
+    {
+      "epoch": 1.5996321476664368,
+      "grad_norm": 0.5862019658088684,
+      "learning_rate": 4.66882775538812e-05,
+      "loss": 1.6631,
+      "step": 6088
+    },
+    {
+      "epoch": 1.6001576510000985,
+      "grad_norm": 0.5458804368972778,
+      "learning_rate": 4.667075521289644e-05,
+      "loss": 1.671,
+      "step": 6090
+    },
+    {
+      "epoch": 1.6006831543337603,
+      "grad_norm": 0.5199334025382996,
+      "learning_rate": 4.665323287191169e-05,
+      "loss": 1.6498,
+      "step": 6092
+    },
+    {
+      "epoch": 1.601208657667422,
+      "grad_norm": 0.5879623293876648,
+      "learning_rate": 4.663571053092694e-05,
+      "loss": 1.667,
+      "step": 6094
+    },
+    {
+      "epoch": 1.601734161001084,
+      "grad_norm": 0.6247827410697937,
+      "learning_rate": 4.661818818994218e-05,
+      "loss": 1.7004,
+      "step": 6096
+    },
+    {
+      "epoch": 1.6022596643347455,
+      "grad_norm": 0.6340193152427673,
+      "learning_rate": 4.6600665848957427e-05,
+      "loss": 1.6622,
+      "step": 6098
+    },
+    {
+      "epoch": 1.6027851676684075,
+      "grad_norm": 0.6243396401405334,
+      "learning_rate": 4.658314350797267e-05,
+      "loss": 1.6983,
+      "step": 6100
+    },
+    {
+      "epoch": 1.603310671002069,
+      "grad_norm": 0.584221601486206,
+      "learning_rate": 4.656562116698791e-05,
+      "loss": 1.6548,
+      "step": 6102
+    },
+    {
+      "epoch": 1.603836174335731,
+      "grad_norm": 0.5587270259857178,
+      "learning_rate": 4.6548098826003156e-05,
+      "loss": 1.6482,
+      "step": 6104
+    },
+    {
+      "epoch": 1.6043616776693927,
+      "grad_norm": 0.8525761365890503,
+      "learning_rate": 4.65305764850184e-05,
+      "loss": 1.6601,
+      "step": 6106
+    },
+    {
+      "epoch": 1.6048871810030545,
+      "grad_norm": 0.5316641330718994,
+      "learning_rate": 4.6513054144033644e-05,
+      "loss": 1.6809,
+      "step": 6108
+    },
+    {
+      "epoch": 1.6054126843367162,
+      "grad_norm": 0.5327313542366028,
+      "learning_rate": 4.649553180304889e-05,
+      "loss": 1.6303,
+      "step": 6110
+    },
+    {
+      "epoch": 1.605938187670378,
+      "grad_norm": 0.5327521562576294,
+      "learning_rate": 4.647800946206413e-05,
+      "loss": 1.6875,
+      "step": 6112
+    },
+    {
+      "epoch": 1.60646369100404,
+      "grad_norm": 0.7889205813407898,
+      "learning_rate": 4.646048712107938e-05,
+      "loss": 1.7043,
+      "step": 6114
+    },
+    {
+      "epoch": 1.6069891943377015,
+      "grad_norm": 0.6641364097595215,
+      "learning_rate": 4.644296478009462e-05,
+      "loss": 1.7062,
+      "step": 6116
+    },
+    {
+      "epoch": 1.6075146976713635,
+      "grad_norm": 0.675348162651062,
+      "learning_rate": 4.642544243910987e-05,
+      "loss": 1.6911,
+      "step": 6118
+    },
+    {
+      "epoch": 1.608040201005025,
+      "grad_norm": 0.541476845741272,
+      "learning_rate": 4.6407920098125116e-05,
+      "loss": 1.6667,
+      "step": 6120
+    },
+    {
+      "epoch": 1.608565704338687,
+      "grad_norm": 0.5840083360671997,
+      "learning_rate": 4.639039775714036e-05,
+      "loss": 1.652,
+      "step": 6122
+    },
+    {
+      "epoch": 1.6090912076723487,
+      "grad_norm": 0.5409373641014099,
+      "learning_rate": 4.6372875416155604e-05,
+      "loss": 1.7031,
+      "step": 6124
+    },
+    {
+      "epoch": 1.6096167110060104,
+      "grad_norm": 0.5161097049713135,
+      "learning_rate": 4.6355353075170845e-05,
+      "loss": 1.6873,
+      "step": 6126
+    },
+    {
+      "epoch": 1.6101422143396722,
+      "grad_norm": 0.6245579719543457,
+      "learning_rate": 4.6337830734186086e-05,
+      "loss": 1.6652,
+      "step": 6128
+    },
+    {
+      "epoch": 1.610667717673334,
+      "grad_norm": 0.60563725233078,
+      "learning_rate": 4.6320308393201333e-05,
+      "loss": 1.6783,
+      "step": 6130
+    },
+    {
+      "epoch": 1.611193221006996,
+      "grad_norm": 0.667809009552002,
+      "learning_rate": 4.630278605221658e-05,
+      "loss": 1.6492,
+      "step": 6132
+    },
+    {
+      "epoch": 1.6117187243406574,
+      "grad_norm": 0.5152156352996826,
+      "learning_rate": 4.628526371123182e-05,
+      "loss": 1.6757,
+      "step": 6134
+    },
+    {
+      "epoch": 1.6122442276743194,
+      "grad_norm": 0.5245918035507202,
+      "learning_rate": 4.626774137024707e-05,
+      "loss": 1.6368,
+      "step": 6136
+    },
+    {
+      "epoch": 1.612769731007981,
+      "grad_norm": 0.5406076908111572,
+      "learning_rate": 4.625021902926231e-05,
+      "loss": 1.6817,
+      "step": 6138
+    },
+    {
+      "epoch": 1.613295234341643,
+      "grad_norm": 0.6298016905784607,
+      "learning_rate": 4.623269668827756e-05,
+      "loss": 1.6542,
+      "step": 6140
+    },
+    {
+      "epoch": 1.6138207376753047,
+      "grad_norm": 0.5466469526290894,
+      "learning_rate": 4.62151743472928e-05,
+      "loss": 1.6367,
+      "step": 6142
+    },
+    {
+      "epoch": 1.6143462410089664,
+      "grad_norm": 0.7066838145256042,
+      "learning_rate": 4.6197652006308046e-05,
+      "loss": 1.6461,
+      "step": 6144
+    },
+    {
+      "epoch": 1.6148717443426281,
+      "grad_norm": 0.5102798938751221,
+      "learning_rate": 4.6180129665323294e-05,
+      "loss": 1.6628,
+      "step": 6146
+    },
+    {
+      "epoch": 1.61539724767629,
+      "grad_norm": 0.72835773229599,
+      "learning_rate": 4.6162607324338534e-05,
+      "loss": 1.7141,
+      "step": 6148
+    },
+    {
+      "epoch": 1.6159227510099519,
+      "grad_norm": 0.5149089097976685,
+      "learning_rate": 4.6145084983353775e-05,
+      "loss": 1.6539,
+      "step": 6150
+    },
+    {
+      "epoch": 1.6164482543436134,
+      "grad_norm": 0.5791059732437134,
+      "learning_rate": 4.612756264236902e-05,
+      "loss": 1.6791,
+      "step": 6152
+    },
+    {
+      "epoch": 1.6169737576772754,
+      "grad_norm": 0.5340200662612915,
+      "learning_rate": 4.6110040301384263e-05,
+      "loss": 1.6807,
+      "step": 6154
+    },
+    {
+      "epoch": 1.6174992610109369,
+      "grad_norm": 0.5943130254745483,
+      "learning_rate": 4.609251796039951e-05,
+      "loss": 1.662,
+      "step": 6156
+    },
+    {
+      "epoch": 1.6180247643445989,
+      "grad_norm": 0.5829207897186279,
+      "learning_rate": 4.607499561941476e-05,
+      "loss": 1.67,
+      "step": 6158
+    },
+    {
+      "epoch": 1.6185502676782606,
+      "grad_norm": 0.5549664497375488,
+      "learning_rate": 4.605747327843e-05,
+      "loss": 1.6686,
+      "step": 6160
+    },
+    {
+      "epoch": 1.6190757710119223,
+      "grad_norm": 0.6359058618545532,
+      "learning_rate": 4.603995093744525e-05,
+      "loss": 1.6694,
+      "step": 6162
+    },
+    {
+      "epoch": 1.619601274345584,
+      "grad_norm": 0.5652121305465698,
+      "learning_rate": 4.602242859646049e-05,
+      "loss": 1.6678,
+      "step": 6164
+    },
+    {
+      "epoch": 1.6201267776792458,
+      "grad_norm": 0.5643856525421143,
+      "learning_rate": 4.6004906255475735e-05,
+      "loss": 1.6915,
+      "step": 6166
+    },
+    {
+      "epoch": 1.6206522810129078,
+      "grad_norm": 0.5165106058120728,
+      "learning_rate": 4.5987383914490976e-05,
+      "loss": 1.6643,
+      "step": 6168
+    },
+    {
+      "epoch": 1.6211777843465693,
+      "grad_norm": 0.6091976761817932,
+      "learning_rate": 4.5969861573506224e-05,
+      "loss": 1.6878,
+      "step": 6170
+    },
+    {
+      "epoch": 1.6217032876802313,
+      "grad_norm": 0.5565605759620667,
+      "learning_rate": 4.595233923252147e-05,
+      "loss": 1.6843,
+      "step": 6172
+    },
+    {
+      "epoch": 1.6222287910138928,
+      "grad_norm": 0.635827362537384,
+      "learning_rate": 4.593481689153671e-05,
+      "loss": 1.6554,
+      "step": 6174
+    },
+    {
+      "epoch": 1.6227542943475548,
+      "grad_norm": 0.6421988010406494,
+      "learning_rate": 4.591729455055195e-05,
+      "loss": 1.6298,
+      "step": 6176
+    },
+    {
+      "epoch": 1.6232797976812166,
+      "grad_norm": 0.7460888624191284,
+      "learning_rate": 4.58997722095672e-05,
+      "loss": 1.7074,
+      "step": 6178
+    },
+    {
+      "epoch": 1.6238053010148783,
+      "grad_norm": 0.5302199721336365,
+      "learning_rate": 4.588224986858244e-05,
+      "loss": 1.6546,
+      "step": 6180
+    },
+    {
+      "epoch": 1.62433080434854,
+      "grad_norm": 0.5736710429191589,
+      "learning_rate": 4.586472752759769e-05,
+      "loss": 1.6816,
+      "step": 6182
+    },
+    {
+      "epoch": 1.6248563076822018,
+      "grad_norm": 0.6789550185203552,
+      "learning_rate": 4.5847205186612936e-05,
+      "loss": 1.6643,
+      "step": 6184
+    },
+    {
+      "epoch": 1.6253818110158638,
+      "grad_norm": 0.5486941337585449,
+      "learning_rate": 4.582968284562818e-05,
+      "loss": 1.68,
+      "step": 6186
+    },
+    {
+      "epoch": 1.6259073143495253,
+      "grad_norm": 0.8008251786231995,
+      "learning_rate": 4.581216050464342e-05,
+      "loss": 1.6386,
+      "step": 6188
+    },
+    {
+      "epoch": 1.6264328176831873,
+      "grad_norm": 0.6284978985786438,
+      "learning_rate": 4.5794638163658665e-05,
+      "loss": 1.6497,
+      "step": 6190
+    },
+    {
+      "epoch": 1.6269583210168488,
+      "grad_norm": 0.5751000642776489,
+      "learning_rate": 4.577711582267391e-05,
+      "loss": 1.6897,
+      "step": 6192
+    },
+    {
+      "epoch": 1.6274838243505108,
+      "grad_norm": 0.5420372486114502,
+      "learning_rate": 4.5759593481689154e-05,
+      "loss": 1.6936,
+      "step": 6194
+    },
+    {
+      "epoch": 1.6280093276841725,
+      "grad_norm": 0.5770832896232605,
+      "learning_rate": 4.57420711407044e-05,
+      "loss": 1.6682,
+      "step": 6196
+    },
+    {
+      "epoch": 1.6285348310178343,
+      "grad_norm": 0.5784463286399841,
+      "learning_rate": 4.572454879971965e-05,
+      "loss": 1.6884,
+      "step": 6198
+    },
+    {
+      "epoch": 1.629060334351496,
+      "grad_norm": 0.5765862464904785,
+      "learning_rate": 4.570702645873489e-05,
+      "loss": 1.6228,
+      "step": 6200
+    },
+    {
+      "epoch": 1.6295858376851577,
+      "grad_norm": 0.6257441639900208,
+      "learning_rate": 4.568950411775013e-05,
+      "loss": 1.6559,
+      "step": 6202
+    },
+    {
+      "epoch": 1.6301113410188197,
+      "grad_norm": 0.5480515360832214,
+      "learning_rate": 4.567198177676538e-05,
+      "loss": 1.6524,
+      "step": 6204
+    },
+    {
+      "epoch": 1.6306368443524812,
+      "grad_norm": 0.5308948755264282,
+      "learning_rate": 4.565445943578062e-05,
+      "loss": 1.6954,
+      "step": 6206
+    },
+    {
+      "epoch": 1.6311623476861432,
+      "grad_norm": 0.640443742275238,
+      "learning_rate": 4.5636937094795866e-05,
+      "loss": 1.6434,
+      "step": 6208
+    },
+    {
+      "epoch": 1.631687851019805,
+      "grad_norm": 0.5399090647697449,
+      "learning_rate": 4.5619414753811114e-05,
+      "loss": 1.6317,
+      "step": 6210
+    },
+    {
+      "epoch": 1.6322133543534667,
+      "grad_norm": 0.5155448913574219,
+      "learning_rate": 4.5601892412826355e-05,
+      "loss": 1.6258,
+      "step": 6212
+    },
+    {
+      "epoch": 1.6327388576871285,
+      "grad_norm": 0.540034830570221,
+      "learning_rate": 4.5584370071841596e-05,
+      "loss": 1.6592,
+      "step": 6214
+    },
+    {
+      "epoch": 1.6332643610207902,
+      "grad_norm": 0.6191601157188416,
+      "learning_rate": 4.556684773085684e-05,
+      "loss": 1.6188,
+      "step": 6216
+    },
+    {
+      "epoch": 1.633789864354452,
+      "grad_norm": 0.5206524133682251,
+      "learning_rate": 4.554932538987209e-05,
+      "loss": 1.667,
+      "step": 6218
+    },
+    {
+      "epoch": 1.6343153676881137,
+      "grad_norm": 0.5997523069381714,
+      "learning_rate": 4.553180304888733e-05,
+      "loss": 1.7165,
+      "step": 6220
+    },
+    {
+      "epoch": 1.6348408710217757,
+      "grad_norm": 0.6088078618049622,
+      "learning_rate": 4.551428070790258e-05,
+      "loss": 1.65,
+      "step": 6222
+    },
+    {
+      "epoch": 1.6353663743554372,
+      "grad_norm": 0.5121122598648071,
+      "learning_rate": 4.5496758366917827e-05,
+      "loss": 1.6661,
+      "step": 6224
+    },
+    {
+      "epoch": 1.6358918776890992,
+      "grad_norm": 0.5405161380767822,
+      "learning_rate": 4.547923602593307e-05,
+      "loss": 1.6812,
+      "step": 6226
+    },
+    {
+      "epoch": 1.636417381022761,
+      "grad_norm": 0.5181125402450562,
+      "learning_rate": 4.546171368494831e-05,
+      "loss": 1.6888,
+      "step": 6228
+    },
+    {
+      "epoch": 1.6369428843564227,
+      "grad_norm": 0.7036988735198975,
+      "learning_rate": 4.5444191343963556e-05,
+      "loss": 1.7048,
+      "step": 6230
+    },
+    {
+      "epoch": 1.6374683876900844,
+      "grad_norm": 0.559990644454956,
+      "learning_rate": 4.54266690029788e-05,
+      "loss": 1.6745,
+      "step": 6232
+    },
+    {
+      "epoch": 1.6379938910237462,
+      "grad_norm": 0.5458931922912598,
+      "learning_rate": 4.5409146661994044e-05,
+      "loss": 1.6668,
+      "step": 6234
+    },
+    {
+      "epoch": 1.6385193943574081,
+      "grad_norm": 0.6637448072433472,
+      "learning_rate": 4.539162432100929e-05,
+      "loss": 1.7194,
+      "step": 6236
+    },
+    {
+      "epoch": 1.6390448976910696,
+      "grad_norm": 0.5551019310951233,
+      "learning_rate": 4.537410198002454e-05,
+      "loss": 1.6846,
+      "step": 6238
+    },
+    {
+      "epoch": 1.6395704010247316,
+      "grad_norm": 0.6413047909736633,
+      "learning_rate": 4.535657963903977e-05,
+      "loss": 1.6505,
+      "step": 6240
+    },
+    {
+      "epoch": 1.6400959043583931,
+      "grad_norm": 0.5083743929862976,
+      "learning_rate": 4.533905729805502e-05,
+      "loss": 1.6929,
+      "step": 6242
+    },
+    {
+      "epoch": 1.6406214076920551,
+      "grad_norm": 0.608717679977417,
+      "learning_rate": 4.532153495707027e-05,
+      "loss": 1.6767,
+      "step": 6244
+    },
+    {
+      "epoch": 1.6411469110257169,
+      "grad_norm": 0.5961493253707886,
+      "learning_rate": 4.530401261608551e-05,
+      "loss": 1.6434,
+      "step": 6246
+    },
+    {
+      "epoch": 1.6416724143593786,
+      "grad_norm": 0.5599090456962585,
+      "learning_rate": 4.528649027510076e-05,
+      "loss": 1.6774,
+      "step": 6248
+    },
+    {
+      "epoch": 1.6421979176930404,
+      "grad_norm": 0.5643408298492432,
+      "learning_rate": 4.5268967934116004e-05,
+      "loss": 1.6631,
+      "step": 6250
+    },
+    {
+      "epoch": 1.642723421026702,
+      "grad_norm": 0.5354277491569519,
+      "learning_rate": 4.5251445593131245e-05,
+      "loss": 1.6685,
+      "step": 6252
+    },
+    {
+      "epoch": 1.643248924360364,
+      "grad_norm": 0.6432374119758606,
+      "learning_rate": 4.5233923252146486e-05,
+      "loss": 1.6746,
+      "step": 6254
+    },
+    {
+      "epoch": 1.6437744276940256,
+      "grad_norm": 0.5566967725753784,
+      "learning_rate": 4.521640091116173e-05,
+      "loss": 1.6512,
+      "step": 6256
+    },
+    {
+      "epoch": 1.6442999310276876,
+      "grad_norm": 0.517951488494873,
+      "learning_rate": 4.519887857017698e-05,
+      "loss": 1.6485,
+      "step": 6258
+    },
+    {
+      "epoch": 1.644825434361349,
+      "grad_norm": 0.5770967602729797,
+      "learning_rate": 4.518135622919222e-05,
+      "loss": 1.6786,
+      "step": 6260
+    },
+    {
+      "epoch": 1.645350937695011,
+      "grad_norm": 0.5681661367416382,
+      "learning_rate": 4.516383388820747e-05,
+      "loss": 1.6464,
+      "step": 6262
+    },
+    {
+      "epoch": 1.6458764410286728,
+      "grad_norm": 0.6962856650352478,
+      "learning_rate": 4.514631154722272e-05,
+      "loss": 1.6435,
+      "step": 6264
+    },
+    {
+      "epoch": 1.6464019443623346,
+      "grad_norm": 0.5683899521827698,
+      "learning_rate": 4.512878920623795e-05,
+      "loss": 1.6862,
+      "step": 6266
+    },
+    {
+      "epoch": 1.6469274476959963,
+      "grad_norm": 0.6042145490646362,
+      "learning_rate": 4.51112668652532e-05,
+      "loss": 1.6574,
+      "step": 6268
+    },
+    {
+      "epoch": 1.647452951029658,
+      "grad_norm": 0.5737462639808655,
+      "learning_rate": 4.5093744524268446e-05,
+      "loss": 1.6785,
+      "step": 6270
+    },
+    {
+      "epoch": 1.64797845436332,
+      "grad_norm": 0.6028643846511841,
+      "learning_rate": 4.507622218328369e-05,
+      "loss": 1.6568,
+      "step": 6272
+    },
+    {
+      "epoch": 1.6485039576969815,
+      "grad_norm": 0.6065294742584229,
+      "learning_rate": 4.5058699842298934e-05,
+      "loss": 1.6882,
+      "step": 6274
+    },
+    {
+      "epoch": 1.6490294610306435,
+      "grad_norm": 0.5672723054885864,
+      "learning_rate": 4.504117750131418e-05,
+      "loss": 1.619,
+      "step": 6276
+    },
+    {
+      "epoch": 1.649554964364305,
+      "grad_norm": 0.727794885635376,
+      "learning_rate": 4.502365516032942e-05,
+      "loss": 1.6583,
+      "step": 6278
+    },
+    {
+      "epoch": 1.650080467697967,
+      "grad_norm": 0.6044321060180664,
+      "learning_rate": 4.5006132819344663e-05,
+      "loss": 1.6795,
+      "step": 6280
+    },
+    {
+      "epoch": 1.6506059710316288,
+      "grad_norm": 0.535193145275116,
+      "learning_rate": 4.498861047835991e-05,
+      "loss": 1.6772,
+      "step": 6282
+    },
+    {
+      "epoch": 1.6511314743652905,
+      "grad_norm": 0.651655375957489,
+      "learning_rate": 4.497108813737516e-05,
+      "loss": 1.6993,
+      "step": 6284
+    },
+    {
+      "epoch": 1.6516569776989523,
+      "grad_norm": 0.6530981659889221,
+      "learning_rate": 4.49535657963904e-05,
+      "loss": 1.6768,
+      "step": 6286
+    },
+    {
+      "epoch": 1.652182481032614,
+      "grad_norm": 0.6580101251602173,
+      "learning_rate": 4.493604345540565e-05,
+      "loss": 1.695,
+      "step": 6288
+    },
+    {
+      "epoch": 1.652707984366276,
+      "grad_norm": 0.6548781991004944,
+      "learning_rate": 4.491852111442089e-05,
+      "loss": 1.682,
+      "step": 6290
+    },
+    {
+      "epoch": 1.6532334876999375,
+      "grad_norm": 0.5895609259605408,
+      "learning_rate": 4.490099877343613e-05,
+      "loss": 1.703,
+      "step": 6292
+    },
+    {
+      "epoch": 1.6537589910335995,
+      "grad_norm": 0.5485368371009827,
+      "learning_rate": 4.4883476432451376e-05,
+      "loss": 1.686,
+      "step": 6294
+    },
+    {
+      "epoch": 1.654284494367261,
+      "grad_norm": 0.6156832575798035,
+      "learning_rate": 4.4865954091466624e-05,
+      "loss": 1.68,
+      "step": 6296
+    },
+    {
+      "epoch": 1.654809997700923,
+      "grad_norm": 0.5617910027503967,
+      "learning_rate": 4.4848431750481864e-05,
+      "loss": 1.6528,
+      "step": 6298
+    },
+    {
+      "epoch": 1.6553355010345847,
+      "grad_norm": 0.568001925945282,
+      "learning_rate": 4.483090940949711e-05,
+      "loss": 1.6656,
+      "step": 6300
+    },
+    {
+      "epoch": 1.6558610043682465,
+      "grad_norm": 0.5492426156997681,
+      "learning_rate": 4.481338706851236e-05,
+      "loss": 1.669,
+      "step": 6302
+    },
+    {
+      "epoch": 1.6563865077019082,
+      "grad_norm": 0.5638085007667542,
+      "learning_rate": 4.47958647275276e-05,
+      "loss": 1.7048,
+      "step": 6304
+    },
+    {
+      "epoch": 1.65691201103557,
+      "grad_norm": 0.5807180404663086,
+      "learning_rate": 4.477834238654284e-05,
+      "loss": 1.6495,
+      "step": 6306
+    },
+    {
+      "epoch": 1.657437514369232,
+      "grad_norm": 0.6146125197410583,
+      "learning_rate": 4.476082004555809e-05,
+      "loss": 1.7158,
+      "step": 6308
+    },
+    {
+      "epoch": 1.6579630177028934,
+      "grad_norm": 0.5622976422309875,
+      "learning_rate": 4.4743297704573336e-05,
+      "loss": 1.6907,
+      "step": 6310
+    },
+    {
+      "epoch": 1.6584885210365554,
+      "grad_norm": 0.49633076786994934,
+      "learning_rate": 4.472577536358858e-05,
+      "loss": 1.6695,
+      "step": 6312
+    },
+    {
+      "epoch": 1.659014024370217,
+      "grad_norm": 0.5975070595741272,
+      "learning_rate": 4.4708253022603825e-05,
+      "loss": 1.7011,
+      "step": 6314
+    },
+    {
+      "epoch": 1.659539527703879,
+      "grad_norm": 0.6197476983070374,
+      "learning_rate": 4.4690730681619065e-05,
+      "loss": 1.6522,
+      "step": 6316
+    },
+    {
+      "epoch": 1.6600650310375407,
+      "grad_norm": 0.6124839782714844,
+      "learning_rate": 4.4673208340634306e-05,
+      "loss": 1.7181,
+      "step": 6318
+    },
+    {
+      "epoch": 1.6605905343712024,
+      "grad_norm": 0.5166561007499695,
+      "learning_rate": 4.4655685999649554e-05,
+      "loss": 1.6536,
+      "step": 6320
+    },
+    {
+      "epoch": 1.6611160377048642,
+      "grad_norm": 0.5568446516990662,
+      "learning_rate": 4.46381636586648e-05,
+      "loss": 1.682,
+      "step": 6322
+    },
+    {
+      "epoch": 1.661641541038526,
+      "grad_norm": 0.5822721719741821,
+      "learning_rate": 4.462064131768004e-05,
+      "loss": 1.6657,
+      "step": 6324
+    },
+    {
+      "epoch": 1.6621670443721879,
+      "grad_norm": 0.6438559293746948,
+      "learning_rate": 4.460311897669529e-05,
+      "loss": 1.6795,
+      "step": 6326
+    },
+    {
+      "epoch": 1.6626925477058494,
+      "grad_norm": 0.7065990567207336,
+      "learning_rate": 4.458559663571053e-05,
+      "loss": 1.6618,
+      "step": 6328
+    },
+    {
+      "epoch": 1.6632180510395114,
+      "grad_norm": 0.512935996055603,
+      "learning_rate": 4.456807429472578e-05,
+      "loss": 1.6742,
+      "step": 6330
+    },
+    {
+      "epoch": 1.663743554373173,
+      "grad_norm": 0.7639873027801514,
+      "learning_rate": 4.455055195374102e-05,
+      "loss": 1.6593,
+      "step": 6332
+    },
+    {
+      "epoch": 1.6642690577068349,
+      "grad_norm": 0.5977439880371094,
+      "learning_rate": 4.4533029612756266e-05,
+      "loss": 1.6691,
+      "step": 6334
+    },
+    {
+      "epoch": 1.6647945610404966,
+      "grad_norm": 0.5824545621871948,
+      "learning_rate": 4.4515507271771514e-05,
+      "loss": 1.6571,
+      "step": 6336
+    },
+    {
+      "epoch": 1.6653200643741584,
+      "grad_norm": 0.5170328617095947,
+      "learning_rate": 4.4497984930786755e-05,
+      "loss": 1.676,
+      "step": 6338
+    },
+    {
+      "epoch": 1.66584556770782,
+      "grad_norm": 0.6465065479278564,
+      "learning_rate": 4.4480462589802e-05,
+      "loss": 1.6516,
+      "step": 6340
+    },
+    {
+      "epoch": 1.6663710710414819,
+      "grad_norm": 0.639741837978363,
+      "learning_rate": 4.446294024881724e-05,
+      "loss": 1.6696,
+      "step": 6342
+    },
+    {
+      "epoch": 1.6668965743751438,
+      "grad_norm": 0.5170881152153015,
+      "learning_rate": 4.444541790783249e-05,
+      "loss": 1.6353,
+      "step": 6344
+    },
+    {
+      "epoch": 1.6674220777088054,
+      "grad_norm": 0.5460655689239502,
+      "learning_rate": 4.442789556684773e-05,
+      "loss": 1.6869,
+      "step": 6346
+    },
+    {
+      "epoch": 1.6679475810424673,
+      "grad_norm": 0.5527986288070679,
+      "learning_rate": 4.441037322586298e-05,
+      "loss": 1.6261,
+      "step": 6348
+    },
+    {
+      "epoch": 1.6684730843761288,
+      "grad_norm": 0.5400204062461853,
+      "learning_rate": 4.4392850884878227e-05,
+      "loss": 1.6559,
+      "step": 6350
+    },
+    {
+      "epoch": 1.6689985877097908,
+      "grad_norm": 0.5666835904121399,
+      "learning_rate": 4.437532854389347e-05,
+      "loss": 1.7025,
+      "step": 6352
+    },
+    {
+      "epoch": 1.6695240910434526,
+      "grad_norm": 0.5993382930755615,
+      "learning_rate": 4.435780620290871e-05,
+      "loss": 1.676,
+      "step": 6354
+    },
+    {
+      "epoch": 1.6700495943771143,
+      "grad_norm": 0.5372394919395447,
+      "learning_rate": 4.4340283861923956e-05,
+      "loss": 1.6549,
+      "step": 6356
+    },
+    {
+      "epoch": 1.670575097710776,
+      "grad_norm": 0.6082696914672852,
+      "learning_rate": 4.4322761520939196e-05,
+      "loss": 1.6643,
+      "step": 6358
+    },
+    {
+      "epoch": 1.6711006010444378,
+      "grad_norm": 0.6554064154624939,
+      "learning_rate": 4.4305239179954444e-05,
+      "loss": 1.692,
+      "step": 6360
+    },
+    {
+      "epoch": 1.6716261043780998,
+      "grad_norm": 0.4961806833744049,
+      "learning_rate": 4.428771683896969e-05,
+      "loss": 1.6921,
+      "step": 6362
+    },
+    {
+      "epoch": 1.6721516077117613,
+      "grad_norm": 0.5501666069030762,
+      "learning_rate": 4.427019449798493e-05,
+      "loss": 1.678,
+      "step": 6364
+    },
+    {
+      "epoch": 1.6726771110454233,
+      "grad_norm": 0.5967716574668884,
+      "learning_rate": 4.425267215700018e-05,
+      "loss": 1.6548,
+      "step": 6366
+    },
+    {
+      "epoch": 1.673202614379085,
+      "grad_norm": 0.654115617275238,
+      "learning_rate": 4.423514981601542e-05,
+      "loss": 1.6859,
+      "step": 6368
+    },
+    {
+      "epoch": 1.6737281177127468,
+      "grad_norm": 0.5566443204879761,
+      "learning_rate": 4.421762747503067e-05,
+      "loss": 1.653,
+      "step": 6370
+    },
+    {
+      "epoch": 1.6742536210464085,
+      "grad_norm": 0.6134412884712219,
+      "learning_rate": 4.420010513404591e-05,
+      "loss": 1.6876,
+      "step": 6372
+    },
+    {
+      "epoch": 1.6747791243800703,
+      "grad_norm": 0.7075713276863098,
+      "learning_rate": 4.418258279306116e-05,
+      "loss": 1.6738,
+      "step": 6374
+    },
+    {
+      "epoch": 1.675304627713732,
+      "grad_norm": 0.5378595590591431,
+      "learning_rate": 4.4165060452076404e-05,
+      "loss": 1.6696,
+      "step": 6376
+    },
+    {
+      "epoch": 1.6758301310473938,
+      "grad_norm": 0.5113415718078613,
+      "learning_rate": 4.4147538111091645e-05,
+      "loss": 1.6662,
+      "step": 6378
+    },
+    {
+      "epoch": 1.6763556343810557,
+      "grad_norm": 0.6252776384353638,
+      "learning_rate": 4.4130015770106886e-05,
+      "loss": 1.6517,
+      "step": 6380
+    },
+    {
+      "epoch": 1.6768811377147173,
+      "grad_norm": 0.5247318744659424,
+      "learning_rate": 4.411249342912213e-05,
+      "loss": 1.6769,
+      "step": 6382
+    },
+    {
+      "epoch": 1.6774066410483792,
+      "grad_norm": 0.5468015670776367,
+      "learning_rate": 4.4094971088137374e-05,
+      "loss": 1.6827,
+      "step": 6384
+    },
+    {
+      "epoch": 1.677932144382041,
+      "grad_norm": 0.5248594880104065,
+      "learning_rate": 4.407744874715262e-05,
+      "loss": 1.691,
+      "step": 6386
+    },
+    {
+      "epoch": 1.6784576477157027,
+      "grad_norm": 0.5515632629394531,
+      "learning_rate": 4.405992640616787e-05,
+      "loss": 1.664,
+      "step": 6388
+    },
+    {
+      "epoch": 1.6789831510493645,
+      "grad_norm": 0.565548300743103,
+      "learning_rate": 4.404240406518311e-05,
+      "loss": 1.6677,
+      "step": 6390
+    },
+    {
+      "epoch": 1.6795086543830262,
+      "grad_norm": 0.5626314878463745,
+      "learning_rate": 4.402488172419835e-05,
+      "loss": 1.6778,
+      "step": 6392
+    },
+    {
+      "epoch": 1.6800341577166882,
+      "grad_norm": 0.5197250247001648,
+      "learning_rate": 4.40073593832136e-05,
+      "loss": 1.6824,
+      "step": 6394
+    },
+    {
+      "epoch": 1.6805596610503497,
+      "grad_norm": 0.5577792525291443,
+      "learning_rate": 4.3989837042228846e-05,
+      "loss": 1.6798,
+      "step": 6396
+    },
+    {
+      "epoch": 1.6810851643840117,
+      "grad_norm": 0.5743740200996399,
+      "learning_rate": 4.397231470124409e-05,
+      "loss": 1.6328,
+      "step": 6398
+    },
+    {
+      "epoch": 1.6816106677176732,
+      "grad_norm": 0.4996255934238434,
+      "learning_rate": 4.3954792360259334e-05,
+      "loss": 1.6781,
+      "step": 6400
+    },
+    {
+      "epoch": 1.6816106677176732,
+      "eval_loss": 1.6763501167297363,
+      "eval_runtime": 487.3304,
+      "eval_samples_per_second": 249.911,
+      "eval_steps_per_second": 31.24,
+      "step": 6400
+    },
+    {
+      "epoch": 1.6821361710513352,
+      "grad_norm": 0.6712773442268372,
+      "learning_rate": 4.393727001927458e-05,
+      "loss": 1.679,
+      "step": 6402
+    },
+    {
+      "epoch": 1.682661674384997,
+      "grad_norm": 0.5399149656295776,
+      "learning_rate": 4.391974767828982e-05,
+      "loss": 1.7156,
+      "step": 6404
+    },
+    {
+      "epoch": 1.6831871777186587,
+      "grad_norm": 0.5838908553123474,
+      "learning_rate": 4.3902225337305063e-05,
+      "loss": 1.6559,
+      "step": 6406
+    },
+    {
+      "epoch": 1.6837126810523204,
+      "grad_norm": 0.6989631056785583,
+      "learning_rate": 4.388470299632031e-05,
+      "loss": 1.6536,
+      "step": 6408
+    },
+    {
+      "epoch": 1.6842381843859822,
+      "grad_norm": 0.6087121963500977,
+      "learning_rate": 4.386718065533555e-05,
+      "loss": 1.68,
+      "step": 6410
+    },
+    {
+      "epoch": 1.6847636877196441,
+      "grad_norm": 0.595737099647522,
+      "learning_rate": 4.38496583143508e-05,
+      "loss": 1.6431,
+      "step": 6412
+    },
+    {
+      "epoch": 1.6852891910533057,
+      "grad_norm": 0.5878545045852661,
+      "learning_rate": 4.383213597336605e-05,
+      "loss": 1.6524,
+      "step": 6414
+    },
+    {
+      "epoch": 1.6858146943869676,
+      "grad_norm": 0.5520877242088318,
+      "learning_rate": 4.381461363238129e-05,
+      "loss": 1.6234,
+      "step": 6416
+    },
+    {
+      "epoch": 1.6863401977206292,
+      "grad_norm": 0.530946671962738,
+      "learning_rate": 4.379709129139653e-05,
+      "loss": 1.6478,
+      "step": 6418
+    },
+    {
+      "epoch": 1.6868657010542911,
+      "grad_norm": 0.5743231177330017,
+      "learning_rate": 4.3779568950411776e-05,
+      "loss": 1.6882,
+      "step": 6420
+    },
+    {
+      "epoch": 1.6873912043879529,
+      "grad_norm": 0.5853824019432068,
+      "learning_rate": 4.3762046609427024e-05,
+      "loss": 1.6798,
+      "step": 6422
+    },
+    {
+      "epoch": 1.6879167077216146,
+      "grad_norm": 0.5864454507827759,
+      "learning_rate": 4.3744524268442264e-05,
+      "loss": 1.656,
+      "step": 6424
+    },
+    {
+      "epoch": 1.6884422110552764,
+      "grad_norm": 0.5097535848617554,
+      "learning_rate": 4.372700192745751e-05,
+      "loss": 1.6534,
+      "step": 6426
+    },
+    {
+      "epoch": 1.6889677143889381,
+      "grad_norm": 0.5435791015625,
+      "learning_rate": 4.370947958647276e-05,
+      "loss": 1.6782,
+      "step": 6428
+    },
+    {
+      "epoch": 1.6894932177226,
+      "grad_norm": 0.6465846300125122,
+      "learning_rate": 4.3691957245487994e-05,
+      "loss": 1.6816,
+      "step": 6430
+    },
+    {
+      "epoch": 1.6900187210562616,
+      "grad_norm": 0.561132550239563,
+      "learning_rate": 4.367443490450324e-05,
+      "loss": 1.6838,
+      "step": 6432
+    },
+    {
+      "epoch": 1.6905442243899236,
+      "grad_norm": 0.530135452747345,
+      "learning_rate": 4.365691256351849e-05,
+      "loss": 1.6541,
+      "step": 6434
+    },
+    {
+      "epoch": 1.691069727723585,
+      "grad_norm": 0.6319286823272705,
+      "learning_rate": 4.363939022253373e-05,
+      "loss": 1.651,
+      "step": 6436
+    },
+    {
+      "epoch": 1.691595231057247,
+      "grad_norm": 0.4940394461154938,
+      "learning_rate": 4.362186788154898e-05,
+      "loss": 1.6613,
+      "step": 6438
+    },
+    {
+      "epoch": 1.6921207343909088,
+      "grad_norm": 0.4995363652706146,
+      "learning_rate": 4.3604345540564225e-05,
+      "loss": 1.6572,
+      "step": 6440
+    },
+    {
+      "epoch": 1.6926462377245706,
+      "grad_norm": 0.5799241662025452,
+      "learning_rate": 4.3586823199579465e-05,
+      "loss": 1.6753,
+      "step": 6442
+    },
+    {
+      "epoch": 1.6931717410582323,
+      "grad_norm": 0.5875564813613892,
+      "learning_rate": 4.3569300858594706e-05,
+      "loss": 1.6322,
+      "step": 6444
+    },
+    {
+      "epoch": 1.693697244391894,
+      "grad_norm": 0.5701809525489807,
+      "learning_rate": 4.3551778517609954e-05,
+      "loss": 1.7094,
+      "step": 6446
+    },
+    {
+      "epoch": 1.694222747725556,
+      "grad_norm": 0.576756477355957,
+      "learning_rate": 4.35342561766252e-05,
+      "loss": 1.667,
+      "step": 6448
+    },
+    {
+      "epoch": 1.6947482510592176,
+      "grad_norm": 0.5512332320213318,
+      "learning_rate": 4.351673383564044e-05,
+      "loss": 1.6592,
+      "step": 6450
+    },
+    {
+      "epoch": 1.6952737543928795,
+      "grad_norm": 0.5287933349609375,
+      "learning_rate": 4.349921149465569e-05,
+      "loss": 1.6677,
+      "step": 6452
+    },
+    {
+      "epoch": 1.695799257726541,
+      "grad_norm": 0.623125433921814,
+      "learning_rate": 4.348168915367094e-05,
+      "loss": 1.6633,
+      "step": 6454
+    },
+    {
+      "epoch": 1.696324761060203,
+      "grad_norm": 0.613586962223053,
+      "learning_rate": 4.346416681268617e-05,
+      "loss": 1.66,
+      "step": 6456
+    },
+    {
+      "epoch": 1.6968502643938648,
+      "grad_norm": 0.5631827116012573,
+      "learning_rate": 4.344664447170142e-05,
+      "loss": 1.6701,
+      "step": 6458
+    },
+    {
+      "epoch": 1.6973757677275265,
+      "grad_norm": 0.4900628328323364,
+      "learning_rate": 4.3429122130716666e-05,
+      "loss": 1.6603,
+      "step": 6460
+    },
+    {
+      "epoch": 1.6979012710611883,
+      "grad_norm": 0.5031628608703613,
+      "learning_rate": 4.341159978973191e-05,
+      "loss": 1.6554,
+      "step": 6462
+    },
+    {
+      "epoch": 1.69842677439485,
+      "grad_norm": 0.5051286816596985,
+      "learning_rate": 4.3394077448747155e-05,
+      "loss": 1.6324,
+      "step": 6464
+    },
+    {
+      "epoch": 1.698952277728512,
+      "grad_norm": 0.5495415925979614,
+      "learning_rate": 4.33765551077624e-05,
+      "loss": 1.679,
+      "step": 6466
+    },
+    {
+      "epoch": 1.6994777810621735,
+      "grad_norm": 0.7252418398857117,
+      "learning_rate": 4.335903276677764e-05,
+      "loss": 1.6898,
+      "step": 6468
+    },
+    {
+      "epoch": 1.7000032843958355,
+      "grad_norm": 0.5228211879730225,
+      "learning_rate": 4.3341510425792884e-05,
+      "loss": 1.644,
+      "step": 6470
+    },
+    {
+      "epoch": 1.700528787729497,
+      "grad_norm": 0.5848027467727661,
+      "learning_rate": 4.332398808480813e-05,
+      "loss": 1.656,
+      "step": 6472
+    },
+    {
+      "epoch": 1.701054291063159,
+      "grad_norm": 0.6442865133285522,
+      "learning_rate": 4.330646574382338e-05,
+      "loss": 1.6261,
+      "step": 6474
+    },
+    {
+      "epoch": 1.7015797943968207,
+      "grad_norm": 0.5885564684867859,
+      "learning_rate": 4.328894340283862e-05,
+      "loss": 1.6406,
+      "step": 6476
+    },
+    {
+      "epoch": 1.7021052977304825,
+      "grad_norm": 0.5726144909858704,
+      "learning_rate": 4.327142106185387e-05,
+      "loss": 1.6843,
+      "step": 6478
+    },
+    {
+      "epoch": 1.7026308010641442,
+      "grad_norm": 0.6747820377349854,
+      "learning_rate": 4.3253898720869115e-05,
+      "loss": 1.6851,
+      "step": 6480
+    },
+    {
+      "epoch": 1.703156304397806,
+      "grad_norm": 0.5923687815666199,
+      "learning_rate": 4.3236376379884356e-05,
+      "loss": 1.6318,
+      "step": 6482
+    },
+    {
+      "epoch": 1.703681807731468,
+      "grad_norm": 0.5728587508201599,
+      "learning_rate": 4.3218854038899596e-05,
+      "loss": 1.682,
+      "step": 6484
+    },
+    {
+      "epoch": 1.7042073110651295,
+      "grad_norm": 0.5209431648254395,
+      "learning_rate": 4.3201331697914844e-05,
+      "loss": 1.662,
+      "step": 6486
+    },
+    {
+      "epoch": 1.7047328143987914,
+      "grad_norm": 0.5561770796775818,
+      "learning_rate": 4.318380935693009e-05,
+      "loss": 1.7016,
+      "step": 6488
+    },
+    {
+      "epoch": 1.705258317732453,
+      "grad_norm": 0.5122435688972473,
+      "learning_rate": 4.316628701594533e-05,
+      "loss": 1.6682,
+      "step": 6490
+    },
+    {
+      "epoch": 1.705783821066115,
+      "grad_norm": 0.6389492750167847,
+      "learning_rate": 4.314876467496058e-05,
+      "loss": 1.6379,
+      "step": 6492
+    },
+    {
+      "epoch": 1.7063093243997767,
+      "grad_norm": 0.5349413752555847,
+      "learning_rate": 4.313124233397582e-05,
+      "loss": 1.6343,
+      "step": 6494
+    },
+    {
+      "epoch": 1.7068348277334384,
+      "grad_norm": 0.5826820731163025,
+      "learning_rate": 4.311371999299106e-05,
+      "loss": 1.6775,
+      "step": 6496
+    },
+    {
+      "epoch": 1.7073603310671002,
+      "grad_norm": 0.5609052181243896,
+      "learning_rate": 4.309619765200631e-05,
+      "loss": 1.7076,
+      "step": 6498
+    },
+    {
+      "epoch": 1.707885834400762,
+      "grad_norm": 0.5599504113197327,
+      "learning_rate": 4.307867531102156e-05,
+      "loss": 1.6997,
+      "step": 6500
+    },
+    {
+      "epoch": 1.7084113377344239,
+      "grad_norm": 0.5563739538192749,
+      "learning_rate": 4.30611529700368e-05,
+      "loss": 1.6595,
+      "step": 6502
+    },
+    {
+      "epoch": 1.7089368410680854,
+      "grad_norm": 0.5203465819358826,
+      "learning_rate": 4.3043630629052045e-05,
+      "loss": 1.6452,
+      "step": 6504
+    },
+    {
+      "epoch": 1.7094623444017474,
+      "grad_norm": 0.59616619348526,
+      "learning_rate": 4.302610828806729e-05,
+      "loss": 1.6495,
+      "step": 6506
+    },
+    {
+      "epoch": 1.709987847735409,
+      "grad_norm": 0.7246098518371582,
+      "learning_rate": 4.300858594708253e-05,
+      "loss": 1.6682,
+      "step": 6508
+    },
+    {
+      "epoch": 1.7105133510690709,
+      "grad_norm": 0.5792231559753418,
+      "learning_rate": 4.2991063606097774e-05,
+      "loss": 1.6474,
+      "step": 6510
+    },
+    {
+      "epoch": 1.7110388544027326,
+      "grad_norm": 0.5333255529403687,
+      "learning_rate": 4.297354126511302e-05,
+      "loss": 1.6795,
+      "step": 6512
+    },
+    {
+      "epoch": 1.7115643577363944,
+      "grad_norm": 0.674374520778656,
+      "learning_rate": 4.295601892412827e-05,
+      "loss": 1.6514,
+      "step": 6514
+    },
+    {
+      "epoch": 1.7120898610700561,
+      "grad_norm": 0.5450239181518555,
+      "learning_rate": 4.293849658314351e-05,
+      "loss": 1.6606,
+      "step": 6516
+    },
+    {
+      "epoch": 1.7126153644037179,
+      "grad_norm": 0.6590548157691956,
+      "learning_rate": 4.292097424215876e-05,
+      "loss": 1.6337,
+      "step": 6518
+    },
+    {
+      "epoch": 1.7131408677373798,
+      "grad_norm": 0.5265931487083435,
+      "learning_rate": 4.2903451901174e-05,
+      "loss": 1.6842,
+      "step": 6520
+    },
+    {
+      "epoch": 1.7136663710710414,
+      "grad_norm": 0.6718656420707703,
+      "learning_rate": 4.288592956018924e-05,
+      "loss": 1.696,
+      "step": 6522
+    },
+    {
+      "epoch": 1.7141918744047033,
+      "grad_norm": 0.6330803632736206,
+      "learning_rate": 4.286840721920449e-05,
+      "loss": 1.671,
+      "step": 6524
+    },
+    {
+      "epoch": 1.714717377738365,
+      "grad_norm": 0.4948212206363678,
+      "learning_rate": 4.2850884878219734e-05,
+      "loss": 1.6464,
+      "step": 6526
+    },
+    {
+      "epoch": 1.7152428810720268,
+      "grad_norm": 0.5330238342285156,
+      "learning_rate": 4.2833362537234975e-05,
+      "loss": 1.6661,
+      "step": 6528
+    },
+    {
+      "epoch": 1.7157683844056886,
+      "grad_norm": 0.5928429961204529,
+      "learning_rate": 4.281584019625022e-05,
+      "loss": 1.645,
+      "step": 6530
+    },
+    {
+      "epoch": 1.7162938877393503,
+      "grad_norm": 0.535369336605072,
+      "learning_rate": 4.2798317855265463e-05,
+      "loss": 1.6554,
+      "step": 6532
+    },
+    {
+      "epoch": 1.716819391073012,
+      "grad_norm": 0.5079066157341003,
+      "learning_rate": 4.278079551428071e-05,
+      "loss": 1.6498,
+      "step": 6534
+    },
+    {
+      "epoch": 1.7173448944066738,
+      "grad_norm": 0.5394106507301331,
+      "learning_rate": 4.276327317329595e-05,
+      "loss": 1.6538,
+      "step": 6536
+    },
+    {
+      "epoch": 1.7178703977403358,
+      "grad_norm": 0.5476702451705933,
+      "learning_rate": 4.27457508323112e-05,
+      "loss": 1.6568,
+      "step": 6538
+    },
+    {
+      "epoch": 1.7183959010739973,
+      "grad_norm": 0.6342707276344299,
+      "learning_rate": 4.272822849132645e-05,
+      "loss": 1.6829,
+      "step": 6540
+    },
+    {
+      "epoch": 1.7189214044076593,
+      "grad_norm": 0.6379374265670776,
+      "learning_rate": 4.271070615034169e-05,
+      "loss": 1.6742,
+      "step": 6542
+    },
+    {
+      "epoch": 1.719446907741321,
+      "grad_norm": 0.6817846894264221,
+      "learning_rate": 4.2693183809356935e-05,
+      "loss": 1.6354,
+      "step": 6544
+    },
+    {
+      "epoch": 1.7199724110749828,
+      "grad_norm": 0.6458591222763062,
+      "learning_rate": 4.2675661468372176e-05,
+      "loss": 1.6599,
+      "step": 6546
+    },
+    {
+      "epoch": 1.7204979144086445,
+      "grad_norm": 0.5006933808326721,
+      "learning_rate": 4.265813912738742e-05,
+      "loss": 1.6438,
+      "step": 6548
+    },
+    {
+      "epoch": 1.7210234177423063,
+      "grad_norm": 0.4796706736087799,
+      "learning_rate": 4.2640616786402664e-05,
+      "loss": 1.6738,
+      "step": 6550
+    },
+    {
+      "epoch": 1.7215489210759682,
+      "grad_norm": 0.5186893939971924,
+      "learning_rate": 4.262309444541791e-05,
+      "loss": 1.691,
+      "step": 6552
+    },
+    {
+      "epoch": 1.7220744244096298,
+      "grad_norm": 0.5311626195907593,
+      "learning_rate": 4.260557210443315e-05,
+      "loss": 1.6739,
+      "step": 6554
+    },
+    {
+      "epoch": 1.7225999277432917,
+      "grad_norm": 0.5143429040908813,
+      "learning_rate": 4.25880497634484e-05,
+      "loss": 1.6533,
+      "step": 6556
+    },
+    {
+      "epoch": 1.7231254310769533,
+      "grad_norm": 0.5430511236190796,
+      "learning_rate": 4.257052742246364e-05,
+      "loss": 1.6758,
+      "step": 6558
+    },
+    {
+      "epoch": 1.7236509344106152,
+      "grad_norm": 0.6149283647537231,
+      "learning_rate": 4.255300508147889e-05,
+      "loss": 1.6581,
+      "step": 6560
+    },
+    {
+      "epoch": 1.724176437744277,
+      "grad_norm": 0.5292539000511169,
+      "learning_rate": 4.253548274049413e-05,
+      "loss": 1.6494,
+      "step": 6562
+    },
+    {
+      "epoch": 1.7247019410779387,
+      "grad_norm": 0.615308403968811,
+      "learning_rate": 4.251796039950938e-05,
+      "loss": 1.6732,
+      "step": 6564
+    },
+    {
+      "epoch": 1.7252274444116005,
+      "grad_norm": 0.562659740447998,
+      "learning_rate": 4.2500438058524625e-05,
+      "loss": 1.6881,
+      "step": 6566
+    },
+    {
+      "epoch": 1.7257529477452622,
+      "grad_norm": 0.6096563339233398,
+      "learning_rate": 4.2482915717539865e-05,
+      "loss": 1.6344,
+      "step": 6568
+    },
+    {
+      "epoch": 1.7262784510789242,
+      "grad_norm": 0.568242073059082,
+      "learning_rate": 4.2465393376555106e-05,
+      "loss": 1.6506,
+      "step": 6570
+    },
+    {
+      "epoch": 1.7268039544125857,
+      "grad_norm": 0.6250702738761902,
+      "learning_rate": 4.2447871035570354e-05,
+      "loss": 1.6647,
+      "step": 6572
+    },
+    {
+      "epoch": 1.7273294577462477,
+      "grad_norm": 0.6344661116600037,
+      "learning_rate": 4.2430348694585594e-05,
+      "loss": 1.6733,
+      "step": 6574
+    },
+    {
+      "epoch": 1.7278549610799092,
+      "grad_norm": 0.5727905631065369,
+      "learning_rate": 4.241282635360084e-05,
+      "loss": 1.6757,
+      "step": 6576
+    },
+    {
+      "epoch": 1.7283804644135712,
+      "grad_norm": 0.5363614559173584,
+      "learning_rate": 4.239530401261609e-05,
+      "loss": 1.6418,
+      "step": 6578
+    },
+    {
+      "epoch": 1.728905967747233,
+      "grad_norm": 0.5695384740829468,
+      "learning_rate": 4.237778167163133e-05,
+      "loss": 1.6424,
+      "step": 6580
+    },
+    {
+      "epoch": 1.7294314710808947,
+      "grad_norm": 0.6201584935188293,
+      "learning_rate": 4.236025933064658e-05,
+      "loss": 1.6648,
+      "step": 6582
+    },
+    {
+      "epoch": 1.7299569744145564,
+      "grad_norm": 0.4974352717399597,
+      "learning_rate": 4.234273698966182e-05,
+      "loss": 1.6666,
+      "step": 6584
+    },
+    {
+      "epoch": 1.7304824777482182,
+      "grad_norm": 0.5813178420066833,
+      "learning_rate": 4.2325214648677066e-05,
+      "loss": 1.6307,
+      "step": 6586
+    },
+    {
+      "epoch": 1.7310079810818801,
+      "grad_norm": 0.5724592804908752,
+      "learning_rate": 4.230769230769231e-05,
+      "loss": 1.6467,
+      "step": 6588
+    },
+    {
+      "epoch": 1.7315334844155417,
+      "grad_norm": 0.5860669612884521,
+      "learning_rate": 4.2290169966707555e-05,
+      "loss": 1.6765,
+      "step": 6590
+    },
+    {
+      "epoch": 1.7320589877492036,
+      "grad_norm": 0.5303966999053955,
+      "learning_rate": 4.22726476257228e-05,
+      "loss": 1.6575,
+      "step": 6592
+    },
+    {
+      "epoch": 1.7325844910828652,
+      "grad_norm": 0.5227830410003662,
+      "learning_rate": 4.225512528473804e-05,
+      "loss": 1.6884,
+      "step": 6594
+    },
+    {
+      "epoch": 1.7331099944165271,
+      "grad_norm": 0.5740933418273926,
+      "learning_rate": 4.2237602943753284e-05,
+      "loss": 1.6472,
+      "step": 6596
+    },
+    {
+      "epoch": 1.7336354977501889,
+      "grad_norm": 0.5894073843955994,
+      "learning_rate": 4.222008060276853e-05,
+      "loss": 1.6755,
+      "step": 6598
+    },
+    {
+      "epoch": 1.7341610010838506,
+      "grad_norm": 0.5687074065208435,
+      "learning_rate": 4.220255826178378e-05,
+      "loss": 1.675,
+      "step": 6600
+    },
+    {
+      "epoch": 1.7346865044175124,
+      "grad_norm": 0.6006156206130981,
+      "learning_rate": 4.218503592079902e-05,
+      "loss": 1.6651,
+      "step": 6602
+    },
+    {
+      "epoch": 1.7352120077511741,
+      "grad_norm": 0.564089834690094,
+      "learning_rate": 4.216751357981427e-05,
+      "loss": 1.6789,
+      "step": 6604
+    },
+    {
+      "epoch": 1.735737511084836,
+      "grad_norm": 0.7533421516418457,
+      "learning_rate": 4.2149991238829515e-05,
+      "loss": 1.6557,
+      "step": 6606
+    },
+    {
+      "epoch": 1.7362630144184976,
+      "grad_norm": 0.5872588753700256,
+      "learning_rate": 4.2132468897844756e-05,
+      "loss": 1.6713,
+      "step": 6608
+    },
+    {
+      "epoch": 1.7367885177521596,
+      "grad_norm": 0.608405590057373,
+      "learning_rate": 4.2114946556859996e-05,
+      "loss": 1.6648,
+      "step": 6610
+    },
+    {
+      "epoch": 1.7373140210858211,
+      "grad_norm": 0.5417534112930298,
+      "learning_rate": 4.2097424215875244e-05,
+      "loss": 1.6667,
+      "step": 6612
+    },
+    {
+      "epoch": 1.737839524419483,
+      "grad_norm": 0.5255427956581116,
+      "learning_rate": 4.2079901874890485e-05,
+      "loss": 1.6847,
+      "step": 6614
+    },
+    {
+      "epoch": 1.7383650277531448,
+      "grad_norm": 0.49976515769958496,
+      "learning_rate": 4.206237953390573e-05,
+      "loss": 1.6426,
+      "step": 6616
+    },
+    {
+      "epoch": 1.7388905310868066,
+      "grad_norm": 0.5245642066001892,
+      "learning_rate": 4.204485719292098e-05,
+      "loss": 1.6557,
+      "step": 6618
+    },
+    {
+      "epoch": 1.7394160344204683,
+      "grad_norm": 0.5113621354103088,
+      "learning_rate": 4.202733485193622e-05,
+      "loss": 1.6725,
+      "step": 6620
+    },
+    {
+      "epoch": 1.73994153775413,
+      "grad_norm": 0.6118736863136292,
+      "learning_rate": 4.200981251095146e-05,
+      "loss": 1.6524,
+      "step": 6622
+    },
+    {
+      "epoch": 1.740467041087792,
+      "grad_norm": 0.6329546570777893,
+      "learning_rate": 4.199229016996671e-05,
+      "loss": 1.7146,
+      "step": 6624
+    },
+    {
+      "epoch": 1.7409925444214536,
+      "grad_norm": 0.5709455013275146,
+      "learning_rate": 4.1974767828981957e-05,
+      "loss": 1.6481,
+      "step": 6626
+    },
+    {
+      "epoch": 1.7415180477551155,
+      "grad_norm": 0.5557751655578613,
+      "learning_rate": 4.19572454879972e-05,
+      "loss": 1.6298,
+      "step": 6628
+    },
+    {
+      "epoch": 1.742043551088777,
+      "grad_norm": 0.5406216979026794,
+      "learning_rate": 4.1939723147012445e-05,
+      "loss": 1.7127,
+      "step": 6630
+    },
+    {
+      "epoch": 1.742569054422439,
+      "grad_norm": 0.6411069631576538,
+      "learning_rate": 4.192220080602769e-05,
+      "loss": 1.674,
+      "step": 6632
+    },
+    {
+      "epoch": 1.7430945577561008,
+      "grad_norm": 0.5912994742393494,
+      "learning_rate": 4.1904678465042927e-05,
+      "loss": 1.7002,
+      "step": 6634
+    },
+    {
+      "epoch": 1.7436200610897625,
+      "grad_norm": 0.5379955172538757,
+      "learning_rate": 4.1887156124058174e-05,
+      "loss": 1.6702,
+      "step": 6636
+    },
+    {
+      "epoch": 1.7441455644234243,
+      "grad_norm": 0.5569443702697754,
+      "learning_rate": 4.186963378307342e-05,
+      "loss": 1.6848,
+      "step": 6638
+    },
+    {
+      "epoch": 1.744671067757086,
+      "grad_norm": 0.5895394086837769,
+      "learning_rate": 4.185211144208866e-05,
+      "loss": 1.6775,
+      "step": 6640
+    },
+    {
+      "epoch": 1.745196571090748,
+      "grad_norm": 0.6112500429153442,
+      "learning_rate": 4.183458910110391e-05,
+      "loss": 1.6713,
+      "step": 6642
+    },
+    {
+      "epoch": 1.7457220744244095,
+      "grad_norm": 0.762654185295105,
+      "learning_rate": 4.181706676011916e-05,
+      "loss": 1.6848,
+      "step": 6644
+    },
+    {
+      "epoch": 1.7462475777580715,
+      "grad_norm": 0.6181445717811584,
+      "learning_rate": 4.17995444191344e-05,
+      "loss": 1.6735,
+      "step": 6646
+    },
+    {
+      "epoch": 1.746773081091733,
+      "grad_norm": 0.5101475119590759,
+      "learning_rate": 4.178202207814964e-05,
+      "loss": 1.6639,
+      "step": 6648
+    },
+    {
+      "epoch": 1.747298584425395,
+      "grad_norm": 0.5235376358032227,
+      "learning_rate": 4.176449973716489e-05,
+      "loss": 1.6485,
+      "step": 6650
+    },
+    {
+      "epoch": 1.7478240877590567,
+      "grad_norm": 0.5604961514472961,
+      "learning_rate": 4.1746977396180134e-05,
+      "loss": 1.6501,
+      "step": 6652
+    },
+    {
+      "epoch": 1.7483495910927185,
+      "grad_norm": 0.759231448173523,
+      "learning_rate": 4.1729455055195375e-05,
+      "loss": 1.6752,
+      "step": 6654
+    },
+    {
+      "epoch": 1.7488750944263802,
+      "grad_norm": 0.5132787823677063,
+      "learning_rate": 4.171193271421062e-05,
+      "loss": 1.672,
+      "step": 6656
+    },
+    {
+      "epoch": 1.749400597760042,
+      "grad_norm": 0.5898250341415405,
+      "learning_rate": 4.169441037322587e-05,
+      "loss": 1.6543,
+      "step": 6658
+    },
+    {
+      "epoch": 1.749926101093704,
+      "grad_norm": 0.5514124631881714,
+      "learning_rate": 4.1676888032241104e-05,
+      "loss": 1.6879,
+      "step": 6660
+    },
+    {
+      "epoch": 1.7504516044273655,
+      "grad_norm": 0.7601991295814514,
+      "learning_rate": 4.165936569125635e-05,
+      "loss": 1.7259,
+      "step": 6662
+    },
+    {
+      "epoch": 1.7509771077610274,
+      "grad_norm": 0.6521760821342468,
+      "learning_rate": 4.16418433502716e-05,
+      "loss": 1.6488,
+      "step": 6664
+    },
+    {
+      "epoch": 1.751502611094689,
+      "grad_norm": 0.6008849143981934,
+      "learning_rate": 4.162432100928684e-05,
+      "loss": 1.6577,
+      "step": 6666
+    },
+    {
+      "epoch": 1.752028114428351,
+      "grad_norm": 0.6038839221000671,
+      "learning_rate": 4.160679866830209e-05,
+      "loss": 1.6503,
+      "step": 6668
+    },
+    {
+      "epoch": 1.7525536177620127,
+      "grad_norm": 0.6234827041625977,
+      "learning_rate": 4.1589276327317335e-05,
+      "loss": 1.6779,
+      "step": 6670
+    },
+    {
+      "epoch": 1.7530791210956744,
+      "grad_norm": 0.5239622592926025,
+      "learning_rate": 4.1571753986332576e-05,
+      "loss": 1.6431,
+      "step": 6672
+    },
+    {
+      "epoch": 1.7536046244293362,
+      "grad_norm": 0.6171594262123108,
+      "learning_rate": 4.155423164534782e-05,
+      "loss": 1.6436,
+      "step": 6674
+    },
+    {
+      "epoch": 1.754130127762998,
+      "grad_norm": 0.651139497756958,
+      "learning_rate": 4.1536709304363064e-05,
+      "loss": 1.6492,
+      "step": 6676
+    },
+    {
+      "epoch": 1.75465563109666,
+      "grad_norm": 0.5682376027107239,
+      "learning_rate": 4.151918696337831e-05,
+      "loss": 1.6417,
+      "step": 6678
+    },
+    {
+      "epoch": 1.7551811344303214,
+      "grad_norm": 0.6295192837715149,
+      "learning_rate": 4.150166462239355e-05,
+      "loss": 1.6839,
+      "step": 6680
+    },
+    {
+      "epoch": 1.7557066377639834,
+      "grad_norm": 0.6096534132957458,
+      "learning_rate": 4.14841422814088e-05,
+      "loss": 1.7125,
+      "step": 6682
+    },
+    {
+      "epoch": 1.7562321410976451,
+      "grad_norm": 0.5466519594192505,
+      "learning_rate": 4.146661994042405e-05,
+      "loss": 1.6566,
+      "step": 6684
+    },
+    {
+      "epoch": 1.7567576444313069,
+      "grad_norm": 0.5740132331848145,
+      "learning_rate": 4.144909759943928e-05,
+      "loss": 1.6564,
+      "step": 6686
+    },
+    {
+      "epoch": 1.7572831477649686,
+      "grad_norm": 0.6351927518844604,
+      "learning_rate": 4.143157525845453e-05,
+      "loss": 1.6752,
+      "step": 6688
+    },
+    {
+      "epoch": 1.7578086510986304,
+      "grad_norm": 0.5739028453826904,
+      "learning_rate": 4.141405291746978e-05,
+      "loss": 1.6834,
+      "step": 6690
+    },
+    {
+      "epoch": 1.7583341544322921,
+      "grad_norm": 0.5718164443969727,
+      "learning_rate": 4.139653057648502e-05,
+      "loss": 1.6433,
+      "step": 6692
+    },
+    {
+      "epoch": 1.7588596577659539,
+      "grad_norm": 0.738676905632019,
+      "learning_rate": 4.1379008235500265e-05,
+      "loss": 1.6528,
+      "step": 6694
+    },
+    {
+      "epoch": 1.7593851610996158,
+      "grad_norm": 0.596734881401062,
+      "learning_rate": 4.136148589451551e-05,
+      "loss": 1.6431,
+      "step": 6696
+    },
+    {
+      "epoch": 1.7599106644332774,
+      "grad_norm": 0.5336854457855225,
+      "learning_rate": 4.1343963553530754e-05,
+      "loss": 1.6692,
+      "step": 6698
+    },
+    {
+      "epoch": 1.7604361677669393,
+      "grad_norm": 0.4876728653907776,
+      "learning_rate": 4.1326441212545994e-05,
+      "loss": 1.6595,
+      "step": 6700
+    },
+    {
+      "epoch": 1.760961671100601,
+      "grad_norm": 0.5300989747047424,
+      "learning_rate": 4.130891887156124e-05,
+      "loss": 1.6767,
+      "step": 6702
+    },
+    {
+      "epoch": 1.7614871744342628,
+      "grad_norm": 0.54608154296875,
+      "learning_rate": 4.129139653057649e-05,
+      "loss": 1.6279,
+      "step": 6704
+    },
+    {
+      "epoch": 1.7620126777679246,
+      "grad_norm": 0.5571487545967102,
+      "learning_rate": 4.127387418959173e-05,
+      "loss": 1.6963,
+      "step": 6706
+    },
+    {
+      "epoch": 1.7625381811015863,
+      "grad_norm": 0.5999481081962585,
+      "learning_rate": 4.125635184860698e-05,
+      "loss": 1.6807,
+      "step": 6708
+    },
+    {
+      "epoch": 1.7630636844352483,
+      "grad_norm": 0.5582924485206604,
+      "learning_rate": 4.123882950762222e-05,
+      "loss": 1.6503,
+      "step": 6710
+    },
+    {
+      "epoch": 1.7635891877689098,
+      "grad_norm": 0.5172569751739502,
+      "learning_rate": 4.122130716663746e-05,
+      "loss": 1.6502,
+      "step": 6712
+    },
+    {
+      "epoch": 1.7641146911025718,
+      "grad_norm": 0.5434536933898926,
+      "learning_rate": 4.120378482565271e-05,
+      "loss": 1.624,
+      "step": 6714
+    },
+    {
+      "epoch": 1.7646401944362333,
+      "grad_norm": 0.5931615233421326,
+      "learning_rate": 4.1186262484667955e-05,
+      "loss": 1.6881,
+      "step": 6716
+    },
+    {
+      "epoch": 1.7651656977698953,
+      "grad_norm": 0.5632887482643127,
+      "learning_rate": 4.1168740143683195e-05,
+      "loss": 1.66,
+      "step": 6718
+    },
+    {
+      "epoch": 1.765691201103557,
+      "grad_norm": 0.6316903233528137,
+      "learning_rate": 4.115121780269844e-05,
+      "loss": 1.6821,
+      "step": 6720
+    },
+    {
+      "epoch": 1.7662167044372188,
+      "grad_norm": 0.5220393538475037,
+      "learning_rate": 4.113369546171369e-05,
+      "loss": 1.6585,
+      "step": 6722
+    },
+    {
+      "epoch": 1.7667422077708805,
+      "grad_norm": 0.5338044166564941,
+      "learning_rate": 4.111617312072893e-05,
+      "loss": 1.6359,
+      "step": 6724
+    },
+    {
+      "epoch": 1.7672677111045423,
+      "grad_norm": 0.5751186013221741,
+      "learning_rate": 4.109865077974417e-05,
+      "loss": 1.6678,
+      "step": 6726
+    },
+    {
+      "epoch": 1.7677932144382043,
+      "grad_norm": 0.5516241788864136,
+      "learning_rate": 4.108112843875942e-05,
+      "loss": 1.6589,
+      "step": 6728
+    },
+    {
+      "epoch": 1.7683187177718658,
+      "grad_norm": 0.5440977811813354,
+      "learning_rate": 4.106360609777467e-05,
+      "loss": 1.6496,
+      "step": 6730
+    },
+    {
+      "epoch": 1.7688442211055277,
+      "grad_norm": 0.5160251259803772,
+      "learning_rate": 4.104608375678991e-05,
+      "loss": 1.6586,
+      "step": 6732
+    },
+    {
+      "epoch": 1.7693697244391893,
+      "grad_norm": 0.6195341348648071,
+      "learning_rate": 4.1028561415805156e-05,
+      "loss": 1.6621,
+      "step": 6734
+    },
+    {
+      "epoch": 1.7698952277728512,
+      "grad_norm": 0.5011487007141113,
+      "learning_rate": 4.1011039074820396e-05,
+      "loss": 1.7067,
+      "step": 6736
+    },
+    {
+      "epoch": 1.770420731106513,
+      "grad_norm": 0.5898102521896362,
+      "learning_rate": 4.0993516733835644e-05,
+      "loss": 1.6775,
+      "step": 6738
+    },
+    {
+      "epoch": 1.7709462344401747,
+      "grad_norm": 0.6446313261985779,
+      "learning_rate": 4.0975994392850885e-05,
+      "loss": 1.6751,
+      "step": 6740
+    },
+    {
+      "epoch": 1.7714717377738365,
+      "grad_norm": 0.5387564301490784,
+      "learning_rate": 4.095847205186613e-05,
+      "loss": 1.6861,
+      "step": 6742
+    },
+    {
+      "epoch": 1.7719972411074982,
+      "grad_norm": 0.6098289489746094,
+      "learning_rate": 4.094094971088138e-05,
+      "loss": 1.6653,
+      "step": 6744
+    },
+    {
+      "epoch": 1.7725227444411602,
+      "grad_norm": 0.5589563846588135,
+      "learning_rate": 4.092342736989662e-05,
+      "loss": 1.6878,
+      "step": 6746
+    },
+    {
+      "epoch": 1.7730482477748217,
+      "grad_norm": 0.6051377058029175,
+      "learning_rate": 4.090590502891187e-05,
+      "loss": 1.6779,
+      "step": 6748
+    },
+    {
+      "epoch": 1.7735737511084837,
+      "grad_norm": 0.5657187104225159,
+      "learning_rate": 4.088838268792711e-05,
+      "loss": 1.6585,
+      "step": 6750
+    },
+    {
+      "epoch": 1.7740992544421452,
+      "grad_norm": 0.9947826862335205,
+      "learning_rate": 4.087086034694235e-05,
+      "loss": 1.6364,
+      "step": 6752
+    },
+    {
+      "epoch": 1.7746247577758072,
+      "grad_norm": 0.5528366565704346,
+      "learning_rate": 4.08533380059576e-05,
+      "loss": 1.6796,
+      "step": 6754
+    },
+    {
+      "epoch": 1.775150261109469,
+      "grad_norm": 0.7105492949485779,
+      "learning_rate": 4.0835815664972845e-05,
+      "loss": 1.6359,
+      "step": 6756
+    },
+    {
+      "epoch": 1.7756757644431307,
+      "grad_norm": 0.5398980975151062,
+      "learning_rate": 4.0818293323988086e-05,
+      "loss": 1.6358,
+      "step": 6758
+    },
+    {
+      "epoch": 1.7762012677767924,
+      "grad_norm": 0.518286943435669,
+      "learning_rate": 4.080077098300333e-05,
+      "loss": 1.6691,
+      "step": 6760
+    },
+    {
+      "epoch": 1.7767267711104542,
+      "grad_norm": 0.5343197584152222,
+      "learning_rate": 4.0783248642018574e-05,
+      "loss": 1.6604,
+      "step": 6762
+    },
+    {
+      "epoch": 1.7772522744441162,
+      "grad_norm": 0.6191185116767883,
+      "learning_rate": 4.076572630103382e-05,
+      "loss": 1.6693,
+      "step": 6764
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.6115537285804749,
+      "learning_rate": 4.074820396004906e-05,
+      "loss": 1.6533,
+      "step": 6766
+    },
+    {
+      "epoch": 1.7783032811114396,
+      "grad_norm": 0.6350643634796143,
+      "learning_rate": 4.073068161906431e-05,
+      "loss": 1.6717,
+      "step": 6768
+    },
+    {
+      "epoch": 1.7788287844451012,
+      "grad_norm": 0.5143194198608398,
+      "learning_rate": 4.071315927807956e-05,
+      "loss": 1.6684,
+      "step": 6770
+    },
+    {
+      "epoch": 1.7793542877787631,
+      "grad_norm": 0.6066332459449768,
+      "learning_rate": 4.06956369370948e-05,
+      "loss": 1.6722,
+      "step": 6772
+    },
+    {
+      "epoch": 1.779879791112425,
+      "grad_norm": 0.5524691939353943,
+      "learning_rate": 4.067811459611004e-05,
+      "loss": 1.6423,
+      "step": 6774
+    },
+    {
+      "epoch": 1.7804052944460866,
+      "grad_norm": 0.6012272834777832,
+      "learning_rate": 4.066059225512529e-05,
+      "loss": 1.6931,
+      "step": 6776
+    },
+    {
+      "epoch": 1.7809307977797484,
+      "grad_norm": 0.58221834897995,
+      "learning_rate": 4.064306991414053e-05,
+      "loss": 1.6591,
+      "step": 6778
+    },
+    {
+      "epoch": 1.7814563011134101,
+      "grad_norm": 0.5327997803688049,
+      "learning_rate": 4.0625547573155775e-05,
+      "loss": 1.6903,
+      "step": 6780
+    },
+    {
+      "epoch": 1.781981804447072,
+      "grad_norm": 0.5887238383293152,
+      "learning_rate": 4.060802523217102e-05,
+      "loss": 1.6571,
+      "step": 6782
+    },
+    {
+      "epoch": 1.7825073077807336,
+      "grad_norm": 0.5583620071411133,
+      "learning_rate": 4.059050289118626e-05,
+      "loss": 1.6552,
+      "step": 6784
+    },
+    {
+      "epoch": 1.7830328111143956,
+      "grad_norm": 0.6321818232536316,
+      "learning_rate": 4.057298055020151e-05,
+      "loss": 1.6565,
+      "step": 6786
+    },
+    {
+      "epoch": 1.7835583144480571,
+      "grad_norm": 0.496971070766449,
+      "learning_rate": 4.055545820921675e-05,
+      "loss": 1.6311,
+      "step": 6788
+    },
+    {
+      "epoch": 1.784083817781719,
+      "grad_norm": 0.5267062783241272,
+      "learning_rate": 4.0537935868232e-05,
+      "loss": 1.6598,
+      "step": 6790
+    },
+    {
+      "epoch": 1.7846093211153808,
+      "grad_norm": 0.502678394317627,
+      "learning_rate": 4.052041352724724e-05,
+      "loss": 1.6493,
+      "step": 6792
+    },
+    {
+      "epoch": 1.7851348244490426,
+      "grad_norm": 0.6034113764762878,
+      "learning_rate": 4.050289118626249e-05,
+      "loss": 1.6914,
+      "step": 6794
+    },
+    {
+      "epoch": 1.7856603277827043,
+      "grad_norm": 0.7164289355278015,
+      "learning_rate": 4.0485368845277735e-05,
+      "loss": 1.7111,
+      "step": 6796
+    },
+    {
+      "epoch": 1.786185831116366,
+      "grad_norm": 0.5487950444221497,
+      "learning_rate": 4.0467846504292976e-05,
+      "loss": 1.6346,
+      "step": 6798
+    },
+    {
+      "epoch": 1.786711334450028,
+      "grad_norm": 0.6355817914009094,
+      "learning_rate": 4.045032416330822e-05,
+      "loss": 1.6228,
+      "step": 6800
+    },
+    {
+      "epoch": 1.786711334450028,
+      "eval_loss": 1.6711539030075073,
+      "eval_runtime": 487.2898,
+      "eval_samples_per_second": 249.931,
+      "eval_steps_per_second": 31.242,
+      "step": 6800
+    },
+    {
+      "epoch": 1.7872368377836896,
+      "grad_norm": 0.5209816694259644,
+      "learning_rate": 4.0432801822323464e-05,
+      "loss": 1.6396,
+      "step": 6802
+    },
+    {
+      "epoch": 1.7877623411173515,
+      "grad_norm": 0.5979394912719727,
+      "learning_rate": 4.0415279481338705e-05,
+      "loss": 1.6686,
+      "step": 6804
+    },
+    {
+      "epoch": 1.788287844451013,
+      "grad_norm": 0.5141789317131042,
+      "learning_rate": 4.039775714035395e-05,
+      "loss": 1.6538,
+      "step": 6806
+    },
+    {
+      "epoch": 1.788813347784675,
+      "grad_norm": 0.5531857013702393,
+      "learning_rate": 4.03802347993692e-05,
+      "loss": 1.6612,
+      "step": 6808
+    },
+    {
+      "epoch": 1.7893388511183368,
+      "grad_norm": 0.5284379720687866,
+      "learning_rate": 4.036271245838444e-05,
+      "loss": 1.6403,
+      "step": 6810
+    },
+    {
+      "epoch": 1.7898643544519985,
+      "grad_norm": 0.5298596620559692,
+      "learning_rate": 4.034519011739968e-05,
+      "loss": 1.6494,
+      "step": 6812
+    },
+    {
+      "epoch": 1.7903898577856603,
+      "grad_norm": 0.5482889413833618,
+      "learning_rate": 4.032766777641493e-05,
+      "loss": 1.6516,
+      "step": 6814
+    },
+    {
+      "epoch": 1.790915361119322,
+      "grad_norm": 0.5631160736083984,
+      "learning_rate": 4.031014543543018e-05,
+      "loss": 1.6419,
+      "step": 6816
+    },
+    {
+      "epoch": 1.791440864452984,
+      "grad_norm": 0.5150030851364136,
+      "learning_rate": 4.029262309444542e-05,
+      "loss": 1.6502,
+      "step": 6818
+    },
+    {
+      "epoch": 1.7919663677866455,
+      "grad_norm": 0.5491872429847717,
+      "learning_rate": 4.0275100753460665e-05,
+      "loss": 1.6489,
+      "step": 6820
+    },
+    {
+      "epoch": 1.7924918711203075,
+      "grad_norm": 0.5404025912284851,
+      "learning_rate": 4.025757841247591e-05,
+      "loss": 1.6455,
+      "step": 6822
+    },
+    {
+      "epoch": 1.793017374453969,
+      "grad_norm": 0.5373459458351135,
+      "learning_rate": 4.0240056071491154e-05,
+      "loss": 1.6487,
+      "step": 6824
+    },
+    {
+      "epoch": 1.793542877787631,
+      "grad_norm": 0.6558994054794312,
+      "learning_rate": 4.0222533730506394e-05,
+      "loss": 1.6603,
+      "step": 6826
+    },
+    {
+      "epoch": 1.7940683811212927,
+      "grad_norm": 0.5680163502693176,
+      "learning_rate": 4.020501138952164e-05,
+      "loss": 1.6607,
+      "step": 6828
+    },
+    {
+      "epoch": 1.7945938844549545,
+      "grad_norm": 0.5356695652008057,
+      "learning_rate": 4.018748904853688e-05,
+      "loss": 1.6526,
+      "step": 6830
+    },
+    {
+      "epoch": 1.7951193877886162,
+      "grad_norm": 0.5755831003189087,
+      "learning_rate": 4.016996670755213e-05,
+      "loss": 1.6748,
+      "step": 6832
+    },
+    {
+      "epoch": 1.795644891122278,
+      "grad_norm": 0.6050164103507996,
+      "learning_rate": 4.015244436656738e-05,
+      "loss": 1.6366,
+      "step": 6834
+    },
+    {
+      "epoch": 1.79617039445594,
+      "grad_norm": 0.5978443026542664,
+      "learning_rate": 4.013492202558262e-05,
+      "loss": 1.6549,
+      "step": 6836
+    },
+    {
+      "epoch": 1.7966958977896015,
+      "grad_norm": 0.806139349937439,
+      "learning_rate": 4.011739968459786e-05,
+      "loss": 1.6486,
+      "step": 6838
+    },
+    {
+      "epoch": 1.7972214011232635,
+      "grad_norm": 0.665317714214325,
+      "learning_rate": 4.009987734361311e-05,
+      "loss": 1.6468,
+      "step": 6840
+    },
+    {
+      "epoch": 1.7977469044569252,
+      "grad_norm": 0.5707154870033264,
+      "learning_rate": 4.0082355002628355e-05,
+      "loss": 1.6592,
+      "step": 6842
+    },
+    {
+      "epoch": 1.798272407790587,
+      "grad_norm": 0.5100306868553162,
+      "learning_rate": 4.0064832661643595e-05,
+      "loss": 1.6732,
+      "step": 6844
+    },
+    {
+      "epoch": 1.7987979111242487,
+      "grad_norm": 0.4903377294540405,
+      "learning_rate": 4.004731032065884e-05,
+      "loss": 1.6383,
+      "step": 6846
+    },
+    {
+      "epoch": 1.7993234144579104,
+      "grad_norm": 0.5019045472145081,
+      "learning_rate": 4.002978797967409e-05,
+      "loss": 1.6668,
+      "step": 6848
+    },
+    {
+      "epoch": 1.7998489177915722,
+      "grad_norm": 0.5553399324417114,
+      "learning_rate": 4.001226563868933e-05,
+      "loss": 1.6666,
+      "step": 6850
+    },
+    {
+      "epoch": 1.800374421125234,
+      "grad_norm": 0.5196052193641663,
+      "learning_rate": 3.999474329770457e-05,
+      "loss": 1.6928,
+      "step": 6852
+    },
+    {
+      "epoch": 1.800899924458896,
+      "grad_norm": 0.5712267756462097,
+      "learning_rate": 3.997722095671982e-05,
+      "loss": 1.6228,
+      "step": 6854
+    },
+    {
+      "epoch": 1.8014254277925574,
+      "grad_norm": 0.6430991291999817,
+      "learning_rate": 3.995969861573507e-05,
+      "loss": 1.6583,
+      "step": 6856
+    },
+    {
+      "epoch": 1.8019509311262194,
+      "grad_norm": 0.5435091257095337,
+      "learning_rate": 3.994217627475031e-05,
+      "loss": 1.6458,
+      "step": 6858
+    },
+    {
+      "epoch": 1.8024764344598811,
+      "grad_norm": 0.5334445238113403,
+      "learning_rate": 3.9924653933765556e-05,
+      "loss": 1.6534,
+      "step": 6860
+    },
+    {
+      "epoch": 1.803001937793543,
+      "grad_norm": 0.6035925149917603,
+      "learning_rate": 3.99071315927808e-05,
+      "loss": 1.6449,
+      "step": 6862
+    },
+    {
+      "epoch": 1.8035274411272046,
+      "grad_norm": 0.6789858341217041,
+      "learning_rate": 3.988960925179604e-05,
+      "loss": 1.6599,
+      "step": 6864
+    },
+    {
+      "epoch": 1.8040529444608664,
+      "grad_norm": 0.6741647720336914,
+      "learning_rate": 3.9872086910811285e-05,
+      "loss": 1.6474,
+      "step": 6866
+    },
+    {
+      "epoch": 1.8045784477945284,
+      "grad_norm": 0.601939857006073,
+      "learning_rate": 3.985456456982653e-05,
+      "loss": 1.657,
+      "step": 6868
+    },
+    {
+      "epoch": 1.8051039511281899,
+      "grad_norm": 0.5084041357040405,
+      "learning_rate": 3.983704222884177e-05,
+      "loss": 1.6707,
+      "step": 6870
+    },
+    {
+      "epoch": 1.8056294544618519,
+      "grad_norm": 0.6498923897743225,
+      "learning_rate": 3.981951988785702e-05,
+      "loss": 1.6738,
+      "step": 6872
+    },
+    {
+      "epoch": 1.8061549577955134,
+      "grad_norm": 0.6679303050041199,
+      "learning_rate": 3.980199754687227e-05,
+      "loss": 1.6565,
+      "step": 6874
+    },
+    {
+      "epoch": 1.8066804611291754,
+      "grad_norm": 0.642929196357727,
+      "learning_rate": 3.978447520588751e-05,
+      "loss": 1.6437,
+      "step": 6876
+    },
+    {
+      "epoch": 1.807205964462837,
+      "grad_norm": 0.5809937119483948,
+      "learning_rate": 3.976695286490275e-05,
+      "loss": 1.6754,
+      "step": 6878
+    },
+    {
+      "epoch": 1.8077314677964988,
+      "grad_norm": 0.5564219355583191,
+      "learning_rate": 3.9749430523918e-05,
+      "loss": 1.6823,
+      "step": 6880
+    },
+    {
+      "epoch": 1.8082569711301606,
+      "grad_norm": 0.5706532001495361,
+      "learning_rate": 3.9731908182933245e-05,
+      "loss": 1.6705,
+      "step": 6882
+    },
+    {
+      "epoch": 1.8087824744638223,
+      "grad_norm": 0.5469827651977539,
+      "learning_rate": 3.9714385841948486e-05,
+      "loss": 1.6463,
+      "step": 6884
+    },
+    {
+      "epoch": 1.8093079777974843,
+      "grad_norm": 0.5436684489250183,
+      "learning_rate": 3.969686350096373e-05,
+      "loss": 1.6658,
+      "step": 6886
+    },
+    {
+      "epoch": 1.8098334811311458,
+      "grad_norm": 0.5233422517776489,
+      "learning_rate": 3.967934115997898e-05,
+      "loss": 1.6741,
+      "step": 6888
+    },
+    {
+      "epoch": 1.8103589844648078,
+      "grad_norm": 0.5667338967323303,
+      "learning_rate": 3.9661818818994215e-05,
+      "loss": 1.6323,
+      "step": 6890
+    },
+    {
+      "epoch": 1.8108844877984693,
+      "grad_norm": 0.5260610580444336,
+      "learning_rate": 3.964429647800946e-05,
+      "loss": 1.6492,
+      "step": 6892
+    },
+    {
+      "epoch": 1.8114099911321313,
+      "grad_norm": 0.554559051990509,
+      "learning_rate": 3.962677413702471e-05,
+      "loss": 1.6686,
+      "step": 6894
+    },
+    {
+      "epoch": 1.811935494465793,
+      "grad_norm": 0.6630009412765503,
+      "learning_rate": 3.960925179603995e-05,
+      "loss": 1.6507,
+      "step": 6896
+    },
+    {
+      "epoch": 1.8124609977994548,
+      "grad_norm": 0.5093562006950378,
+      "learning_rate": 3.95917294550552e-05,
+      "loss": 1.6631,
+      "step": 6898
+    },
+    {
+      "epoch": 1.8129865011331165,
+      "grad_norm": 0.5125998854637146,
+      "learning_rate": 3.9574207114070446e-05,
+      "loss": 1.6141,
+      "step": 6900
+    },
+    {
+      "epoch": 1.8135120044667783,
+      "grad_norm": 0.5436182022094727,
+      "learning_rate": 3.955668477308569e-05,
+      "loss": 1.6298,
+      "step": 6902
+    },
+    {
+      "epoch": 1.8140375078004403,
+      "grad_norm": 0.5174747705459595,
+      "learning_rate": 3.953916243210093e-05,
+      "loss": 1.6381,
+      "step": 6904
+    },
+    {
+      "epoch": 1.8145630111341018,
+      "grad_norm": 0.48531410098075867,
+      "learning_rate": 3.9521640091116175e-05,
+      "loss": 1.6465,
+      "step": 6906
+    },
+    {
+      "epoch": 1.8150885144677638,
+      "grad_norm": 0.5112138390541077,
+      "learning_rate": 3.950411775013142e-05,
+      "loss": 1.6469,
+      "step": 6908
+    },
+    {
+      "epoch": 1.8156140178014253,
+      "grad_norm": 0.578628420829773,
+      "learning_rate": 3.948659540914666e-05,
+      "loss": 1.6874,
+      "step": 6910
+    },
+    {
+      "epoch": 1.8161395211350873,
+      "grad_norm": 0.5760912895202637,
+      "learning_rate": 3.946907306816191e-05,
+      "loss": 1.6661,
+      "step": 6912
+    },
+    {
+      "epoch": 1.816665024468749,
+      "grad_norm": 0.5140530467033386,
+      "learning_rate": 3.945155072717715e-05,
+      "loss": 1.6359,
+      "step": 6914
+    },
+    {
+      "epoch": 1.8171905278024107,
+      "grad_norm": 0.5136293172836304,
+      "learning_rate": 3.943402838619239e-05,
+      "loss": 1.6839,
+      "step": 6916
+    },
+    {
+      "epoch": 1.8177160311360725,
+      "grad_norm": 0.6095285415649414,
+      "learning_rate": 3.941650604520764e-05,
+      "loss": 1.6552,
+      "step": 6918
+    },
+    {
+      "epoch": 1.8182415344697342,
+      "grad_norm": 0.5882896780967712,
+      "learning_rate": 3.939898370422289e-05,
+      "loss": 1.6532,
+      "step": 6920
+    },
+    {
+      "epoch": 1.8187670378033962,
+      "grad_norm": 0.5088122487068176,
+      "learning_rate": 3.938146136323813e-05,
+      "loss": 1.6719,
+      "step": 6922
+    },
+    {
+      "epoch": 1.8192925411370577,
+      "grad_norm": 0.5035478472709656,
+      "learning_rate": 3.9363939022253376e-05,
+      "loss": 1.641,
+      "step": 6924
+    },
+    {
+      "epoch": 1.8198180444707197,
+      "grad_norm": 0.559293270111084,
+      "learning_rate": 3.9346416681268624e-05,
+      "loss": 1.6971,
+      "step": 6926
+    },
+    {
+      "epoch": 1.8203435478043812,
+      "grad_norm": 0.5681344866752625,
+      "learning_rate": 3.9328894340283864e-05,
+      "loss": 1.6569,
+      "step": 6928
+    },
+    {
+      "epoch": 1.8208690511380432,
+      "grad_norm": 0.6714912056922913,
+      "learning_rate": 3.9311371999299105e-05,
+      "loss": 1.7089,
+      "step": 6930
+    },
+    {
+      "epoch": 1.821394554471705,
+      "grad_norm": 0.5596345067024231,
+      "learning_rate": 3.929384965831435e-05,
+      "loss": 1.6888,
+      "step": 6932
+    },
+    {
+      "epoch": 1.8219200578053667,
+      "grad_norm": 0.5444710850715637,
+      "learning_rate": 3.92763273173296e-05,
+      "loss": 1.6772,
+      "step": 6934
+    },
+    {
+      "epoch": 1.8224455611390284,
+      "grad_norm": 0.5050802230834961,
+      "learning_rate": 3.925880497634484e-05,
+      "loss": 1.6613,
+      "step": 6936
+    },
+    {
+      "epoch": 1.8229710644726902,
+      "grad_norm": 0.5749676823616028,
+      "learning_rate": 3.924128263536009e-05,
+      "loss": 1.6448,
+      "step": 6938
+    },
+    {
+      "epoch": 1.8234965678063522,
+      "grad_norm": 0.577195405960083,
+      "learning_rate": 3.922376029437533e-05,
+      "loss": 1.6614,
+      "step": 6940
+    },
+    {
+      "epoch": 1.8240220711400137,
+      "grad_norm": 0.5361042022705078,
+      "learning_rate": 3.920623795339057e-05,
+      "loss": 1.6764,
+      "step": 6942
+    },
+    {
+      "epoch": 1.8245475744736757,
+      "grad_norm": 0.5485758781433105,
+      "learning_rate": 3.918871561240582e-05,
+      "loss": 1.6486,
+      "step": 6944
+    },
+    {
+      "epoch": 1.8250730778073372,
+      "grad_norm": 0.7427318096160889,
+      "learning_rate": 3.9171193271421065e-05,
+      "loss": 1.6755,
+      "step": 6946
+    },
+    {
+      "epoch": 1.8255985811409992,
+      "grad_norm": 0.5816475749015808,
+      "learning_rate": 3.9153670930436306e-05,
+      "loss": 1.6518,
+      "step": 6948
+    },
+    {
+      "epoch": 1.826124084474661,
+      "grad_norm": 0.48851844668388367,
+      "learning_rate": 3.9136148589451554e-05,
+      "loss": 1.6655,
+      "step": 6950
+    },
+    {
+      "epoch": 1.8266495878083227,
+      "grad_norm": 0.5446299910545349,
+      "learning_rate": 3.91186262484668e-05,
+      "loss": 1.6374,
+      "step": 6952
+    },
+    {
+      "epoch": 1.8271750911419844,
+      "grad_norm": 0.514245331287384,
+      "learning_rate": 3.910110390748204e-05,
+      "loss": 1.6463,
+      "step": 6954
+    },
+    {
+      "epoch": 1.8277005944756461,
+      "grad_norm": 0.5527070760726929,
+      "learning_rate": 3.908358156649728e-05,
+      "loss": 1.6874,
+      "step": 6956
+    },
+    {
+      "epoch": 1.8282260978093081,
+      "grad_norm": 0.5745643377304077,
+      "learning_rate": 3.906605922551253e-05,
+      "loss": 1.6334,
+      "step": 6958
+    },
+    {
+      "epoch": 1.8287516011429696,
+      "grad_norm": 0.5475890636444092,
+      "learning_rate": 3.904853688452778e-05,
+      "loss": 1.6336,
+      "step": 6960
+    },
+    {
+      "epoch": 1.8292771044766316,
+      "grad_norm": 0.5883870124816895,
+      "learning_rate": 3.903101454354302e-05,
+      "loss": 1.6226,
+      "step": 6962
+    },
+    {
+      "epoch": 1.8298026078102931,
+      "grad_norm": 0.5167810916900635,
+      "learning_rate": 3.9013492202558266e-05,
+      "loss": 1.6612,
+      "step": 6964
+    },
+    {
+      "epoch": 1.830328111143955,
+      "grad_norm": 0.5100778937339783,
+      "learning_rate": 3.899596986157351e-05,
+      "loss": 1.6666,
+      "step": 6966
+    },
+    {
+      "epoch": 1.8308536144776169,
+      "grad_norm": 0.5465711355209351,
+      "learning_rate": 3.897844752058875e-05,
+      "loss": 1.6586,
+      "step": 6968
+    },
+    {
+      "epoch": 1.8313791178112786,
+      "grad_norm": 0.5632458329200745,
+      "learning_rate": 3.8960925179603995e-05,
+      "loss": 1.6729,
+      "step": 6970
+    },
+    {
+      "epoch": 1.8319046211449403,
+      "grad_norm": 0.7153643369674683,
+      "learning_rate": 3.894340283861924e-05,
+      "loss": 1.6767,
+      "step": 6972
+    },
+    {
+      "epoch": 1.832430124478602,
+      "grad_norm": 0.5567420721054077,
+      "learning_rate": 3.8925880497634484e-05,
+      "loss": 1.6836,
+      "step": 6974
+    },
+    {
+      "epoch": 1.832955627812264,
+      "grad_norm": 0.6071416735649109,
+      "learning_rate": 3.890835815664973e-05,
+      "loss": 1.6403,
+      "step": 6976
+    },
+    {
+      "epoch": 1.8334811311459256,
+      "grad_norm": 0.5392048954963684,
+      "learning_rate": 3.889083581566497e-05,
+      "loss": 1.6667,
+      "step": 6978
+    },
+    {
+      "epoch": 1.8340066344795876,
+      "grad_norm": 0.7555952668190002,
+      "learning_rate": 3.887331347468022e-05,
+      "loss": 1.6623,
+      "step": 6980
+    },
+    {
+      "epoch": 1.834532137813249,
+      "grad_norm": 0.6968433856964111,
+      "learning_rate": 3.885579113369546e-05,
+      "loss": 1.6556,
+      "step": 6982
+    },
+    {
+      "epoch": 1.835057641146911,
+      "grad_norm": 0.6319105625152588,
+      "learning_rate": 3.883826879271071e-05,
+      "loss": 1.7025,
+      "step": 6984
+    },
+    {
+      "epoch": 1.8355831444805728,
+      "grad_norm": 0.5621939301490784,
+      "learning_rate": 3.8820746451725956e-05,
+      "loss": 1.639,
+      "step": 6986
+    },
+    {
+      "epoch": 1.8361086478142346,
+      "grad_norm": 0.5498519539833069,
+      "learning_rate": 3.8803224110741196e-05,
+      "loss": 1.6726,
+      "step": 6988
+    },
+    {
+      "epoch": 1.8366341511478963,
+      "grad_norm": 0.6178312301635742,
+      "learning_rate": 3.8785701769756444e-05,
+      "loss": 1.6525,
+      "step": 6990
+    },
+    {
+      "epoch": 1.837159654481558,
+      "grad_norm": 0.4725324511528015,
+      "learning_rate": 3.8768179428771685e-05,
+      "loss": 1.6423,
+      "step": 6992
+    },
+    {
+      "epoch": 1.83768515781522,
+      "grad_norm": 0.5918049812316895,
+      "learning_rate": 3.875065708778693e-05,
+      "loss": 1.6728,
+      "step": 6994
+    },
+    {
+      "epoch": 1.8382106611488815,
+      "grad_norm": 0.6157652735710144,
+      "learning_rate": 3.873313474680217e-05,
+      "loss": 1.6549,
+      "step": 6996
+    },
+    {
+      "epoch": 1.8387361644825435,
+      "grad_norm": 0.596299409866333,
+      "learning_rate": 3.871561240581742e-05,
+      "loss": 1.6488,
+      "step": 6998
+    },
+    {
+      "epoch": 1.8392616678162053,
+      "grad_norm": 0.4981100559234619,
+      "learning_rate": 3.869809006483267e-05,
+      "loss": 1.6163,
+      "step": 7000
+    },
+    {
+      "epoch": 1.839787171149867,
+      "grad_norm": 0.5667177438735962,
+      "learning_rate": 3.868056772384791e-05,
+      "loss": 1.7114,
+      "step": 7002
+    },
+    {
+      "epoch": 1.8403126744835288,
+      "grad_norm": 0.5904120206832886,
+      "learning_rate": 3.866304538286315e-05,
+      "loss": 1.644,
+      "step": 7004
+    },
+    {
+      "epoch": 1.8408381778171905,
+      "grad_norm": 0.5185456871986389,
+      "learning_rate": 3.86455230418784e-05,
+      "loss": 1.6537,
+      "step": 7006
+    },
+    {
+      "epoch": 1.8413636811508522,
+      "grad_norm": 0.5449099540710449,
+      "learning_rate": 3.862800070089364e-05,
+      "loss": 1.6807,
+      "step": 7008
+    },
+    {
+      "epoch": 1.841889184484514,
+      "grad_norm": 0.7209144234657288,
+      "learning_rate": 3.8610478359908886e-05,
+      "loss": 1.6648,
+      "step": 7010
+    },
+    {
+      "epoch": 1.842414687818176,
+      "grad_norm": 0.6890124082565308,
+      "learning_rate": 3.859295601892413e-05,
+      "loss": 1.647,
+      "step": 7012
+    },
+    {
+      "epoch": 1.8429401911518375,
+      "grad_norm": 0.5385224223136902,
+      "learning_rate": 3.8575433677939374e-05,
+      "loss": 1.6475,
+      "step": 7014
+    },
+    {
+      "epoch": 1.8434656944854995,
+      "grad_norm": 0.5653911232948303,
+      "learning_rate": 3.8557911336954615e-05,
+      "loss": 1.6752,
+      "step": 7016
+    },
+    {
+      "epoch": 1.8439911978191612,
+      "grad_norm": 0.5241896510124207,
+      "learning_rate": 3.854038899596986e-05,
+      "loss": 1.6747,
+      "step": 7018
+    },
+    {
+      "epoch": 1.844516701152823,
+      "grad_norm": 0.7143029570579529,
+      "learning_rate": 3.852286665498511e-05,
+      "loss": 1.6538,
+      "step": 7020
+    },
+    {
+      "epoch": 1.8450422044864847,
+      "grad_norm": 0.5827885270118713,
+      "learning_rate": 3.850534431400035e-05,
+      "loss": 1.6987,
+      "step": 7022
+    },
+    {
+      "epoch": 1.8455677078201465,
+      "grad_norm": 0.6297259330749512,
+      "learning_rate": 3.84878219730156e-05,
+      "loss": 1.6628,
+      "step": 7024
+    },
+    {
+      "epoch": 1.8460932111538084,
+      "grad_norm": 0.5605854392051697,
+      "learning_rate": 3.8470299632030846e-05,
+      "loss": 1.6216,
+      "step": 7026
+    },
+    {
+      "epoch": 1.84661871448747,
+      "grad_norm": 0.6488915085792542,
+      "learning_rate": 3.845277729104609e-05,
+      "loss": 1.7036,
+      "step": 7028
+    },
+    {
+      "epoch": 1.847144217821132,
+      "grad_norm": 0.6576531529426575,
+      "learning_rate": 3.843525495006133e-05,
+      "loss": 1.6846,
+      "step": 7030
+    },
+    {
+      "epoch": 1.8476697211547934,
+      "grad_norm": 0.49783241748809814,
+      "learning_rate": 3.8417732609076575e-05,
+      "loss": 1.6566,
+      "step": 7032
+    },
+    {
+      "epoch": 1.8481952244884554,
+      "grad_norm": 0.5974870324134827,
+      "learning_rate": 3.8400210268091816e-05,
+      "loss": 1.6787,
+      "step": 7034
+    },
+    {
+      "epoch": 1.8487207278221172,
+      "grad_norm": 0.5442907810211182,
+      "learning_rate": 3.838268792710706e-05,
+      "loss": 1.6414,
+      "step": 7036
+    },
+    {
+      "epoch": 1.849246231155779,
+      "grad_norm": 0.4982667565345764,
+      "learning_rate": 3.836516558612231e-05,
+      "loss": 1.664,
+      "step": 7038
+    },
+    {
+      "epoch": 1.8497717344894407,
+      "grad_norm": 0.5198219418525696,
+      "learning_rate": 3.834764324513755e-05,
+      "loss": 1.6708,
+      "step": 7040
+    },
+    {
+      "epoch": 1.8502972378231024,
+      "grad_norm": 0.6068596839904785,
+      "learning_rate": 3.833012090415279e-05,
+      "loss": 1.6467,
+      "step": 7042
+    },
+    {
+      "epoch": 1.8508227411567644,
+      "grad_norm": 0.5471599102020264,
+      "learning_rate": 3.831259856316804e-05,
+      "loss": 1.6371,
+      "step": 7044
+    },
+    {
+      "epoch": 1.851348244490426,
+      "grad_norm": 0.7527357339859009,
+      "learning_rate": 3.829507622218329e-05,
+      "loss": 1.6562,
+      "step": 7046
+    },
+    {
+      "epoch": 1.8518737478240879,
+      "grad_norm": 0.5454807877540588,
+      "learning_rate": 3.827755388119853e-05,
+      "loss": 1.6923,
+      "step": 7048
+    },
+    {
+      "epoch": 1.8523992511577494,
+      "grad_norm": 0.5361452102661133,
+      "learning_rate": 3.8260031540213776e-05,
+      "loss": 1.6479,
+      "step": 7050
+    },
+    {
+      "epoch": 1.8529247544914114,
+      "grad_norm": 0.5446274876594543,
+      "learning_rate": 3.8242509199229024e-05,
+      "loss": 1.6576,
+      "step": 7052
+    },
+    {
+      "epoch": 1.853450257825073,
+      "grad_norm": 0.6107861995697021,
+      "learning_rate": 3.822498685824426e-05,
+      "loss": 1.6777,
+      "step": 7054
+    },
+    {
+      "epoch": 1.8539757611587349,
+      "grad_norm": 0.526027262210846,
+      "learning_rate": 3.8207464517259505e-05,
+      "loss": 1.6423,
+      "step": 7056
+    },
+    {
+      "epoch": 1.8545012644923966,
+      "grad_norm": 0.5729789137840271,
+      "learning_rate": 3.818994217627475e-05,
+      "loss": 1.6364,
+      "step": 7058
+    },
+    {
+      "epoch": 1.8550267678260584,
+      "grad_norm": 0.5303583145141602,
+      "learning_rate": 3.8172419835289993e-05,
+      "loss": 1.6984,
+      "step": 7060
+    },
+    {
+      "epoch": 1.8555522711597203,
+      "grad_norm": 0.6077755093574524,
+      "learning_rate": 3.815489749430524e-05,
+      "loss": 1.6934,
+      "step": 7062
+    },
+    {
+      "epoch": 1.8560777744933818,
+      "grad_norm": 0.5593380331993103,
+      "learning_rate": 3.813737515332049e-05,
+      "loss": 1.6546,
+      "step": 7064
+    },
+    {
+      "epoch": 1.8566032778270438,
+      "grad_norm": 0.5528572797775269,
+      "learning_rate": 3.811985281233573e-05,
+      "loss": 1.6722,
+      "step": 7066
+    },
+    {
+      "epoch": 1.8571287811607053,
+      "grad_norm": 0.5037624835968018,
+      "learning_rate": 3.810233047135097e-05,
+      "loss": 1.6701,
+      "step": 7068
+    },
+    {
+      "epoch": 1.8576542844943673,
+      "grad_norm": 0.5588459968566895,
+      "learning_rate": 3.808480813036622e-05,
+      "loss": 1.6871,
+      "step": 7070
+    },
+    {
+      "epoch": 1.858179787828029,
+      "grad_norm": 0.5192500948905945,
+      "learning_rate": 3.8067285789381465e-05,
+      "loss": 1.6539,
+      "step": 7072
+    },
+    {
+      "epoch": 1.8587052911616908,
+      "grad_norm": 0.5383935570716858,
+      "learning_rate": 3.8049763448396706e-05,
+      "loss": 1.672,
+      "step": 7074
+    },
+    {
+      "epoch": 1.8592307944953526,
+      "grad_norm": 0.6632468104362488,
+      "learning_rate": 3.8032241107411954e-05,
+      "loss": 1.6735,
+      "step": 7076
+    },
+    {
+      "epoch": 1.8597562978290143,
+      "grad_norm": 0.5410235524177551,
+      "learning_rate": 3.80147187664272e-05,
+      "loss": 1.6374,
+      "step": 7078
+    },
+    {
+      "epoch": 1.8602818011626763,
+      "grad_norm": 0.7134387493133545,
+      "learning_rate": 3.7997196425442435e-05,
+      "loss": 1.6661,
+      "step": 7080
+    },
+    {
+      "epoch": 1.8608073044963378,
+      "grad_norm": 0.54608154296875,
+      "learning_rate": 3.797967408445768e-05,
+      "loss": 1.6297,
+      "step": 7082
+    },
+    {
+      "epoch": 1.8613328078299998,
+      "grad_norm": 0.6386792659759521,
+      "learning_rate": 3.796215174347293e-05,
+      "loss": 1.6489,
+      "step": 7084
+    },
+    {
+      "epoch": 1.8618583111636613,
+      "grad_norm": 0.5409175157546997,
+      "learning_rate": 3.794462940248817e-05,
+      "loss": 1.6609,
+      "step": 7086
+    },
+    {
+      "epoch": 1.8623838144973233,
+      "grad_norm": 0.611273467540741,
+      "learning_rate": 3.792710706150342e-05,
+      "loss": 1.6685,
+      "step": 7088
+    },
+    {
+      "epoch": 1.862909317830985,
+      "grad_norm": 0.6014897227287292,
+      "learning_rate": 3.7909584720518666e-05,
+      "loss": 1.6068,
+      "step": 7090
+    },
+    {
+      "epoch": 1.8634348211646468,
+      "grad_norm": 0.5631469488143921,
+      "learning_rate": 3.789206237953391e-05,
+      "loss": 1.6478,
+      "step": 7092
+    },
+    {
+      "epoch": 1.8639603244983085,
+      "grad_norm": 0.566448450088501,
+      "learning_rate": 3.787454003854915e-05,
+      "loss": 1.6861,
+      "step": 7094
+    },
+    {
+      "epoch": 1.8644858278319703,
+      "grad_norm": 0.5702322721481323,
+      "learning_rate": 3.7857017697564395e-05,
+      "loss": 1.6507,
+      "step": 7096
+    },
+    {
+      "epoch": 1.8650113311656322,
+      "grad_norm": 0.5720840096473694,
+      "learning_rate": 3.783949535657964e-05,
+      "loss": 1.6615,
+      "step": 7098
+    },
+    {
+      "epoch": 1.8655368344992938,
+      "grad_norm": 0.586739718914032,
+      "learning_rate": 3.7821973015594884e-05,
+      "loss": 1.6837,
+      "step": 7100
+    },
+    {
+      "epoch": 1.8660623378329557,
+      "grad_norm": 0.5152260065078735,
+      "learning_rate": 3.780445067461013e-05,
+      "loss": 1.6565,
+      "step": 7102
+    },
+    {
+      "epoch": 1.8665878411666172,
+      "grad_norm": 0.5608677864074707,
+      "learning_rate": 3.778692833362538e-05,
+      "loss": 1.6448,
+      "step": 7104
+    },
+    {
+      "epoch": 1.8671133445002792,
+      "grad_norm": 0.5977829694747925,
+      "learning_rate": 3.776940599264062e-05,
+      "loss": 1.6415,
+      "step": 7106
+    },
+    {
+      "epoch": 1.867638847833941,
+      "grad_norm": 0.6419270038604736,
+      "learning_rate": 3.775188365165586e-05,
+      "loss": 1.6032,
+      "step": 7108
+    },
+    {
+      "epoch": 1.8681643511676027,
+      "grad_norm": 0.6474549770355225,
+      "learning_rate": 3.773436131067111e-05,
+      "loss": 1.6717,
+      "step": 7110
+    },
+    {
+      "epoch": 1.8686898545012645,
+      "grad_norm": 0.6148583889007568,
+      "learning_rate": 3.7716838969686356e-05,
+      "loss": 1.6476,
+      "step": 7112
+    },
+    {
+      "epoch": 1.8692153578349262,
+      "grad_norm": 0.5739107728004456,
+      "learning_rate": 3.7699316628701596e-05,
+      "loss": 1.652,
+      "step": 7114
+    },
+    {
+      "epoch": 1.8697408611685882,
+      "grad_norm": 0.5666532516479492,
+      "learning_rate": 3.7681794287716844e-05,
+      "loss": 1.6611,
+      "step": 7116
+    },
+    {
+      "epoch": 1.8702663645022497,
+      "grad_norm": 0.6231557726860046,
+      "learning_rate": 3.7664271946732085e-05,
+      "loss": 1.6285,
+      "step": 7118
+    },
+    {
+      "epoch": 1.8707918678359117,
+      "grad_norm": 0.5692910552024841,
+      "learning_rate": 3.7646749605747325e-05,
+      "loss": 1.6195,
+      "step": 7120
+    },
+    {
+      "epoch": 1.8713173711695732,
+      "grad_norm": 0.5744662880897522,
+      "learning_rate": 3.762922726476257e-05,
+      "loss": 1.6603,
+      "step": 7122
+    },
+    {
+      "epoch": 1.8718428745032352,
+      "grad_norm": 0.5582786798477173,
+      "learning_rate": 3.761170492377782e-05,
+      "loss": 1.6719,
+      "step": 7124
+    },
+    {
+      "epoch": 1.872368377836897,
+      "grad_norm": 0.5148811340332031,
+      "learning_rate": 3.759418258279306e-05,
+      "loss": 1.6329,
+      "step": 7126
+    },
+    {
+      "epoch": 1.8728938811705587,
+      "grad_norm": 0.5253287553787231,
+      "learning_rate": 3.757666024180831e-05,
+      "loss": 1.6576,
+      "step": 7128
+    },
+    {
+      "epoch": 1.8734193845042204,
+      "grad_norm": 0.5456867814064026,
+      "learning_rate": 3.7559137900823557e-05,
+      "loss": 1.6682,
+      "step": 7130
+    },
+    {
+      "epoch": 1.8739448878378822,
+      "grad_norm": 0.5340244770050049,
+      "learning_rate": 3.75416155598388e-05,
+      "loss": 1.6758,
+      "step": 7132
+    },
+    {
+      "epoch": 1.8744703911715441,
+      "grad_norm": 0.5905424356460571,
+      "learning_rate": 3.752409321885404e-05,
+      "loss": 1.6707,
+      "step": 7134
+    },
+    {
+      "epoch": 1.8749958945052057,
+      "grad_norm": 0.5440637469291687,
+      "learning_rate": 3.7506570877869286e-05,
+      "loss": 1.6507,
+      "step": 7136
+    },
+    {
+      "epoch": 1.8755213978388676,
+      "grad_norm": 0.751801073551178,
+      "learning_rate": 3.748904853688453e-05,
+      "loss": 1.6656,
+      "step": 7138
+    },
+    {
+      "epoch": 1.8760469011725294,
+      "grad_norm": 0.5282323956489563,
+      "learning_rate": 3.7471526195899774e-05,
+      "loss": 1.6808,
+      "step": 7140
+    },
+    {
+      "epoch": 1.8765724045061911,
+      "grad_norm": 0.5995266437530518,
+      "learning_rate": 3.745400385491502e-05,
+      "loss": 1.6374,
+      "step": 7142
+    },
+    {
+      "epoch": 1.8770979078398529,
+      "grad_norm": 0.5131629705429077,
+      "learning_rate": 3.743648151393026e-05,
+      "loss": 1.631,
+      "step": 7144
+    },
+    {
+      "epoch": 1.8776234111735146,
+      "grad_norm": 0.5343523025512695,
+      "learning_rate": 3.74189591729455e-05,
+      "loss": 1.6253,
+      "step": 7146
+    },
+    {
+      "epoch": 1.8781489145071764,
+      "grad_norm": 0.6353945136070251,
+      "learning_rate": 3.740143683196075e-05,
+      "loss": 1.6843,
+      "step": 7148
+    },
+    {
+      "epoch": 1.878674417840838,
+      "grad_norm": 0.5876971483230591,
+      "learning_rate": 3.7383914490976e-05,
+      "loss": 1.6692,
+      "step": 7150
+    },
+    {
+      "epoch": 1.8791999211745,
+      "grad_norm": 0.5077223181724548,
+      "learning_rate": 3.736639214999124e-05,
+      "loss": 1.6095,
+      "step": 7152
+    },
+    {
+      "epoch": 1.8797254245081616,
+      "grad_norm": 0.5225921273231506,
+      "learning_rate": 3.734886980900649e-05,
+      "loss": 1.6475,
+      "step": 7154
+    },
+    {
+      "epoch": 1.8802509278418236,
+      "grad_norm": 0.549243688583374,
+      "learning_rate": 3.733134746802173e-05,
+      "loss": 1.6529,
+      "step": 7156
+    },
+    {
+      "epoch": 1.8807764311754853,
+      "grad_norm": 0.5567914843559265,
+      "learning_rate": 3.7313825127036975e-05,
+      "loss": 1.6621,
+      "step": 7158
+    },
+    {
+      "epoch": 1.881301934509147,
+      "grad_norm": 0.5527283549308777,
+      "learning_rate": 3.7296302786052216e-05,
+      "loss": 1.6705,
+      "step": 7160
+    },
+    {
+      "epoch": 1.8818274378428088,
+      "grad_norm": 0.5174548625946045,
+      "learning_rate": 3.727878044506746e-05,
+      "loss": 1.6606,
+      "step": 7162
+    },
+    {
+      "epoch": 1.8823529411764706,
+      "grad_norm": 0.527554988861084,
+      "learning_rate": 3.726125810408271e-05,
+      "loss": 1.6543,
+      "step": 7164
+    },
+    {
+      "epoch": 1.8828784445101323,
+      "grad_norm": 0.6432197093963623,
+      "learning_rate": 3.724373576309795e-05,
+      "loss": 1.6551,
+      "step": 7166
+    },
+    {
+      "epoch": 1.883403947843794,
+      "grad_norm": 0.6546508073806763,
+      "learning_rate": 3.72262134221132e-05,
+      "loss": 1.644,
+      "step": 7168
+    },
+    {
+      "epoch": 1.883929451177456,
+      "grad_norm": 0.5889551043510437,
+      "learning_rate": 3.720869108112844e-05,
+      "loss": 1.6767,
+      "step": 7170
+    },
+    {
+      "epoch": 1.8844549545111176,
+      "grad_norm": 0.5616403222084045,
+      "learning_rate": 3.719116874014368e-05,
+      "loss": 1.661,
+      "step": 7172
+    },
+    {
+      "epoch": 1.8849804578447795,
+      "grad_norm": 0.7210134267807007,
+      "learning_rate": 3.717364639915893e-05,
+      "loss": 1.6527,
+      "step": 7174
+    },
+    {
+      "epoch": 1.8855059611784413,
+      "grad_norm": 0.5026001930236816,
+      "learning_rate": 3.7156124058174176e-05,
+      "loss": 1.6589,
+      "step": 7176
+    },
+    {
+      "epoch": 1.886031464512103,
+      "grad_norm": 0.5873162746429443,
+      "learning_rate": 3.713860171718942e-05,
+      "loss": 1.622,
+      "step": 7178
+    },
+    {
+      "epoch": 1.8865569678457648,
+      "grad_norm": 0.729069709777832,
+      "learning_rate": 3.7121079376204664e-05,
+      "loss": 1.6229,
+      "step": 7180
+    },
+    {
+      "epoch": 1.8870824711794265,
+      "grad_norm": 0.5042127966880798,
+      "learning_rate": 3.7103557035219905e-05,
+      "loss": 1.6329,
+      "step": 7182
+    },
+    {
+      "epoch": 1.8876079745130885,
+      "grad_norm": 0.5304409861564636,
+      "learning_rate": 3.708603469423515e-05,
+      "loss": 1.6427,
+      "step": 7184
+    },
+    {
+      "epoch": 1.88813347784675,
+      "grad_norm": 0.6537191271781921,
+      "learning_rate": 3.7068512353250393e-05,
+      "loss": 1.662,
+      "step": 7186
+    },
+    {
+      "epoch": 1.888658981180412,
+      "grad_norm": 0.5206599235534668,
+      "learning_rate": 3.705099001226564e-05,
+      "loss": 1.6375,
+      "step": 7188
+    },
+    {
+      "epoch": 1.8891844845140735,
+      "grad_norm": 0.6516169905662537,
+      "learning_rate": 3.703346767128089e-05,
+      "loss": 1.6364,
+      "step": 7190
+    },
+    {
+      "epoch": 1.8897099878477355,
+      "grad_norm": 0.5258002877235413,
+      "learning_rate": 3.701594533029613e-05,
+      "loss": 1.6826,
+      "step": 7192
+    },
+    {
+      "epoch": 1.8902354911813972,
+      "grad_norm": 0.5022348761558533,
+      "learning_rate": 3.699842298931138e-05,
+      "loss": 1.64,
+      "step": 7194
+    },
+    {
+      "epoch": 1.890760994515059,
+      "grad_norm": 0.585818350315094,
+      "learning_rate": 3.698090064832662e-05,
+      "loss": 1.6968,
+      "step": 7196
+    },
+    {
+      "epoch": 1.8912864978487207,
+      "grad_norm": 0.5346968770027161,
+      "learning_rate": 3.696337830734186e-05,
+      "loss": 1.6514,
+      "step": 7198
+    },
+    {
+      "epoch": 1.8918120011823825,
+      "grad_norm": 0.6125622391700745,
+      "learning_rate": 3.6945855966357106e-05,
+      "loss": 1.6467,
+      "step": 7200
+    },
+    {
+      "epoch": 1.8918120011823825,
+      "eval_loss": 1.66790771484375,
+      "eval_runtime": 487.3046,
+      "eval_samples_per_second": 249.924,
+      "eval_steps_per_second": 31.241,
+      "step": 7200
+    },
+    {
+      "epoch": 1.8923375045160444,
+      "grad_norm": 0.5719324350357056,
+      "learning_rate": 3.6928333625372354e-05,
+      "loss": 1.6776,
+      "step": 7202
+    },
+    {
+      "epoch": 1.892863007849706,
+      "grad_norm": 0.5245485901832581,
+      "learning_rate": 3.6910811284387594e-05,
+      "loss": 1.6324,
+      "step": 7204
+    },
+    {
+      "epoch": 1.893388511183368,
+      "grad_norm": 0.5099211931228638,
+      "learning_rate": 3.689328894340284e-05,
+      "loss": 1.6764,
+      "step": 7206
+    },
+    {
+      "epoch": 1.8939140145170295,
+      "grad_norm": 0.7997536659240723,
+      "learning_rate": 3.687576660241808e-05,
+      "loss": 1.6619,
+      "step": 7208
+    },
+    {
+      "epoch": 1.8944395178506914,
+      "grad_norm": 0.5983949899673462,
+      "learning_rate": 3.685824426143333e-05,
+      "loss": 1.6557,
+      "step": 7210
+    },
+    {
+      "epoch": 1.8949650211843532,
+      "grad_norm": 0.6213306784629822,
+      "learning_rate": 3.684072192044857e-05,
+      "loss": 1.6559,
+      "step": 7212
+    },
+    {
+      "epoch": 1.895490524518015,
+      "grad_norm": 0.5697503685951233,
+      "learning_rate": 3.682319957946382e-05,
+      "loss": 1.6754,
+      "step": 7214
+    },
+    {
+      "epoch": 1.8960160278516767,
+      "grad_norm": 0.507168710231781,
+      "learning_rate": 3.6805677238479066e-05,
+      "loss": 1.6618,
+      "step": 7216
+    },
+    {
+      "epoch": 1.8965415311853384,
+      "grad_norm": 0.6169989109039307,
+      "learning_rate": 3.678815489749431e-05,
+      "loss": 1.6329,
+      "step": 7218
+    },
+    {
+      "epoch": 1.8970670345190004,
+      "grad_norm": 0.516326904296875,
+      "learning_rate": 3.677063255650955e-05,
+      "loss": 1.65,
+      "step": 7220
+    },
+    {
+      "epoch": 1.897592537852662,
+      "grad_norm": 0.5588873624801636,
+      "learning_rate": 3.6753110215524795e-05,
+      "loss": 1.6707,
+      "step": 7222
+    },
+    {
+      "epoch": 1.8981180411863239,
+      "grad_norm": 0.7022035717964172,
+      "learning_rate": 3.6735587874540036e-05,
+      "loss": 1.6935,
+      "step": 7224
+    },
+    {
+      "epoch": 1.8986435445199854,
+      "grad_norm": 0.5556198954582214,
+      "learning_rate": 3.6718065533555284e-05,
+      "loss": 1.6955,
+      "step": 7226
+    },
+    {
+      "epoch": 1.8991690478536474,
+      "grad_norm": 0.6451961398124695,
+      "learning_rate": 3.670054319257053e-05,
+      "loss": 1.6607,
+      "step": 7228
+    },
+    {
+      "epoch": 1.8996945511873091,
+      "grad_norm": 0.603360116481781,
+      "learning_rate": 3.668302085158577e-05,
+      "loss": 1.626,
+      "step": 7230
+    },
+    {
+      "epoch": 1.9002200545209709,
+      "grad_norm": 0.7140105962753296,
+      "learning_rate": 3.666549851060102e-05,
+      "loss": 1.6587,
+      "step": 7232
+    },
+    {
+      "epoch": 1.9007455578546326,
+      "grad_norm": 0.5646045804023743,
+      "learning_rate": 3.664797616961626e-05,
+      "loss": 1.6681,
+      "step": 7234
+    },
+    {
+      "epoch": 1.9012710611882944,
+      "grad_norm": 0.5183248519897461,
+      "learning_rate": 3.663045382863151e-05,
+      "loss": 1.6912,
+      "step": 7236
+    },
+    {
+      "epoch": 1.9017965645219563,
+      "grad_norm": 0.6256281137466431,
+      "learning_rate": 3.661293148764675e-05,
+      "loss": 1.6699,
+      "step": 7238
+    },
+    {
+      "epoch": 1.9023220678556179,
+      "grad_norm": 0.5462591052055359,
+      "learning_rate": 3.6595409146661996e-05,
+      "loss": 1.6628,
+      "step": 7240
+    },
+    {
+      "epoch": 1.9028475711892798,
+      "grad_norm": 0.715267539024353,
+      "learning_rate": 3.6577886805677244e-05,
+      "loss": 1.6907,
+      "step": 7242
+    },
+    {
+      "epoch": 1.9033730745229414,
+      "grad_norm": 0.8691731095314026,
+      "learning_rate": 3.6560364464692485e-05,
+      "loss": 1.6982,
+      "step": 7244
+    },
+    {
+      "epoch": 1.9038985778566033,
+      "grad_norm": 0.55055171251297,
+      "learning_rate": 3.6542842123707725e-05,
+      "loss": 1.6757,
+      "step": 7246
+    },
+    {
+      "epoch": 1.904424081190265,
+      "grad_norm": 0.5819158554077148,
+      "learning_rate": 3.652531978272297e-05,
+      "loss": 1.6571,
+      "step": 7248
+    },
+    {
+      "epoch": 1.9049495845239268,
+      "grad_norm": 0.6203599572181702,
+      "learning_rate": 3.650779744173822e-05,
+      "loss": 1.7292,
+      "step": 7250
+    },
+    {
+      "epoch": 1.9054750878575886,
+      "grad_norm": 0.5759249925613403,
+      "learning_rate": 3.649027510075346e-05,
+      "loss": 1.6227,
+      "step": 7252
+    },
+    {
+      "epoch": 1.9060005911912503,
+      "grad_norm": 0.581551730632782,
+      "learning_rate": 3.647275275976871e-05,
+      "loss": 1.6741,
+      "step": 7254
+    },
+    {
+      "epoch": 1.9065260945249123,
+      "grad_norm": 0.6072301864624023,
+      "learning_rate": 3.6455230418783956e-05,
+      "loss": 1.6663,
+      "step": 7256
+    },
+    {
+      "epoch": 1.9070515978585738,
+      "grad_norm": 0.5655650496482849,
+      "learning_rate": 3.643770807779919e-05,
+      "loss": 1.6729,
+      "step": 7258
+    },
+    {
+      "epoch": 1.9075771011922358,
+      "grad_norm": 0.5449069738388062,
+      "learning_rate": 3.642018573681444e-05,
+      "loss": 1.6514,
+      "step": 7260
+    },
+    {
+      "epoch": 1.9081026045258973,
+      "grad_norm": 0.573408842086792,
+      "learning_rate": 3.6402663395829686e-05,
+      "loss": 1.6718,
+      "step": 7262
+    },
+    {
+      "epoch": 1.9086281078595593,
+      "grad_norm": 0.7997104525566101,
+      "learning_rate": 3.6385141054844926e-05,
+      "loss": 1.702,
+      "step": 7264
+    },
+    {
+      "epoch": 1.909153611193221,
+      "grad_norm": 0.565268337726593,
+      "learning_rate": 3.6367618713860174e-05,
+      "loss": 1.6244,
+      "step": 7266
+    },
+    {
+      "epoch": 1.9096791145268828,
+      "grad_norm": 0.5995902419090271,
+      "learning_rate": 3.635009637287542e-05,
+      "loss": 1.6694,
+      "step": 7268
+    },
+    {
+      "epoch": 1.9102046178605445,
+      "grad_norm": 0.6907638907432556,
+      "learning_rate": 3.633257403189066e-05,
+      "loss": 1.6425,
+      "step": 7270
+    },
+    {
+      "epoch": 1.9107301211942063,
+      "grad_norm": 0.49709975719451904,
+      "learning_rate": 3.63150516909059e-05,
+      "loss": 1.6564,
+      "step": 7272
+    },
+    {
+      "epoch": 1.9112556245278682,
+      "grad_norm": 0.5460817813873291,
+      "learning_rate": 3.629752934992115e-05,
+      "loss": 1.6822,
+      "step": 7274
+    },
+    {
+      "epoch": 1.9117811278615298,
+      "grad_norm": 0.6267193555831909,
+      "learning_rate": 3.62800070089364e-05,
+      "loss": 1.6325,
+      "step": 7276
+    },
+    {
+      "epoch": 1.9123066311951917,
+      "grad_norm": 0.5849470496177673,
+      "learning_rate": 3.626248466795164e-05,
+      "loss": 1.6596,
+      "step": 7278
+    },
+    {
+      "epoch": 1.9128321345288533,
+      "grad_norm": 0.5416064262390137,
+      "learning_rate": 3.6244962326966887e-05,
+      "loss": 1.6786,
+      "step": 7280
+    },
+    {
+      "epoch": 1.9133576378625152,
+      "grad_norm": 0.6678909063339233,
+      "learning_rate": 3.6227439985982134e-05,
+      "loss": 1.6392,
+      "step": 7282
+    },
+    {
+      "epoch": 1.913883141196177,
+      "grad_norm": 0.5513505935668945,
+      "learning_rate": 3.620991764499737e-05,
+      "loss": 1.6942,
+      "step": 7284
+    },
+    {
+      "epoch": 1.9144086445298387,
+      "grad_norm": 0.7723992466926575,
+      "learning_rate": 3.6192395304012616e-05,
+      "loss": 1.6853,
+      "step": 7286
+    },
+    {
+      "epoch": 1.9149341478635005,
+      "grad_norm": 0.610593318939209,
+      "learning_rate": 3.617487296302786e-05,
+      "loss": 1.6927,
+      "step": 7288
+    },
+    {
+      "epoch": 1.9154596511971622,
+      "grad_norm": 0.6118401288986206,
+      "learning_rate": 3.6157350622043104e-05,
+      "loss": 1.6397,
+      "step": 7290
+    },
+    {
+      "epoch": 1.9159851545308242,
+      "grad_norm": 0.5796778798103333,
+      "learning_rate": 3.613982828105835e-05,
+      "loss": 1.6625,
+      "step": 7292
+    },
+    {
+      "epoch": 1.9165106578644857,
+      "grad_norm": 0.520003080368042,
+      "learning_rate": 3.61223059400736e-05,
+      "loss": 1.6902,
+      "step": 7294
+    },
+    {
+      "epoch": 1.9170361611981477,
+      "grad_norm": 0.6326977610588074,
+      "learning_rate": 3.610478359908884e-05,
+      "loss": 1.6767,
+      "step": 7296
+    },
+    {
+      "epoch": 1.9175616645318094,
+      "grad_norm": 0.4994673728942871,
+      "learning_rate": 3.608726125810408e-05,
+      "loss": 1.625,
+      "step": 7298
+    },
+    {
+      "epoch": 1.9180871678654712,
+      "grad_norm": 0.4929233193397522,
+      "learning_rate": 3.606973891711933e-05,
+      "loss": 1.7034,
+      "step": 7300
+    },
+    {
+      "epoch": 1.918612671199133,
+      "grad_norm": 0.5794808268547058,
+      "learning_rate": 3.6052216576134576e-05,
+      "loss": 1.6522,
+      "step": 7302
+    },
+    {
+      "epoch": 1.9191381745327947,
+      "grad_norm": 0.5688311457633972,
+      "learning_rate": 3.603469423514982e-05,
+      "loss": 1.7031,
+      "step": 7304
+    },
+    {
+      "epoch": 1.9196636778664564,
+      "grad_norm": 0.5700064897537231,
+      "learning_rate": 3.6017171894165064e-05,
+      "loss": 1.6469,
+      "step": 7306
+    },
+    {
+      "epoch": 1.9201891812001182,
+      "grad_norm": 0.4965689480304718,
+      "learning_rate": 3.599964955318031e-05,
+      "loss": 1.6539,
+      "step": 7308
+    },
+    {
+      "epoch": 1.9207146845337801,
+      "grad_norm": 0.6212711334228516,
+      "learning_rate": 3.5982127212195546e-05,
+      "loss": 1.6558,
+      "step": 7310
+    },
+    {
+      "epoch": 1.9212401878674417,
+      "grad_norm": 0.5985603928565979,
+      "learning_rate": 3.5964604871210793e-05,
+      "loss": 1.6462,
+      "step": 7312
+    },
+    {
+      "epoch": 1.9217656912011036,
+      "grad_norm": 0.6384910345077515,
+      "learning_rate": 3.594708253022604e-05,
+      "loss": 1.6279,
+      "step": 7314
+    },
+    {
+      "epoch": 1.9222911945347654,
+      "grad_norm": 0.5548694133758545,
+      "learning_rate": 3.592956018924128e-05,
+      "loss": 1.6488,
+      "step": 7316
+    },
+    {
+      "epoch": 1.9228166978684271,
+      "grad_norm": 0.5115952491760254,
+      "learning_rate": 3.591203784825653e-05,
+      "loss": 1.6557,
+      "step": 7318
+    },
+    {
+      "epoch": 1.9233422012020889,
+      "grad_norm": 0.564540684223175,
+      "learning_rate": 3.589451550727178e-05,
+      "loss": 1.6506,
+      "step": 7320
+    },
+    {
+      "epoch": 1.9238677045357506,
+      "grad_norm": 0.5424622297286987,
+      "learning_rate": 3.587699316628702e-05,
+      "loss": 1.6837,
+      "step": 7322
+    },
+    {
+      "epoch": 1.9243932078694124,
+      "grad_norm": 0.6576436758041382,
+      "learning_rate": 3.585947082530226e-05,
+      "loss": 1.6663,
+      "step": 7324
+    },
+    {
+      "epoch": 1.9249187112030741,
+      "grad_norm": 0.6056269407272339,
+      "learning_rate": 3.5841948484317506e-05,
+      "loss": 1.6427,
+      "step": 7326
+    },
+    {
+      "epoch": 1.925444214536736,
+      "grad_norm": 0.5156649947166443,
+      "learning_rate": 3.5824426143332754e-05,
+      "loss": 1.649,
+      "step": 7328
+    },
+    {
+      "epoch": 1.9259697178703976,
+      "grad_norm": 0.5065081119537354,
+      "learning_rate": 3.5806903802347994e-05,
+      "loss": 1.6422,
+      "step": 7330
+    },
+    {
+      "epoch": 1.9264952212040596,
+      "grad_norm": 0.5199773907661438,
+      "learning_rate": 3.578938146136324e-05,
+      "loss": 1.634,
+      "step": 7332
+    },
+    {
+      "epoch": 1.9270207245377213,
+      "grad_norm": 0.5321483016014099,
+      "learning_rate": 3.577185912037849e-05,
+      "loss": 1.6501,
+      "step": 7334
+    },
+    {
+      "epoch": 1.927546227871383,
+      "grad_norm": 0.5092171430587769,
+      "learning_rate": 3.5754336779393724e-05,
+      "loss": 1.6337,
+      "step": 7336
+    },
+    {
+      "epoch": 1.9280717312050448,
+      "grad_norm": 0.4993257224559784,
+      "learning_rate": 3.573681443840897e-05,
+      "loss": 1.647,
+      "step": 7338
+    },
+    {
+      "epoch": 1.9285972345387066,
+      "grad_norm": 0.7736045718193054,
+      "learning_rate": 3.571929209742422e-05,
+      "loss": 1.645,
+      "step": 7340
+    },
+    {
+      "epoch": 1.9291227378723685,
+      "grad_norm": 0.5889518857002258,
+      "learning_rate": 3.570176975643946e-05,
+      "loss": 1.6412,
+      "step": 7342
+    },
+    {
+      "epoch": 1.92964824120603,
+      "grad_norm": 0.5440062880516052,
+      "learning_rate": 3.568424741545471e-05,
+      "loss": 1.6537,
+      "step": 7344
+    },
+    {
+      "epoch": 1.930173744539692,
+      "grad_norm": 0.5857016444206238,
+      "learning_rate": 3.5666725074469955e-05,
+      "loss": 1.6627,
+      "step": 7346
+    },
+    {
+      "epoch": 1.9306992478733536,
+      "grad_norm": 0.5349956750869751,
+      "learning_rate": 3.5649202733485195e-05,
+      "loss": 1.6621,
+      "step": 7348
+    },
+    {
+      "epoch": 1.9312247512070155,
+      "grad_norm": 0.6116125583648682,
+      "learning_rate": 3.5631680392500436e-05,
+      "loss": 1.6555,
+      "step": 7350
+    },
+    {
+      "epoch": 1.9317502545406773,
+      "grad_norm": 0.5522750020027161,
+      "learning_rate": 3.5614158051515684e-05,
+      "loss": 1.6551,
+      "step": 7352
+    },
+    {
+      "epoch": 1.932275757874339,
+      "grad_norm": 0.503121554851532,
+      "learning_rate": 3.559663571053093e-05,
+      "loss": 1.6105,
+      "step": 7354
+    },
+    {
+      "epoch": 1.9328012612080008,
+      "grad_norm": 0.7201147079467773,
+      "learning_rate": 3.557911336954617e-05,
+      "loss": 1.6763,
+      "step": 7356
+    },
+    {
+      "epoch": 1.9333267645416625,
+      "grad_norm": 0.507443904876709,
+      "learning_rate": 3.556159102856142e-05,
+      "loss": 1.6476,
+      "step": 7358
+    },
+    {
+      "epoch": 1.9338522678753245,
+      "grad_norm": 0.5974195003509521,
+      "learning_rate": 3.554406868757666e-05,
+      "loss": 1.6446,
+      "step": 7360
+    },
+    {
+      "epoch": 1.934377771208986,
+      "grad_norm": 0.5196303725242615,
+      "learning_rate": 3.552654634659191e-05,
+      "loss": 1.638,
+      "step": 7362
+    },
+    {
+      "epoch": 1.934903274542648,
+      "grad_norm": 0.5367706418037415,
+      "learning_rate": 3.550902400560715e-05,
+      "loss": 1.6712,
+      "step": 7364
+    },
+    {
+      "epoch": 1.9354287778763095,
+      "grad_norm": 0.5725764632225037,
+      "learning_rate": 3.5491501664622396e-05,
+      "loss": 1.6764,
+      "step": 7366
+    },
+    {
+      "epoch": 1.9359542812099715,
+      "grad_norm": 0.6095370054244995,
+      "learning_rate": 3.5473979323637644e-05,
+      "loss": 1.6385,
+      "step": 7368
+    },
+    {
+      "epoch": 1.9364797845436332,
+      "grad_norm": 0.6413910388946533,
+      "learning_rate": 3.5456456982652885e-05,
+      "loss": 1.6665,
+      "step": 7370
+    },
+    {
+      "epoch": 1.937005287877295,
+      "grad_norm": 0.5054334998130798,
+      "learning_rate": 3.543893464166813e-05,
+      "loss": 1.6639,
+      "step": 7372
+    },
+    {
+      "epoch": 1.9375307912109567,
+      "grad_norm": 0.5909721255302429,
+      "learning_rate": 3.542141230068337e-05,
+      "loss": 1.6658,
+      "step": 7374
+    },
+    {
+      "epoch": 1.9380562945446185,
+      "grad_norm": 0.5415735244750977,
+      "learning_rate": 3.5403889959698614e-05,
+      "loss": 1.6872,
+      "step": 7376
+    },
+    {
+      "epoch": 1.9385817978782804,
+      "grad_norm": 0.5185040831565857,
+      "learning_rate": 3.538636761871386e-05,
+      "loss": 1.6371,
+      "step": 7378
+    },
+    {
+      "epoch": 1.939107301211942,
+      "grad_norm": 0.4956663250923157,
+      "learning_rate": 3.536884527772911e-05,
+      "loss": 1.6036,
+      "step": 7380
+    },
+    {
+      "epoch": 1.939632804545604,
+      "grad_norm": 0.5410330891609192,
+      "learning_rate": 3.535132293674435e-05,
+      "loss": 1.6482,
+      "step": 7382
+    },
+    {
+      "epoch": 1.9401583078792655,
+      "grad_norm": 0.6475924253463745,
+      "learning_rate": 3.53338005957596e-05,
+      "loss": 1.6297,
+      "step": 7384
+    },
+    {
+      "epoch": 1.9406838112129274,
+      "grad_norm": 0.5302926301956177,
+      "learning_rate": 3.531627825477484e-05,
+      "loss": 1.6652,
+      "step": 7386
+    },
+    {
+      "epoch": 1.9412093145465892,
+      "grad_norm": 0.6099222898483276,
+      "learning_rate": 3.5298755913790086e-05,
+      "loss": 1.6463,
+      "step": 7388
+    },
+    {
+      "epoch": 1.941734817880251,
+      "grad_norm": 0.6674348711967468,
+      "learning_rate": 3.5281233572805326e-05,
+      "loss": 1.6421,
+      "step": 7390
+    },
+    {
+      "epoch": 1.9422603212139127,
+      "grad_norm": 0.5395660400390625,
+      "learning_rate": 3.5263711231820574e-05,
+      "loss": 1.6465,
+      "step": 7392
+    },
+    {
+      "epoch": 1.9427858245475744,
+      "grad_norm": 0.629122257232666,
+      "learning_rate": 3.524618889083582e-05,
+      "loss": 1.6914,
+      "step": 7394
+    },
+    {
+      "epoch": 1.9433113278812364,
+      "grad_norm": 0.5263876914978027,
+      "learning_rate": 3.522866654985106e-05,
+      "loss": 1.7023,
+      "step": 7396
+    },
+    {
+      "epoch": 1.943836831214898,
+      "grad_norm": 0.5672745108604431,
+      "learning_rate": 3.52111442088663e-05,
+      "loss": 1.652,
+      "step": 7398
+    },
+    {
+      "epoch": 1.94436233454856,
+      "grad_norm": 0.551514744758606,
+      "learning_rate": 3.519362186788155e-05,
+      "loss": 1.6294,
+      "step": 7400
+    },
+    {
+      "epoch": 1.9448878378822214,
+      "grad_norm": 0.593605101108551,
+      "learning_rate": 3.517609952689679e-05,
+      "loss": 1.6561,
+      "step": 7402
+    },
+    {
+      "epoch": 1.9454133412158834,
+      "grad_norm": 0.5020228028297424,
+      "learning_rate": 3.515857718591204e-05,
+      "loss": 1.6724,
+      "step": 7404
+    },
+    {
+      "epoch": 1.9459388445495451,
+      "grad_norm": 0.5694274306297302,
+      "learning_rate": 3.5141054844927287e-05,
+      "loss": 1.6771,
+      "step": 7406
+    },
+    {
+      "epoch": 1.9464643478832069,
+      "grad_norm": 0.6046936511993408,
+      "learning_rate": 3.512353250394253e-05,
+      "loss": 1.6627,
+      "step": 7408
+    },
+    {
+      "epoch": 1.9469898512168686,
+      "grad_norm": 0.5441648364067078,
+      "learning_rate": 3.5106010162957775e-05,
+      "loss": 1.638,
+      "step": 7410
+    },
+    {
+      "epoch": 1.9475153545505304,
+      "grad_norm": 0.5168318748474121,
+      "learning_rate": 3.5088487821973016e-05,
+      "loss": 1.6782,
+      "step": 7412
+    },
+    {
+      "epoch": 1.9480408578841923,
+      "grad_norm": 0.5330861210823059,
+      "learning_rate": 3.507096548098826e-05,
+      "loss": 1.6427,
+      "step": 7414
+    },
+    {
+      "epoch": 1.9485663612178539,
+      "grad_norm": 0.6046448945999146,
+      "learning_rate": 3.5053443140003504e-05,
+      "loss": 1.644,
+      "step": 7416
+    },
+    {
+      "epoch": 1.9490918645515158,
+      "grad_norm": 0.5944949388504028,
+      "learning_rate": 3.503592079901875e-05,
+      "loss": 1.6684,
+      "step": 7418
+    },
+    {
+      "epoch": 1.9496173678851774,
+      "grad_norm": 0.5597866773605347,
+      "learning_rate": 3.5018398458034e-05,
+      "loss": 1.646,
+      "step": 7420
+    },
+    {
+      "epoch": 1.9501428712188393,
+      "grad_norm": 0.6573653817176819,
+      "learning_rate": 3.500087611704924e-05,
+      "loss": 1.6601,
+      "step": 7422
+    },
+    {
+      "epoch": 1.950668374552501,
+      "grad_norm": 0.5286207795143127,
+      "learning_rate": 3.498335377606448e-05,
+      "loss": 1.6259,
+      "step": 7424
+    },
+    {
+      "epoch": 1.9511938778861628,
+      "grad_norm": 0.5400940179824829,
+      "learning_rate": 3.496583143507973e-05,
+      "loss": 1.6581,
+      "step": 7426
+    },
+    {
+      "epoch": 1.9517193812198246,
+      "grad_norm": 0.6813338994979858,
+      "learning_rate": 3.494830909409497e-05,
+      "loss": 1.6552,
+      "step": 7428
+    },
+    {
+      "epoch": 1.9522448845534863,
+      "grad_norm": 0.5815770030021667,
+      "learning_rate": 3.493078675311022e-05,
+      "loss": 1.6845,
+      "step": 7430
+    },
+    {
+      "epoch": 1.9527703878871483,
+      "grad_norm": 0.5522500872612,
+      "learning_rate": 3.4913264412125464e-05,
+      "loss": 1.6749,
+      "step": 7432
+    },
+    {
+      "epoch": 1.9532958912208098,
+      "grad_norm": 0.5736134052276611,
+      "learning_rate": 3.4895742071140705e-05,
+      "loss": 1.6413,
+      "step": 7434
+    },
+    {
+      "epoch": 1.9538213945544718,
+      "grad_norm": 0.4932587742805481,
+      "learning_rate": 3.487821973015595e-05,
+      "loss": 1.6416,
+      "step": 7436
+    },
+    {
+      "epoch": 1.9543468978881333,
+      "grad_norm": 0.5147337913513184,
+      "learning_rate": 3.486069738917119e-05,
+      "loss": 1.6626,
+      "step": 7438
+    },
+    {
+      "epoch": 1.9548724012217953,
+      "grad_norm": 0.7022672891616821,
+      "learning_rate": 3.484317504818644e-05,
+      "loss": 1.6723,
+      "step": 7440
+    },
+    {
+      "epoch": 1.955397904555457,
+      "grad_norm": 0.6867913007736206,
+      "learning_rate": 3.482565270720168e-05,
+      "loss": 1.6431,
+      "step": 7442
+    },
+    {
+      "epoch": 1.9559234078891188,
+      "grad_norm": 0.5425695180892944,
+      "learning_rate": 3.480813036621693e-05,
+      "loss": 1.6566,
+      "step": 7444
+    },
+    {
+      "epoch": 1.9564489112227805,
+      "grad_norm": 0.6468666195869446,
+      "learning_rate": 3.479060802523218e-05,
+      "loss": 1.6638,
+      "step": 7446
+    },
+    {
+      "epoch": 1.9569744145564423,
+      "grad_norm": 0.5925945043563843,
+      "learning_rate": 3.477308568424742e-05,
+      "loss": 1.6443,
+      "step": 7448
+    },
+    {
+      "epoch": 1.9574999178901042,
+      "grad_norm": 0.6654757857322693,
+      "learning_rate": 3.475556334326266e-05,
+      "loss": 1.6064,
+      "step": 7450
+    },
+    {
+      "epoch": 1.9580254212237658,
+      "grad_norm": 0.5577012300491333,
+      "learning_rate": 3.4738041002277906e-05,
+      "loss": 1.6556,
+      "step": 7452
+    },
+    {
+      "epoch": 1.9585509245574277,
+      "grad_norm": 0.7134981751441956,
+      "learning_rate": 3.472051866129315e-05,
+      "loss": 1.6739,
+      "step": 7454
+    },
+    {
+      "epoch": 1.9590764278910895,
+      "grad_norm": 0.5233617424964905,
+      "learning_rate": 3.4702996320308394e-05,
+      "loss": 1.6466,
+      "step": 7456
+    },
+    {
+      "epoch": 1.9596019312247512,
+      "grad_norm": 0.5530445575714111,
+      "learning_rate": 3.468547397932364e-05,
+      "loss": 1.6994,
+      "step": 7458
+    },
+    {
+      "epoch": 1.960127434558413,
+      "grad_norm": 0.5197651982307434,
+      "learning_rate": 3.466795163833888e-05,
+      "loss": 1.6614,
+      "step": 7460
+    },
+    {
+      "epoch": 1.9606529378920747,
+      "grad_norm": 0.48130765557289124,
+      "learning_rate": 3.4650429297354123e-05,
+      "loss": 1.6449,
+      "step": 7462
+    },
+    {
+      "epoch": 1.9611784412257365,
+      "grad_norm": 0.49137425422668457,
+      "learning_rate": 3.463290695636937e-05,
+      "loss": 1.6527,
+      "step": 7464
+    },
+    {
+      "epoch": 1.9617039445593982,
+      "grad_norm": 0.5547472238540649,
+      "learning_rate": 3.461538461538462e-05,
+      "loss": 1.6491,
+      "step": 7466
+    },
+    {
+      "epoch": 1.9622294478930602,
+      "grad_norm": 0.5760181546211243,
+      "learning_rate": 3.459786227439986e-05,
+      "loss": 1.6794,
+      "step": 7468
+    },
+    {
+      "epoch": 1.9627549512267217,
+      "grad_norm": 0.5289864540100098,
+      "learning_rate": 3.458033993341511e-05,
+      "loss": 1.6548,
+      "step": 7470
+    },
+    {
+      "epoch": 1.9632804545603837,
+      "grad_norm": 0.5183879733085632,
+      "learning_rate": 3.4562817592430355e-05,
+      "loss": 1.6629,
+      "step": 7472
+    },
+    {
+      "epoch": 1.9638059578940454,
+      "grad_norm": 0.5636990666389465,
+      "learning_rate": 3.4545295251445595e-05,
+      "loss": 1.6591,
+      "step": 7474
+    },
+    {
+      "epoch": 1.9643314612277072,
+      "grad_norm": 0.5583269596099854,
+      "learning_rate": 3.4527772910460836e-05,
+      "loss": 1.6576,
+      "step": 7476
+    },
+    {
+      "epoch": 1.964856964561369,
+      "grad_norm": 0.5348683595657349,
+      "learning_rate": 3.4510250569476084e-05,
+      "loss": 1.6148,
+      "step": 7478
+    },
+    {
+      "epoch": 1.9653824678950307,
+      "grad_norm": 0.5336459875106812,
+      "learning_rate": 3.4492728228491324e-05,
+      "loss": 1.6704,
+      "step": 7480
+    },
+    {
+      "epoch": 1.9659079712286924,
+      "grad_norm": 0.5364224314689636,
+      "learning_rate": 3.447520588750657e-05,
+      "loss": 1.6391,
+      "step": 7482
+    },
+    {
+      "epoch": 1.9664334745623542,
+      "grad_norm": 0.5437600612640381,
+      "learning_rate": 3.445768354652182e-05,
+      "loss": 1.6817,
+      "step": 7484
+    },
+    {
+      "epoch": 1.9669589778960161,
+      "grad_norm": 0.6408882737159729,
+      "learning_rate": 3.444016120553706e-05,
+      "loss": 1.6505,
+      "step": 7486
+    },
+    {
+      "epoch": 1.9674844812296777,
+      "grad_norm": 0.5990426540374756,
+      "learning_rate": 3.44226388645523e-05,
+      "loss": 1.6634,
+      "step": 7488
+    },
+    {
+      "epoch": 1.9680099845633396,
+      "grad_norm": 0.5437130928039551,
+      "learning_rate": 3.440511652356755e-05,
+      "loss": 1.6716,
+      "step": 7490
+    },
+    {
+      "epoch": 1.9685354878970014,
+      "grad_norm": 0.6260076761245728,
+      "learning_rate": 3.4387594182582796e-05,
+      "loss": 1.6569,
+      "step": 7492
+    },
+    {
+      "epoch": 1.9690609912306631,
+      "grad_norm": 0.5225280523300171,
+      "learning_rate": 3.437007184159804e-05,
+      "loss": 1.6649,
+      "step": 7494
+    },
+    {
+      "epoch": 1.9695864945643249,
+      "grad_norm": 0.5389422178268433,
+      "learning_rate": 3.4352549500613285e-05,
+      "loss": 1.6723,
+      "step": 7496
+    },
+    {
+      "epoch": 1.9701119978979866,
+      "grad_norm": 0.6174726486206055,
+      "learning_rate": 3.433502715962853e-05,
+      "loss": 1.669,
+      "step": 7498
+    },
+    {
+      "epoch": 1.9706375012316486,
+      "grad_norm": 0.5065149664878845,
+      "learning_rate": 3.431750481864377e-05,
+      "loss": 1.6638,
+      "step": 7500
+    },
+    {
+      "epoch": 1.9711630045653101,
+      "grad_norm": 0.6950640678405762,
+      "learning_rate": 3.4299982477659014e-05,
+      "loss": 1.6713,
+      "step": 7502
+    },
+    {
+      "epoch": 1.971688507898972,
+      "grad_norm": 0.507963240146637,
+      "learning_rate": 3.428246013667426e-05,
+      "loss": 1.6332,
+      "step": 7504
+    },
+    {
+      "epoch": 1.9722140112326336,
+      "grad_norm": 0.6077314019203186,
+      "learning_rate": 3.426493779568951e-05,
+      "loss": 1.6641,
+      "step": 7506
+    },
+    {
+      "epoch": 1.9727395145662956,
+      "grad_norm": 0.5328965783119202,
+      "learning_rate": 3.424741545470475e-05,
+      "loss": 1.6593,
+      "step": 7508
+    },
+    {
+      "epoch": 1.9732650178999573,
+      "grad_norm": 0.6714265942573547,
+      "learning_rate": 3.422989311372e-05,
+      "loss": 1.6444,
+      "step": 7510
+    },
+    {
+      "epoch": 1.973790521233619,
+      "grad_norm": 0.5329544544219971,
+      "learning_rate": 3.4212370772735245e-05,
+      "loss": 1.6337,
+      "step": 7512
+    },
+    {
+      "epoch": 1.9743160245672808,
+      "grad_norm": 0.5360748767852783,
+      "learning_rate": 3.419484843175048e-05,
+      "loss": 1.6535,
+      "step": 7514
+    },
+    {
+      "epoch": 1.9748415279009426,
+      "grad_norm": 0.5412916541099548,
+      "learning_rate": 3.4177326090765726e-05,
+      "loss": 1.6504,
+      "step": 7516
+    },
+    {
+      "epoch": 1.9753670312346046,
+      "grad_norm": 0.6592147946357727,
+      "learning_rate": 3.4159803749780974e-05,
+      "loss": 1.6766,
+      "step": 7518
+    },
+    {
+      "epoch": 1.975892534568266,
+      "grad_norm": 0.5632504820823669,
+      "learning_rate": 3.4142281408796215e-05,
+      "loss": 1.6587,
+      "step": 7520
+    },
+    {
+      "epoch": 1.976418037901928,
+      "grad_norm": 0.5235944390296936,
+      "learning_rate": 3.412475906781146e-05,
+      "loss": 1.7091,
+      "step": 7522
+    },
+    {
+      "epoch": 1.9769435412355896,
+      "grad_norm": 0.7552381753921509,
+      "learning_rate": 3.410723672682671e-05,
+      "loss": 1.6765,
+      "step": 7524
+    },
+    {
+      "epoch": 1.9774690445692515,
+      "grad_norm": 0.5463106036186218,
+      "learning_rate": 3.408971438584195e-05,
+      "loss": 1.7028,
+      "step": 7526
+    },
+    {
+      "epoch": 1.9779945479029133,
+      "grad_norm": 0.5811282396316528,
+      "learning_rate": 3.407219204485719e-05,
+      "loss": 1.626,
+      "step": 7528
+    },
+    {
+      "epoch": 1.978520051236575,
+      "grad_norm": 0.5381944179534912,
+      "learning_rate": 3.405466970387244e-05,
+      "loss": 1.6557,
+      "step": 7530
+    },
+    {
+      "epoch": 1.9790455545702368,
+      "grad_norm": 0.5382105112075806,
+      "learning_rate": 3.4037147362887687e-05,
+      "loss": 1.6717,
+      "step": 7532
+    },
+    {
+      "epoch": 1.9795710579038985,
+      "grad_norm": 0.5555400252342224,
+      "learning_rate": 3.401962502190293e-05,
+      "loss": 1.6546,
+      "step": 7534
+    },
+    {
+      "epoch": 1.9800965612375605,
+      "grad_norm": 0.5068405270576477,
+      "learning_rate": 3.4002102680918175e-05,
+      "loss": 1.6378,
+      "step": 7536
+    },
+    {
+      "epoch": 1.980622064571222,
+      "grad_norm": 0.5872784852981567,
+      "learning_rate": 3.398458033993342e-05,
+      "loss": 1.6615,
+      "step": 7538
+    },
+    {
+      "epoch": 1.981147567904884,
+      "grad_norm": 0.5751393437385559,
+      "learning_rate": 3.3967057998948656e-05,
+      "loss": 1.6712,
+      "step": 7540
+    },
+    {
+      "epoch": 1.9816730712385455,
+      "grad_norm": 0.5746015310287476,
+      "learning_rate": 3.3949535657963904e-05,
+      "loss": 1.667,
+      "step": 7542
+    },
+    {
+      "epoch": 1.9821985745722075,
+      "grad_norm": 0.6387732028961182,
+      "learning_rate": 3.393201331697915e-05,
+      "loss": 1.6253,
+      "step": 7544
+    },
+    {
+      "epoch": 1.9827240779058692,
+      "grad_norm": 0.5271580815315247,
+      "learning_rate": 3.391449097599439e-05,
+      "loss": 1.6442,
+      "step": 7546
+    },
+    {
+      "epoch": 1.983249581239531,
+      "grad_norm": 0.5445979237556458,
+      "learning_rate": 3.389696863500964e-05,
+      "loss": 1.6363,
+      "step": 7548
+    },
+    {
+      "epoch": 1.9837750845731927,
+      "grad_norm": 0.6095319986343384,
+      "learning_rate": 3.387944629402489e-05,
+      "loss": 1.6486,
+      "step": 7550
+    },
+    {
+      "epoch": 1.9843005879068545,
+      "grad_norm": 0.6398605704307556,
+      "learning_rate": 3.386192395304013e-05,
+      "loss": 1.6427,
+      "step": 7552
+    },
+    {
+      "epoch": 1.9848260912405165,
+      "grad_norm": 0.6042175889015198,
+      "learning_rate": 3.384440161205537e-05,
+      "loss": 1.6522,
+      "step": 7554
+    },
+    {
+      "epoch": 1.985351594574178,
+      "grad_norm": 0.550961971282959,
+      "learning_rate": 3.382687927107062e-05,
+      "loss": 1.7006,
+      "step": 7556
+    },
+    {
+      "epoch": 1.98587709790784,
+      "grad_norm": 0.5935929417610168,
+      "learning_rate": 3.3809356930085864e-05,
+      "loss": 1.6537,
+      "step": 7558
+    },
+    {
+      "epoch": 1.9864026012415015,
+      "grad_norm": 0.6039239764213562,
+      "learning_rate": 3.3791834589101105e-05,
+      "loss": 1.6623,
+      "step": 7560
+    },
+    {
+      "epoch": 1.9869281045751634,
+      "grad_norm": 0.5280980467796326,
+      "learning_rate": 3.377431224811635e-05,
+      "loss": 1.6695,
+      "step": 7562
+    },
+    {
+      "epoch": 1.9874536079088252,
+      "grad_norm": 0.5845438838005066,
+      "learning_rate": 3.375678990713159e-05,
+      "loss": 1.6387,
+      "step": 7564
+    },
+    {
+      "epoch": 1.987979111242487,
+      "grad_norm": 0.5099235773086548,
+      "learning_rate": 3.3739267566146834e-05,
+      "loss": 1.644,
+      "step": 7566
+    },
+    {
+      "epoch": 1.9885046145761487,
+      "grad_norm": 0.6559625864028931,
+      "learning_rate": 3.372174522516208e-05,
+      "loss": 1.6683,
+      "step": 7568
+    },
+    {
+      "epoch": 1.9890301179098104,
+      "grad_norm": 0.5141841173171997,
+      "learning_rate": 3.370422288417733e-05,
+      "loss": 1.6487,
+      "step": 7570
+    },
+    {
+      "epoch": 1.9895556212434724,
+      "grad_norm": 0.5415523052215576,
+      "learning_rate": 3.368670054319257e-05,
+      "loss": 1.6654,
+      "step": 7572
+    },
+    {
+      "epoch": 1.990081124577134,
+      "grad_norm": 0.5395537614822388,
+      "learning_rate": 3.366917820220782e-05,
+      "loss": 1.6397,
+      "step": 7574
+    },
+    {
+      "epoch": 1.990606627910796,
+      "grad_norm": 0.7312270402908325,
+      "learning_rate": 3.3651655861223065e-05,
+      "loss": 1.628,
+      "step": 7576
+    },
+    {
+      "epoch": 1.9911321312444574,
+      "grad_norm": 0.7088197469711304,
+      "learning_rate": 3.3634133520238306e-05,
+      "loss": 1.6725,
+      "step": 7578
+    },
+    {
+      "epoch": 1.9916576345781194,
+      "grad_norm": 0.5116075277328491,
+      "learning_rate": 3.361661117925355e-05,
+      "loss": 1.6307,
+      "step": 7580
+    },
+    {
+      "epoch": 1.9921831379117811,
+      "grad_norm": 0.5469351410865784,
+      "learning_rate": 3.3599088838268794e-05,
+      "loss": 1.6465,
+      "step": 7582
+    },
+    {
+      "epoch": 1.992708641245443,
+      "grad_norm": 0.5921772718429565,
+      "learning_rate": 3.358156649728404e-05,
+      "loss": 1.6809,
+      "step": 7584
+    },
+    {
+      "epoch": 1.9932341445791046,
+      "grad_norm": 0.5668914318084717,
+      "learning_rate": 3.356404415629928e-05,
+      "loss": 1.639,
+      "step": 7586
+    },
+    {
+      "epoch": 1.9937596479127664,
+      "grad_norm": 0.5449174642562866,
+      "learning_rate": 3.354652181531453e-05,
+      "loss": 1.6501,
+      "step": 7588
+    },
+    {
+      "epoch": 1.9942851512464284,
+      "grad_norm": 0.6408663988113403,
+      "learning_rate": 3.352899947432977e-05,
+      "loss": 1.6949,
+      "step": 7590
+    },
+    {
+      "epoch": 1.9948106545800899,
+      "grad_norm": 0.6358173489570618,
+      "learning_rate": 3.351147713334501e-05,
+      "loss": 1.6808,
+      "step": 7592
+    },
+    {
+      "epoch": 1.9953361579137519,
+      "grad_norm": 0.5143889784812927,
+      "learning_rate": 3.349395479236026e-05,
+      "loss": 1.6659,
+      "step": 7594
+    },
+    {
+      "epoch": 1.9958616612474134,
+      "grad_norm": 0.5499483346939087,
+      "learning_rate": 3.347643245137551e-05,
+      "loss": 1.6649,
+      "step": 7596
+    },
+    {
+      "epoch": 1.9963871645810753,
+      "grad_norm": 0.5296257138252258,
+      "learning_rate": 3.345891011039075e-05,
+      "loss": 1.6784,
+      "step": 7598
+    },
+    {
+      "epoch": 1.996912667914737,
+      "grad_norm": 0.6414445638656616,
+      "learning_rate": 3.3441387769405995e-05,
+      "loss": 1.6672,
+      "step": 7600
+    },
+    {
+      "epoch": 1.996912667914737,
+      "eval_loss": 1.6618915796279907,
+      "eval_runtime": 487.2163,
+      "eval_samples_per_second": 249.969,
+      "eval_steps_per_second": 31.247,
+      "step": 7600
+    },
+    {
+      "epoch": 1.9974381712483988,
+      "grad_norm": 0.5835072994232178,
+      "learning_rate": 3.3423865428421236e-05,
+      "loss": 1.6681,
+      "step": 7602
+    },
+    {
+      "epoch": 1.9979636745820606,
+      "grad_norm": 0.5567203164100647,
+      "learning_rate": 3.3406343087436484e-05,
+      "loss": 1.6592,
+      "step": 7604
+    },
+    {
+      "epoch": 1.9984891779157223,
+      "grad_norm": 0.6631194949150085,
+      "learning_rate": 3.3388820746451724e-05,
+      "loss": 1.6551,
+      "step": 7606
+    },
+    {
+      "epoch": 1.9990146812493843,
+      "grad_norm": 0.6589709520339966,
+      "learning_rate": 3.337129840546697e-05,
+      "loss": 1.6358,
+      "step": 7608
+    },
+    {
+      "epoch": 1.9995401845830458,
+      "grad_norm": 0.5504697561264038,
+      "learning_rate": 3.335377606448222e-05,
+      "loss": 1.6477,
+      "step": 7610
+    },
+    {
+      "epoch": 2.000065687916708,
+      "grad_norm": 0.4931085705757141,
+      "learning_rate": 3.333625372349746e-05,
+      "loss": 1.6487,
+      "step": 7612
+    },
+    {
+      "epoch": 2.0005911912503693,
+      "grad_norm": 0.6642125844955444,
+      "learning_rate": 3.331873138251271e-05,
+      "loss": 1.6306,
+      "step": 7614
+    },
+    {
+      "epoch": 2.0011166945840313,
+      "grad_norm": 0.5301389098167419,
+      "learning_rate": 3.330120904152795e-05,
+      "loss": 1.574,
+      "step": 7616
+    },
+    {
+      "epoch": 2.001642197917693,
+      "grad_norm": 0.5445417761802673,
+      "learning_rate": 3.3283686700543196e-05,
+      "loss": 1.6205,
+      "step": 7618
+    },
+    {
+      "epoch": 2.002167701251355,
+      "grad_norm": 0.5667280554771423,
+      "learning_rate": 3.326616435955844e-05,
+      "loss": 1.6149,
+      "step": 7620
+    },
+    {
+      "epoch": 2.0026932045850168,
+      "grad_norm": 0.5716310739517212,
+      "learning_rate": 3.3248642018573685e-05,
+      "loss": 1.6333,
+      "step": 7622
+    },
+    {
+      "epoch": 2.0032187079186783,
+      "grad_norm": 0.7009231448173523,
+      "learning_rate": 3.323111967758893e-05,
+      "loss": 1.6281,
+      "step": 7624
+    },
+    {
+      "epoch": 2.0037442112523403,
+      "grad_norm": 0.5482553839683533,
+      "learning_rate": 3.321359733660417e-05,
+      "loss": 1.6054,
+      "step": 7626
+    },
+    {
+      "epoch": 2.004269714586002,
+      "grad_norm": 0.48852571845054626,
+      "learning_rate": 3.3196074995619414e-05,
+      "loss": 1.636,
+      "step": 7628
+    },
+    {
+      "epoch": 2.0047952179196638,
+      "grad_norm": 0.5163108110427856,
+      "learning_rate": 3.317855265463466e-05,
+      "loss": 1.6036,
+      "step": 7630
+    },
+    {
+      "epoch": 2.0053207212533253,
+      "grad_norm": 0.5240321755409241,
+      "learning_rate": 3.31610303136499e-05,
+      "loss": 1.5722,
+      "step": 7632
+    },
+    {
+      "epoch": 2.0058462245869872,
+      "grad_norm": 0.5390346646308899,
+      "learning_rate": 3.314350797266515e-05,
+      "loss": 1.5939,
+      "step": 7634
+    },
+    {
+      "epoch": 2.006371727920649,
+      "grad_norm": 0.5812144875526428,
+      "learning_rate": 3.31259856316804e-05,
+      "loss": 1.6445,
+      "step": 7636
+    },
+    {
+      "epoch": 2.0068972312543107,
+      "grad_norm": 0.6792988777160645,
+      "learning_rate": 3.310846329069564e-05,
+      "loss": 1.6003,
+      "step": 7638
+    },
+    {
+      "epoch": 2.0074227345879727,
+      "grad_norm": 0.6539336442947388,
+      "learning_rate": 3.309094094971088e-05,
+      "loss": 1.632,
+      "step": 7640
+    },
+    {
+      "epoch": 2.0079482379216342,
+      "grad_norm": 0.6297439336776733,
+      "learning_rate": 3.3073418608726126e-05,
+      "loss": 1.6148,
+      "step": 7642
+    },
+    {
+      "epoch": 2.008473741255296,
+      "grad_norm": 0.6377638578414917,
+      "learning_rate": 3.3055896267741374e-05,
+      "loss": 1.6097,
+      "step": 7644
+    },
+    {
+      "epoch": 2.0089992445889577,
+      "grad_norm": 0.5924256443977356,
+      "learning_rate": 3.3038373926756615e-05,
+      "loss": 1.6239,
+      "step": 7646
+    },
+    {
+      "epoch": 2.0095247479226197,
+      "grad_norm": 0.541309118270874,
+      "learning_rate": 3.302085158577186e-05,
+      "loss": 1.6145,
+      "step": 7648
+    },
+    {
+      "epoch": 2.0100502512562812,
+      "grad_norm": 0.708676278591156,
+      "learning_rate": 3.300332924478711e-05,
+      "loss": 1.6037,
+      "step": 7650
+    },
+    {
+      "epoch": 2.010575754589943,
+      "grad_norm": 0.6031850576400757,
+      "learning_rate": 3.298580690380235e-05,
+      "loss": 1.6071,
+      "step": 7652
+    },
+    {
+      "epoch": 2.011101257923605,
+      "grad_norm": 0.6296379566192627,
+      "learning_rate": 3.296828456281759e-05,
+      "loss": 1.6348,
+      "step": 7654
+    },
+    {
+      "epoch": 2.0116267612572667,
+      "grad_norm": 0.5863556861877441,
+      "learning_rate": 3.295076222183284e-05,
+      "loss": 1.6446,
+      "step": 7656
+    },
+    {
+      "epoch": 2.0121522645909287,
+      "grad_norm": 0.5582921504974365,
+      "learning_rate": 3.293323988084808e-05,
+      "loss": 1.62,
+      "step": 7658
+    },
+    {
+      "epoch": 2.01267776792459,
+      "grad_norm": 0.5858394503593445,
+      "learning_rate": 3.291571753986333e-05,
+      "loss": 1.5987,
+      "step": 7660
+    },
+    {
+      "epoch": 2.013203271258252,
+      "grad_norm": 0.6063563227653503,
+      "learning_rate": 3.2898195198878575e-05,
+      "loss": 1.6469,
+      "step": 7662
+    },
+    {
+      "epoch": 2.0137287745919137,
+      "grad_norm": 0.536972165107727,
+      "learning_rate": 3.2880672857893816e-05,
+      "loss": 1.6042,
+      "step": 7664
+    },
+    {
+      "epoch": 2.0142542779255757,
+      "grad_norm": 0.6241976618766785,
+      "learning_rate": 3.2863150516909056e-05,
+      "loss": 1.5904,
+      "step": 7666
+    },
+    {
+      "epoch": 2.014779781259237,
+      "grad_norm": 0.5789350271224976,
+      "learning_rate": 3.2845628175924304e-05,
+      "loss": 1.5959,
+      "step": 7668
+    },
+    {
+      "epoch": 2.015305284592899,
+      "grad_norm": 0.6051028370857239,
+      "learning_rate": 3.282810583493955e-05,
+      "loss": 1.628,
+      "step": 7670
+    },
+    {
+      "epoch": 2.015830787926561,
+      "grad_norm": 0.658258855342865,
+      "learning_rate": 3.281058349395479e-05,
+      "loss": 1.6064,
+      "step": 7672
+    },
+    {
+      "epoch": 2.0163562912602226,
+      "grad_norm": 0.5936440825462341,
+      "learning_rate": 3.279306115297004e-05,
+      "loss": 1.6102,
+      "step": 7674
+    },
+    {
+      "epoch": 2.0168817945938846,
+      "grad_norm": 0.5593782663345337,
+      "learning_rate": 3.277553881198529e-05,
+      "loss": 1.6103,
+      "step": 7676
+    },
+    {
+      "epoch": 2.017407297927546,
+      "grad_norm": 0.5200657248497009,
+      "learning_rate": 3.275801647100053e-05,
+      "loss": 1.5972,
+      "step": 7678
+    },
+    {
+      "epoch": 2.017932801261208,
+      "grad_norm": 0.6176472902297974,
+      "learning_rate": 3.274049413001577e-05,
+      "loss": 1.5623,
+      "step": 7680
+    },
+    {
+      "epoch": 2.0184583045948696,
+      "grad_norm": 0.7342426776885986,
+      "learning_rate": 3.272297178903102e-05,
+      "loss": 1.6424,
+      "step": 7682
+    },
+    {
+      "epoch": 2.0189838079285316,
+      "grad_norm": 0.5766749978065491,
+      "learning_rate": 3.270544944804626e-05,
+      "loss": 1.6093,
+      "step": 7684
+    },
+    {
+      "epoch": 2.019509311262193,
+      "grad_norm": 0.5769773125648499,
+      "learning_rate": 3.2687927107061505e-05,
+      "loss": 1.6372,
+      "step": 7686
+    },
+    {
+      "epoch": 2.020034814595855,
+      "grad_norm": 0.6358572244644165,
+      "learning_rate": 3.267040476607675e-05,
+      "loss": 1.6274,
+      "step": 7688
+    },
+    {
+      "epoch": 2.020560317929517,
+      "grad_norm": 0.5427133440971375,
+      "learning_rate": 3.265288242509199e-05,
+      "loss": 1.594,
+      "step": 7690
+    },
+    {
+      "epoch": 2.0210858212631786,
+      "grad_norm": 0.5329941511154175,
+      "learning_rate": 3.2635360084107234e-05,
+      "loss": 1.606,
+      "step": 7692
+    },
+    {
+      "epoch": 2.0216113245968406,
+      "grad_norm": 0.5538175106048584,
+      "learning_rate": 3.261783774312248e-05,
+      "loss": 1.5938,
+      "step": 7694
+    },
+    {
+      "epoch": 2.022136827930502,
+      "grad_norm": 0.5610730051994324,
+      "learning_rate": 3.260031540213773e-05,
+      "loss": 1.604,
+      "step": 7696
+    },
+    {
+      "epoch": 2.022662331264164,
+      "grad_norm": 0.6616785526275635,
+      "learning_rate": 3.258279306115297e-05,
+      "loss": 1.6107,
+      "step": 7698
+    },
+    {
+      "epoch": 2.0231878345978256,
+      "grad_norm": 0.5210354924201965,
+      "learning_rate": 3.256527072016822e-05,
+      "loss": 1.6338,
+      "step": 7700
+    },
+    {
+      "epoch": 2.0237133379314876,
+      "grad_norm": 0.7099502682685852,
+      "learning_rate": 3.2547748379183465e-05,
+      "loss": 1.632,
+      "step": 7702
+    },
+    {
+      "epoch": 2.024238841265149,
+      "grad_norm": 0.5911880731582642,
+      "learning_rate": 3.25302260381987e-05,
+      "loss": 1.5943,
+      "step": 7704
+    },
+    {
+      "epoch": 2.024764344598811,
+      "grad_norm": 0.5507931709289551,
+      "learning_rate": 3.251270369721395e-05,
+      "loss": 1.605,
+      "step": 7706
+    },
+    {
+      "epoch": 2.025289847932473,
+      "grad_norm": 0.5524379014968872,
+      "learning_rate": 3.2495181356229194e-05,
+      "loss": 1.5664,
+      "step": 7708
+    },
+    {
+      "epoch": 2.0258153512661345,
+      "grad_norm": 0.5344357490539551,
+      "learning_rate": 3.2477659015244435e-05,
+      "loss": 1.6136,
+      "step": 7710
+    },
+    {
+      "epoch": 2.0263408545997965,
+      "grad_norm": 0.6390090584754944,
+      "learning_rate": 3.246013667425968e-05,
+      "loss": 1.6033,
+      "step": 7712
+    },
+    {
+      "epoch": 2.026866357933458,
+      "grad_norm": 0.5854797959327698,
+      "learning_rate": 3.244261433327493e-05,
+      "loss": 1.5812,
+      "step": 7714
+    },
+    {
+      "epoch": 2.02739186126712,
+      "grad_norm": 0.6743679642677307,
+      "learning_rate": 3.242509199229017e-05,
+      "loss": 1.577,
+      "step": 7716
+    },
+    {
+      "epoch": 2.0279173646007815,
+      "grad_norm": 0.6346087455749512,
+      "learning_rate": 3.240756965130541e-05,
+      "loss": 1.6094,
+      "step": 7718
+    },
+    {
+      "epoch": 2.0284428679344435,
+      "grad_norm": 0.6077213883399963,
+      "learning_rate": 3.239004731032066e-05,
+      "loss": 1.6264,
+      "step": 7720
+    },
+    {
+      "epoch": 2.028968371268105,
+      "grad_norm": 0.6114009618759155,
+      "learning_rate": 3.237252496933591e-05,
+      "loss": 1.6323,
+      "step": 7722
+    },
+    {
+      "epoch": 2.029493874601767,
+      "grad_norm": 0.5229764580726624,
+      "learning_rate": 3.235500262835115e-05,
+      "loss": 1.6128,
+      "step": 7724
+    },
+    {
+      "epoch": 2.030019377935429,
+      "grad_norm": 0.7623997330665588,
+      "learning_rate": 3.2337480287366395e-05,
+      "loss": 1.5912,
+      "step": 7726
+    },
+    {
+      "epoch": 2.0305448812690905,
+      "grad_norm": 0.7716650366783142,
+      "learning_rate": 3.231995794638164e-05,
+      "loss": 1.6001,
+      "step": 7728
+    },
+    {
+      "epoch": 2.0310703846027525,
+      "grad_norm": 0.5649428963661194,
+      "learning_rate": 3.230243560539688e-05,
+      "loss": 1.6171,
+      "step": 7730
+    },
+    {
+      "epoch": 2.031595887936414,
+      "grad_norm": 0.689694344997406,
+      "learning_rate": 3.2284913264412124e-05,
+      "loss": 1.6245,
+      "step": 7732
+    },
+    {
+      "epoch": 2.032121391270076,
+      "grad_norm": 0.6371480226516724,
+      "learning_rate": 3.226739092342737e-05,
+      "loss": 1.6262,
+      "step": 7734
+    },
+    {
+      "epoch": 2.0326468946037375,
+      "grad_norm": 0.5378861427307129,
+      "learning_rate": 3.224986858244261e-05,
+      "loss": 1.6057,
+      "step": 7736
+    },
+    {
+      "epoch": 2.0331723979373995,
+      "grad_norm": 0.633033037185669,
+      "learning_rate": 3.223234624145786e-05,
+      "loss": 1.6442,
+      "step": 7738
+    },
+    {
+      "epoch": 2.033697901271061,
+      "grad_norm": 0.6772140860557556,
+      "learning_rate": 3.221482390047311e-05,
+      "loss": 1.5639,
+      "step": 7740
+    },
+    {
+      "epoch": 2.034223404604723,
+      "grad_norm": 0.5355550050735474,
+      "learning_rate": 3.219730155948835e-05,
+      "loss": 1.5942,
+      "step": 7742
+    },
+    {
+      "epoch": 2.034748907938385,
+      "grad_norm": 0.5276470184326172,
+      "learning_rate": 3.217977921850359e-05,
+      "loss": 1.6259,
+      "step": 7744
+    },
+    {
+      "epoch": 2.0352744112720464,
+      "grad_norm": 0.5952677726745605,
+      "learning_rate": 3.216225687751884e-05,
+      "loss": 1.6073,
+      "step": 7746
+    },
+    {
+      "epoch": 2.0357999146057084,
+      "grad_norm": 0.5393653512001038,
+      "learning_rate": 3.2144734536534085e-05,
+      "loss": 1.612,
+      "step": 7748
+    },
+    {
+      "epoch": 2.03632541793937,
+      "grad_norm": 0.5733909010887146,
+      "learning_rate": 3.2127212195549325e-05,
+      "loss": 1.6513,
+      "step": 7750
+    },
+    {
+      "epoch": 2.036850921273032,
+      "grad_norm": 0.6015291810035706,
+      "learning_rate": 3.210968985456457e-05,
+      "loss": 1.613,
+      "step": 7752
+    },
+    {
+      "epoch": 2.0373764246066934,
+      "grad_norm": 0.6741379499435425,
+      "learning_rate": 3.209216751357982e-05,
+      "loss": 1.6277,
+      "step": 7754
+    },
+    {
+      "epoch": 2.0379019279403554,
+      "grad_norm": 0.6242867112159729,
+      "learning_rate": 3.207464517259506e-05,
+      "loss": 1.6299,
+      "step": 7756
+    },
+    {
+      "epoch": 2.038427431274017,
+      "grad_norm": 0.6203280091285706,
+      "learning_rate": 3.20571228316103e-05,
+      "loss": 1.624,
+      "step": 7758
+    },
+    {
+      "epoch": 2.038952934607679,
+      "grad_norm": 0.5922704339027405,
+      "learning_rate": 3.203960049062555e-05,
+      "loss": 1.6176,
+      "step": 7760
+    },
+    {
+      "epoch": 2.039478437941341,
+      "grad_norm": 0.5987135767936707,
+      "learning_rate": 3.20220781496408e-05,
+      "loss": 1.631,
+      "step": 7762
+    },
+    {
+      "epoch": 2.0400039412750024,
+      "grad_norm": 0.658090353012085,
+      "learning_rate": 3.200455580865604e-05,
+      "loss": 1.6455,
+      "step": 7764
+    },
+    {
+      "epoch": 2.0405294446086644,
+      "grad_norm": 0.5447620749473572,
+      "learning_rate": 3.1987033467671286e-05,
+      "loss": 1.6257,
+      "step": 7766
+    },
+    {
+      "epoch": 2.041054947942326,
+      "grad_norm": 0.5644233226776123,
+      "learning_rate": 3.1969511126686526e-05,
+      "loss": 1.6187,
+      "step": 7768
+    },
+    {
+      "epoch": 2.041580451275988,
+      "grad_norm": 0.6040834188461304,
+      "learning_rate": 3.195198878570177e-05,
+      "loss": 1.651,
+      "step": 7770
+    },
+    {
+      "epoch": 2.0421059546096494,
+      "grad_norm": 0.5864139199256897,
+      "learning_rate": 3.1934466444717015e-05,
+      "loss": 1.6151,
+      "step": 7772
+    },
+    {
+      "epoch": 2.0426314579433114,
+      "grad_norm": 0.5407016277313232,
+      "learning_rate": 3.191694410373226e-05,
+      "loss": 1.6075,
+      "step": 7774
+    },
+    {
+      "epoch": 2.043156961276973,
+      "grad_norm": 0.6198828816413879,
+      "learning_rate": 3.18994217627475e-05,
+      "loss": 1.6075,
+      "step": 7776
+    },
+    {
+      "epoch": 2.043682464610635,
+      "grad_norm": 0.5452790856361389,
+      "learning_rate": 3.188189942176275e-05,
+      "loss": 1.5942,
+      "step": 7778
+    },
+    {
+      "epoch": 2.044207967944297,
+      "grad_norm": 0.5661458969116211,
+      "learning_rate": 3.1864377080778e-05,
+      "loss": 1.64,
+      "step": 7780
+    },
+    {
+      "epoch": 2.0447334712779583,
+      "grad_norm": 0.5775606036186218,
+      "learning_rate": 3.184685473979324e-05,
+      "loss": 1.6221,
+      "step": 7782
+    },
+    {
+      "epoch": 2.0452589746116203,
+      "grad_norm": 0.6352986693382263,
+      "learning_rate": 3.182933239880848e-05,
+      "loss": 1.5736,
+      "step": 7784
+    },
+    {
+      "epoch": 2.045784477945282,
+      "grad_norm": 0.6562779545783997,
+      "learning_rate": 3.181181005782373e-05,
+      "loss": 1.659,
+      "step": 7786
+    },
+    {
+      "epoch": 2.046309981278944,
+      "grad_norm": 0.6051666140556335,
+      "learning_rate": 3.1794287716838975e-05,
+      "loss": 1.6418,
+      "step": 7788
+    },
+    {
+      "epoch": 2.0468354846126053,
+      "grad_norm": 0.5075708031654358,
+      "learning_rate": 3.1776765375854216e-05,
+      "loss": 1.6209,
+      "step": 7790
+    },
+    {
+      "epoch": 2.0473609879462673,
+      "grad_norm": 0.646185576915741,
+      "learning_rate": 3.175924303486946e-05,
+      "loss": 1.6081,
+      "step": 7792
+    },
+    {
+      "epoch": 2.047886491279929,
+      "grad_norm": 0.5304019451141357,
+      "learning_rate": 3.1741720693884704e-05,
+      "loss": 1.6065,
+      "step": 7794
+    },
+    {
+      "epoch": 2.048411994613591,
+      "grad_norm": 0.6288559436798096,
+      "learning_rate": 3.1724198352899945e-05,
+      "loss": 1.63,
+      "step": 7796
+    },
+    {
+      "epoch": 2.0489374979472528,
+      "grad_norm": 0.5334764122962952,
+      "learning_rate": 3.170667601191519e-05,
+      "loss": 1.6161,
+      "step": 7798
+    },
+    {
+      "epoch": 2.0494630012809143,
+      "grad_norm": 0.5528808236122131,
+      "learning_rate": 3.168915367093044e-05,
+      "loss": 1.5925,
+      "step": 7800
+    },
+    {
+      "epoch": 2.0499885046145763,
+      "grad_norm": 0.5791876912117004,
+      "learning_rate": 3.167163132994568e-05,
+      "loss": 1.6322,
+      "step": 7802
+    },
+    {
+      "epoch": 2.050514007948238,
+      "grad_norm": 0.6923157572746277,
+      "learning_rate": 3.165410898896093e-05,
+      "loss": 1.6301,
+      "step": 7804
+    },
+    {
+      "epoch": 2.0510395112818998,
+      "grad_norm": 0.5527494549751282,
+      "learning_rate": 3.163658664797617e-05,
+      "loss": 1.6363,
+      "step": 7806
+    },
+    {
+      "epoch": 2.0515650146155613,
+      "grad_norm": 0.554539144039154,
+      "learning_rate": 3.1619064306991417e-05,
+      "loss": 1.6324,
+      "step": 7808
+    },
+    {
+      "epoch": 2.0520905179492233,
+      "grad_norm": 0.6409008502960205,
+      "learning_rate": 3.160154196600666e-05,
+      "loss": 1.6378,
+      "step": 7810
+    },
+    {
+      "epoch": 2.0526160212828852,
+      "grad_norm": 0.6073037385940552,
+      "learning_rate": 3.1584019625021905e-05,
+      "loss": 1.5801,
+      "step": 7812
+    },
+    {
+      "epoch": 2.0531415246165468,
+      "grad_norm": 0.6152986288070679,
+      "learning_rate": 3.156649728403715e-05,
+      "loss": 1.6128,
+      "step": 7814
+    },
+    {
+      "epoch": 2.0536670279502087,
+      "grad_norm": 0.5760658383369446,
+      "learning_rate": 3.154897494305239e-05,
+      "loss": 1.5979,
+      "step": 7816
+    },
+    {
+      "epoch": 2.0541925312838702,
+      "grad_norm": 0.5909692645072937,
+      "learning_rate": 3.153145260206764e-05,
+      "loss": 1.5931,
+      "step": 7818
+    },
+    {
+      "epoch": 2.054718034617532,
+      "grad_norm": 0.5456188917160034,
+      "learning_rate": 3.151393026108288e-05,
+      "loss": 1.6557,
+      "step": 7820
+    },
+    {
+      "epoch": 2.0552435379511937,
+      "grad_norm": 0.6995749473571777,
+      "learning_rate": 3.149640792009812e-05,
+      "loss": 1.5974,
+      "step": 7822
+    },
+    {
+      "epoch": 2.0557690412848557,
+      "grad_norm": 0.7031815648078918,
+      "learning_rate": 3.147888557911337e-05,
+      "loss": 1.6331,
+      "step": 7824
+    },
+    {
+      "epoch": 2.0562945446185172,
+      "grad_norm": 0.5628200769424438,
+      "learning_rate": 3.146136323812862e-05,
+      "loss": 1.6148,
+      "step": 7826
+    },
+    {
+      "epoch": 2.056820047952179,
+      "grad_norm": 0.5670493841171265,
+      "learning_rate": 3.144384089714386e-05,
+      "loss": 1.6436,
+      "step": 7828
+    },
+    {
+      "epoch": 2.057345551285841,
+      "grad_norm": 0.5815064311027527,
+      "learning_rate": 3.1426318556159106e-05,
+      "loss": 1.6107,
+      "step": 7830
+    },
+    {
+      "epoch": 2.0578710546195027,
+      "grad_norm": 0.4898243844509125,
+      "learning_rate": 3.140879621517435e-05,
+      "loss": 1.5852,
+      "step": 7832
+    },
+    {
+      "epoch": 2.0583965579531647,
+      "grad_norm": 0.71734219789505,
+      "learning_rate": 3.1391273874189594e-05,
+      "loss": 1.6009,
+      "step": 7834
+    },
+    {
+      "epoch": 2.058922061286826,
+      "grad_norm": 0.5565480589866638,
+      "learning_rate": 3.1373751533204835e-05,
+      "loss": 1.611,
+      "step": 7836
+    },
+    {
+      "epoch": 2.059447564620488,
+      "grad_norm": 0.5884687900543213,
+      "learning_rate": 3.135622919222008e-05,
+      "loss": 1.5922,
+      "step": 7838
+    },
+    {
+      "epoch": 2.0599730679541497,
+      "grad_norm": 0.5969564318656921,
+      "learning_rate": 3.133870685123533e-05,
+      "loss": 1.6237,
+      "step": 7840
+    },
+    {
+      "epoch": 2.0604985712878117,
+      "grad_norm": 0.6498041152954102,
+      "learning_rate": 3.132118451025057e-05,
+      "loss": 1.6127,
+      "step": 7842
+    },
+    {
+      "epoch": 2.061024074621473,
+      "grad_norm": 0.6092948913574219,
+      "learning_rate": 3.130366216926581e-05,
+      "loss": 1.6377,
+      "step": 7844
+    },
+    {
+      "epoch": 2.061549577955135,
+      "grad_norm": 0.5632656812667847,
+      "learning_rate": 3.128613982828106e-05,
+      "loss": 1.5976,
+      "step": 7846
+    },
+    {
+      "epoch": 2.062075081288797,
+      "grad_norm": 0.6170439720153809,
+      "learning_rate": 3.12686174872963e-05,
+      "loss": 1.5962,
+      "step": 7848
+    },
+    {
+      "epoch": 2.0626005846224587,
+      "grad_norm": 0.5882241725921631,
+      "learning_rate": 3.125109514631155e-05,
+      "loss": 1.5918,
+      "step": 7850
+    },
+    {
+      "epoch": 2.0631260879561206,
+      "grad_norm": 0.5418891906738281,
+      "learning_rate": 3.1233572805326795e-05,
+      "loss": 1.6289,
+      "step": 7852
+    },
+    {
+      "epoch": 2.063651591289782,
+      "grad_norm": 0.6917753219604492,
+      "learning_rate": 3.1216050464342036e-05,
+      "loss": 1.6565,
+      "step": 7854
+    },
+    {
+      "epoch": 2.064177094623444,
+      "grad_norm": 0.596343457698822,
+      "learning_rate": 3.1198528123357284e-05,
+      "loss": 1.6087,
+      "step": 7856
+    },
+    {
+      "epoch": 2.0647025979571056,
+      "grad_norm": 0.5705437660217285,
+      "learning_rate": 3.1181005782372524e-05,
+      "loss": 1.607,
+      "step": 7858
+    },
+    {
+      "epoch": 2.0652281012907676,
+      "grad_norm": 0.5608698129653931,
+      "learning_rate": 3.116348344138777e-05,
+      "loss": 1.5935,
+      "step": 7860
+    },
+    {
+      "epoch": 2.065753604624429,
+      "grad_norm": 0.5956361293792725,
+      "learning_rate": 3.114596110040301e-05,
+      "loss": 1.5784,
+      "step": 7862
+    },
+    {
+      "epoch": 2.066279107958091,
+      "grad_norm": 0.5428637862205505,
+      "learning_rate": 3.112843875941826e-05,
+      "loss": 1.6384,
+      "step": 7864
+    },
+    {
+      "epoch": 2.066804611291753,
+      "grad_norm": 0.5758678317070007,
+      "learning_rate": 3.111091641843351e-05,
+      "loss": 1.6397,
+      "step": 7866
+    },
+    {
+      "epoch": 2.0673301146254146,
+      "grad_norm": 0.605904221534729,
+      "learning_rate": 3.109339407744875e-05,
+      "loss": 1.6312,
+      "step": 7868
+    },
+    {
+      "epoch": 2.0678556179590766,
+      "grad_norm": 0.5100085735321045,
+      "learning_rate": 3.107587173646399e-05,
+      "loss": 1.583,
+      "step": 7870
+    },
+    {
+      "epoch": 2.068381121292738,
+      "grad_norm": 0.5583624839782715,
+      "learning_rate": 3.105834939547924e-05,
+      "loss": 1.6319,
+      "step": 7872
+    },
+    {
+      "epoch": 2.0689066246264,
+      "grad_norm": 0.7279824018478394,
+      "learning_rate": 3.1040827054494485e-05,
+      "loss": 1.6142,
+      "step": 7874
+    },
+    {
+      "epoch": 2.0694321279600616,
+      "grad_norm": 0.5635095238685608,
+      "learning_rate": 3.1023304713509725e-05,
+      "loss": 1.6363,
+      "step": 7876
+    },
+    {
+      "epoch": 2.0699576312937236,
+      "grad_norm": 0.6370921730995178,
+      "learning_rate": 3.100578237252497e-05,
+      "loss": 1.6132,
+      "step": 7878
+    },
+    {
+      "epoch": 2.070483134627385,
+      "grad_norm": 0.6772926449775696,
+      "learning_rate": 3.098826003154022e-05,
+      "loss": 1.6033,
+      "step": 7880
+    },
+    {
+      "epoch": 2.071008637961047,
+      "grad_norm": 0.7656962275505066,
+      "learning_rate": 3.0970737690555454e-05,
+      "loss": 1.5948,
+      "step": 7882
+    },
+    {
+      "epoch": 2.071534141294709,
+      "grad_norm": 0.6695175766944885,
+      "learning_rate": 3.09532153495707e-05,
+      "loss": 1.6419,
+      "step": 7884
+    },
+    {
+      "epoch": 2.0720596446283706,
+      "grad_norm": 0.7357918620109558,
+      "learning_rate": 3.093569300858595e-05,
+      "loss": 1.5987,
+      "step": 7886
+    },
+    {
+      "epoch": 2.0725851479620325,
+      "grad_norm": 0.5544195175170898,
+      "learning_rate": 3.091817066760119e-05,
+      "loss": 1.6389,
+      "step": 7888
+    },
+    {
+      "epoch": 2.073110651295694,
+      "grad_norm": 0.674032986164093,
+      "learning_rate": 3.090064832661644e-05,
+      "loss": 1.6046,
+      "step": 7890
+    },
+    {
+      "epoch": 2.073636154629356,
+      "grad_norm": 0.5453786253929138,
+      "learning_rate": 3.0883125985631686e-05,
+      "loss": 1.6194,
+      "step": 7892
+    },
+    {
+      "epoch": 2.0741616579630175,
+      "grad_norm": 0.6969361901283264,
+      "learning_rate": 3.0865603644646926e-05,
+      "loss": 1.5962,
+      "step": 7894
+    },
+    {
+      "epoch": 2.0746871612966795,
+      "grad_norm": 0.8166598677635193,
+      "learning_rate": 3.084808130366217e-05,
+      "loss": 1.5563,
+      "step": 7896
+    },
+    {
+      "epoch": 2.075212664630341,
+      "grad_norm": 0.6729124784469604,
+      "learning_rate": 3.0830558962677415e-05,
+      "loss": 1.5873,
+      "step": 7898
+    },
+    {
+      "epoch": 2.075738167964003,
+      "grad_norm": 0.5767272114753723,
+      "learning_rate": 3.081303662169266e-05,
+      "loss": 1.6117,
+      "step": 7900
+    },
+    {
+      "epoch": 2.076263671297665,
+      "grad_norm": 0.5525824427604675,
+      "learning_rate": 3.07955142807079e-05,
+      "loss": 1.6047,
+      "step": 7902
+    },
+    {
+      "epoch": 2.0767891746313265,
+      "grad_norm": 0.6294286251068115,
+      "learning_rate": 3.077799193972315e-05,
+      "loss": 1.6076,
+      "step": 7904
+    },
+    {
+      "epoch": 2.0773146779649885,
+      "grad_norm": 0.5846551060676575,
+      "learning_rate": 3.07604695987384e-05,
+      "loss": 1.6207,
+      "step": 7906
+    },
+    {
+      "epoch": 2.07784018129865,
+      "grad_norm": 0.5915541052818298,
+      "learning_rate": 3.074294725775363e-05,
+      "loss": 1.6222,
+      "step": 7908
+    },
+    {
+      "epoch": 2.078365684632312,
+      "grad_norm": 0.5178691148757935,
+      "learning_rate": 3.072542491676888e-05,
+      "loss": 1.6086,
+      "step": 7910
+    },
+    {
+      "epoch": 2.0788911879659735,
+      "grad_norm": 0.5767921805381775,
+      "learning_rate": 3.070790257578413e-05,
+      "loss": 1.6102,
+      "step": 7912
+    },
+    {
+      "epoch": 2.0794166912996355,
+      "grad_norm": 0.7136160135269165,
+      "learning_rate": 3.069038023479937e-05,
+      "loss": 1.6103,
+      "step": 7914
+    },
+    {
+      "epoch": 2.079942194633297,
+      "grad_norm": 0.5835669040679932,
+      "learning_rate": 3.0672857893814616e-05,
+      "loss": 1.6199,
+      "step": 7916
+    },
+    {
+      "epoch": 2.080467697966959,
+      "grad_norm": 0.6543889045715332,
+      "learning_rate": 3.065533555282986e-05,
+      "loss": 1.6452,
+      "step": 7918
+    },
+    {
+      "epoch": 2.080993201300621,
+      "grad_norm": 0.5414628386497498,
+      "learning_rate": 3.0637813211845104e-05,
+      "loss": 1.6277,
+      "step": 7920
+    },
+    {
+      "epoch": 2.0815187046342825,
+      "grad_norm": 0.5655497908592224,
+      "learning_rate": 3.0620290870860345e-05,
+      "loss": 1.6294,
+      "step": 7922
+    },
+    {
+      "epoch": 2.0820442079679444,
+      "grad_norm": 0.5398779511451721,
+      "learning_rate": 3.060276852987559e-05,
+      "loss": 1.5906,
+      "step": 7924
+    },
+    {
+      "epoch": 2.082569711301606,
+      "grad_norm": 0.6147701740264893,
+      "learning_rate": 3.058524618889084e-05,
+      "loss": 1.6274,
+      "step": 7926
+    },
+    {
+      "epoch": 2.083095214635268,
+      "grad_norm": 0.5175689458847046,
+      "learning_rate": 3.056772384790608e-05,
+      "loss": 1.6206,
+      "step": 7928
+    },
+    {
+      "epoch": 2.0836207179689294,
+      "grad_norm": 0.5908101797103882,
+      "learning_rate": 3.055020150692133e-05,
+      "loss": 1.5894,
+      "step": 7930
+    },
+    {
+      "epoch": 2.0841462213025914,
+      "grad_norm": 0.7519561648368835,
+      "learning_rate": 3.0532679165936576e-05,
+      "loss": 1.5862,
+      "step": 7932
+    },
+    {
+      "epoch": 2.0846717246362534,
+      "grad_norm": 0.6898811459541321,
+      "learning_rate": 3.0515156824951813e-05,
+      "loss": 1.5917,
+      "step": 7934
+    },
+    {
+      "epoch": 2.085197227969915,
+      "grad_norm": 0.6252642869949341,
+      "learning_rate": 3.0497634483967057e-05,
+      "loss": 1.616,
+      "step": 7936
+    },
+    {
+      "epoch": 2.085722731303577,
+      "grad_norm": 0.6489799618721008,
+      "learning_rate": 3.0480112142982305e-05,
+      "loss": 1.5802,
+      "step": 7938
+    },
+    {
+      "epoch": 2.0862482346372384,
+      "grad_norm": 0.544003427028656,
+      "learning_rate": 3.046258980199755e-05,
+      "loss": 1.6274,
+      "step": 7940
+    },
+    {
+      "epoch": 2.0867737379709004,
+      "grad_norm": 0.6666077971458435,
+      "learning_rate": 3.0445067461012793e-05,
+      "loss": 1.5999,
+      "step": 7942
+    },
+    {
+      "epoch": 2.087299241304562,
+      "grad_norm": 0.6988151669502258,
+      "learning_rate": 3.042754512002804e-05,
+      "loss": 1.6314,
+      "step": 7944
+    },
+    {
+      "epoch": 2.087824744638224,
+      "grad_norm": 0.5597438812255859,
+      "learning_rate": 3.0410022779043278e-05,
+      "loss": 1.6348,
+      "step": 7946
+    },
+    {
+      "epoch": 2.0883502479718854,
+      "grad_norm": 0.5571982264518738,
+      "learning_rate": 3.0392500438058526e-05,
+      "loss": 1.6356,
+      "step": 7948
+    },
+    {
+      "epoch": 2.0888757513055474,
+      "grad_norm": 0.5218850374221802,
+      "learning_rate": 3.037497809707377e-05,
+      "loss": 1.588,
+      "step": 7950
+    },
+    {
+      "epoch": 2.089401254639209,
+      "grad_norm": 0.627077579498291,
+      "learning_rate": 3.0357455756089014e-05,
+      "loss": 1.5931,
+      "step": 7952
+    },
+    {
+      "epoch": 2.089926757972871,
+      "grad_norm": 0.6151776313781738,
+      "learning_rate": 3.0339933415104262e-05,
+      "loss": 1.6073,
+      "step": 7954
+    },
+    {
+      "epoch": 2.090452261306533,
+      "grad_norm": 0.5469711422920227,
+      "learning_rate": 3.0322411074119506e-05,
+      "loss": 1.5982,
+      "step": 7956
+    },
+    {
+      "epoch": 2.0909777646401944,
+      "grad_norm": 0.6806207299232483,
+      "learning_rate": 3.030488873313475e-05,
+      "loss": 1.6376,
+      "step": 7958
+    },
+    {
+      "epoch": 2.0915032679738563,
+      "grad_norm": 0.6746312379837036,
+      "learning_rate": 3.028736639214999e-05,
+      "loss": 1.6113,
+      "step": 7960
+    },
+    {
+      "epoch": 2.092028771307518,
+      "grad_norm": 0.6440801024436951,
+      "learning_rate": 3.0269844051165235e-05,
+      "loss": 1.6223,
+      "step": 7962
+    },
+    {
+      "epoch": 2.09255427464118,
+      "grad_norm": 0.5747029185295105,
+      "learning_rate": 3.0252321710180483e-05,
+      "loss": 1.5959,
+      "step": 7964
+    },
+    {
+      "epoch": 2.0930797779748413,
+      "grad_norm": 0.5977646708488464,
+      "learning_rate": 3.0234799369195727e-05,
+      "loss": 1.6403,
+      "step": 7966
+    },
+    {
+      "epoch": 2.0936052813085033,
+      "grad_norm": 0.5634188055992126,
+      "learning_rate": 3.021727702821097e-05,
+      "loss": 1.6062,
+      "step": 7968
+    },
+    {
+      "epoch": 2.0941307846421653,
+      "grad_norm": 0.661744236946106,
+      "learning_rate": 3.019975468722622e-05,
+      "loss": 1.6394,
+      "step": 7970
+    },
+    {
+      "epoch": 2.094656287975827,
+      "grad_norm": 0.5878733396530151,
+      "learning_rate": 3.0182232346241456e-05,
+      "loss": 1.5781,
+      "step": 7972
+    },
+    {
+      "epoch": 2.095181791309489,
+      "grad_norm": 0.5121323466300964,
+      "learning_rate": 3.0164710005256703e-05,
+      "loss": 1.6238,
+      "step": 7974
+    },
+    {
+      "epoch": 2.0957072946431503,
+      "grad_norm": 0.6021749377250671,
+      "learning_rate": 3.0147187664271948e-05,
+      "loss": 1.6077,
+      "step": 7976
+    },
+    {
+      "epoch": 2.0962327979768123,
+      "grad_norm": 0.5374654531478882,
+      "learning_rate": 3.0129665323287192e-05,
+      "loss": 1.6252,
+      "step": 7978
+    },
+    {
+      "epoch": 2.096758301310474,
+      "grad_norm": 0.5462518930435181,
+      "learning_rate": 3.011214298230244e-05,
+      "loss": 1.619,
+      "step": 7980
+    },
+    {
+      "epoch": 2.0972838046441358,
+      "grad_norm": 0.5597232580184937,
+      "learning_rate": 3.0094620641317684e-05,
+      "loss": 1.6151,
+      "step": 7982
+    },
+    {
+      "epoch": 2.0978093079777973,
+      "grad_norm": 0.6084597110748291,
+      "learning_rate": 3.0077098300332924e-05,
+      "loss": 1.6398,
+      "step": 7984
+    },
+    {
+      "epoch": 2.0983348113114593,
+      "grad_norm": 0.5504382252693176,
+      "learning_rate": 3.005957595934817e-05,
+      "loss": 1.5902,
+      "step": 7986
+    },
+    {
+      "epoch": 2.0988603146451212,
+      "grad_norm": 0.5516804456710815,
+      "learning_rate": 3.0042053618363413e-05,
+      "loss": 1.625,
+      "step": 7988
+    },
+    {
+      "epoch": 2.0993858179787828,
+      "grad_norm": 0.5523831248283386,
+      "learning_rate": 3.002453127737866e-05,
+      "loss": 1.6291,
+      "step": 7990
+    },
+    {
+      "epoch": 2.0999113213124447,
+      "grad_norm": 0.6193360686302185,
+      "learning_rate": 3.0007008936393904e-05,
+      "loss": 1.5764,
+      "step": 7992
+    },
+    {
+      "epoch": 2.1004368246461063,
+      "grad_norm": 0.620893657207489,
+      "learning_rate": 2.998948659540915e-05,
+      "loss": 1.6295,
+      "step": 7994
+    },
+    {
+      "epoch": 2.1009623279797682,
+      "grad_norm": 0.5495589971542358,
+      "learning_rate": 2.9971964254424396e-05,
+      "loss": 1.6159,
+      "step": 7996
+    },
+    {
+      "epoch": 2.1014878313134298,
+      "grad_norm": 0.5830954909324646,
+      "learning_rate": 2.9954441913439634e-05,
+      "loss": 1.5844,
+      "step": 7998
+    },
+    {
+      "epoch": 2.1020133346470917,
+      "grad_norm": 0.6515240669250488,
+      "learning_rate": 2.993691957245488e-05,
+      "loss": 1.6092,
+      "step": 8000
+    },
+    {
+      "epoch": 2.1020133346470917,
+      "eval_loss": 1.6652451753616333,
+      "eval_runtime": 487.2001,
+      "eval_samples_per_second": 249.977,
+      "eval_steps_per_second": 31.248,
+      "step": 8000
+    },
+    {
+      "epoch": 2.1025388379807533,
+      "grad_norm": 0.5562619566917419,
+      "learning_rate": 2.9919397231470125e-05,
+      "loss": 1.6272,
+      "step": 8002
+    },
+    {
+      "epoch": 2.103064341314415,
+      "grad_norm": 0.8904392123222351,
+      "learning_rate": 2.990187489048537e-05,
+      "loss": 1.6286,
+      "step": 8004
+    },
+    {
+      "epoch": 2.103589844648077,
+      "grad_norm": 0.6131122708320618,
+      "learning_rate": 2.9884352549500617e-05,
+      "loss": 1.6225,
+      "step": 8006
+    },
+    {
+      "epoch": 2.1041153479817387,
+      "grad_norm": 0.5321511030197144,
+      "learning_rate": 2.986683020851586e-05,
+      "loss": 1.5881,
+      "step": 8008
+    },
+    {
+      "epoch": 2.1046408513154007,
+      "grad_norm": 0.5121209621429443,
+      "learning_rate": 2.9849307867531102e-05,
+      "loss": 1.5884,
+      "step": 8010
+    },
+    {
+      "epoch": 2.105166354649062,
+      "grad_norm": 0.5307869911193848,
+      "learning_rate": 2.9831785526546346e-05,
+      "loss": 1.5926,
+      "step": 8012
+    },
+    {
+      "epoch": 2.105691857982724,
+      "grad_norm": 0.5231491327285767,
+      "learning_rate": 2.981426318556159e-05,
+      "loss": 1.6319,
+      "step": 8014
+    },
+    {
+      "epoch": 2.1062173613163857,
+      "grad_norm": 0.5306436419487,
+      "learning_rate": 2.9796740844576838e-05,
+      "loss": 1.6526,
+      "step": 8016
+    },
+    {
+      "epoch": 2.1067428646500477,
+      "grad_norm": 0.6463943123817444,
+      "learning_rate": 2.9779218503592082e-05,
+      "loss": 1.6074,
+      "step": 8018
+    },
+    {
+      "epoch": 2.107268367983709,
+      "grad_norm": 0.5345489382743835,
+      "learning_rate": 2.9761696162607326e-05,
+      "loss": 1.6413,
+      "step": 8020
+    },
+    {
+      "epoch": 2.107793871317371,
+      "grad_norm": 0.6284579634666443,
+      "learning_rate": 2.9744173821622574e-05,
+      "loss": 1.63,
+      "step": 8022
+    },
+    {
+      "epoch": 2.108319374651033,
+      "grad_norm": 0.5651626586914062,
+      "learning_rate": 2.972665148063781e-05,
+      "loss": 1.5837,
+      "step": 8024
+    },
+    {
+      "epoch": 2.1088448779846947,
+      "grad_norm": 0.692808210849762,
+      "learning_rate": 2.970912913965306e-05,
+      "loss": 1.6361,
+      "step": 8026
+    },
+    {
+      "epoch": 2.1093703813183566,
+      "grad_norm": 0.5368518829345703,
+      "learning_rate": 2.9691606798668303e-05,
+      "loss": 1.6577,
+      "step": 8028
+    },
+    {
+      "epoch": 2.109895884652018,
+      "grad_norm": 0.561772346496582,
+      "learning_rate": 2.9674084457683547e-05,
+      "loss": 1.5959,
+      "step": 8030
+    },
+    {
+      "epoch": 2.11042138798568,
+      "grad_norm": 0.6566132307052612,
+      "learning_rate": 2.9656562116698795e-05,
+      "loss": 1.622,
+      "step": 8032
+    },
+    {
+      "epoch": 2.1109468913193417,
+      "grad_norm": 0.7207258939743042,
+      "learning_rate": 2.963903977571404e-05,
+      "loss": 1.6318,
+      "step": 8034
+    },
+    {
+      "epoch": 2.1114723946530036,
+      "grad_norm": 0.5623325705528259,
+      "learning_rate": 2.962151743472928e-05,
+      "loss": 1.5964,
+      "step": 8036
+    },
+    {
+      "epoch": 2.111997897986665,
+      "grad_norm": 0.6407482624053955,
+      "learning_rate": 2.9603995093744524e-05,
+      "loss": 1.5759,
+      "step": 8038
+    },
+    {
+      "epoch": 2.112523401320327,
+      "grad_norm": 0.5547378659248352,
+      "learning_rate": 2.9586472752759768e-05,
+      "loss": 1.6112,
+      "step": 8040
+    },
+    {
+      "epoch": 2.113048904653989,
+      "grad_norm": 0.5998631715774536,
+      "learning_rate": 2.9568950411775016e-05,
+      "loss": 1.6102,
+      "step": 8042
+    },
+    {
+      "epoch": 2.1135744079876506,
+      "grad_norm": 0.5907427072525024,
+      "learning_rate": 2.955142807079026e-05,
+      "loss": 1.6056,
+      "step": 8044
+    },
+    {
+      "epoch": 2.1140999113213126,
+      "grad_norm": 0.5377500653266907,
+      "learning_rate": 2.9533905729805504e-05,
+      "loss": 1.626,
+      "step": 8046
+    },
+    {
+      "epoch": 2.114625414654974,
+      "grad_norm": 0.5881298780441284,
+      "learning_rate": 2.9516383388820745e-05,
+      "loss": 1.6245,
+      "step": 8048
+    },
+    {
+      "epoch": 2.115150917988636,
+      "grad_norm": 0.6338145136833191,
+      "learning_rate": 2.949886104783599e-05,
+      "loss": 1.6133,
+      "step": 8050
+    },
+    {
+      "epoch": 2.1156764213222976,
+      "grad_norm": 0.7909775972366333,
+      "learning_rate": 2.9481338706851236e-05,
+      "loss": 1.625,
+      "step": 8052
+    },
+    {
+      "epoch": 2.1162019246559596,
+      "grad_norm": 0.5986071228981018,
+      "learning_rate": 2.946381636586648e-05,
+      "loss": 1.6209,
+      "step": 8054
+    },
+    {
+      "epoch": 2.116727427989621,
+      "grad_norm": 0.6373353600502014,
+      "learning_rate": 2.9446294024881725e-05,
+      "loss": 1.5849,
+      "step": 8056
+    },
+    {
+      "epoch": 2.117252931323283,
+      "grad_norm": 0.7317929267883301,
+      "learning_rate": 2.9428771683896972e-05,
+      "loss": 1.6032,
+      "step": 8058
+    },
+    {
+      "epoch": 2.117778434656945,
+      "grad_norm": 0.5454069375991821,
+      "learning_rate": 2.9411249342912217e-05,
+      "loss": 1.6171,
+      "step": 8060
+    },
+    {
+      "epoch": 2.1183039379906066,
+      "grad_norm": 0.5561736226081848,
+      "learning_rate": 2.9393727001927457e-05,
+      "loss": 1.6434,
+      "step": 8062
+    },
+    {
+      "epoch": 2.1188294413242685,
+      "grad_norm": 0.685099720954895,
+      "learning_rate": 2.93762046609427e-05,
+      "loss": 1.5534,
+      "step": 8064
+    },
+    {
+      "epoch": 2.11935494465793,
+      "grad_norm": 0.5449316501617432,
+      "learning_rate": 2.935868231995795e-05,
+      "loss": 1.5926,
+      "step": 8066
+    },
+    {
+      "epoch": 2.119880447991592,
+      "grad_norm": 0.5778328776359558,
+      "learning_rate": 2.9341159978973193e-05,
+      "loss": 1.6166,
+      "step": 8068
+    },
+    {
+      "epoch": 2.1204059513252536,
+      "grad_norm": 0.5505948662757874,
+      "learning_rate": 2.9323637637988437e-05,
+      "loss": 1.5876,
+      "step": 8070
+    },
+    {
+      "epoch": 2.1209314546589155,
+      "grad_norm": 0.5759081244468689,
+      "learning_rate": 2.9306115297003685e-05,
+      "loss": 1.6145,
+      "step": 8072
+    },
+    {
+      "epoch": 2.121456957992577,
+      "grad_norm": 0.5547658205032349,
+      "learning_rate": 2.9288592956018922e-05,
+      "loss": 1.564,
+      "step": 8074
+    },
+    {
+      "epoch": 2.121982461326239,
+      "grad_norm": 0.6756736040115356,
+      "learning_rate": 2.927107061503417e-05,
+      "loss": 1.6301,
+      "step": 8076
+    },
+    {
+      "epoch": 2.122507964659901,
+      "grad_norm": 0.8322302103042603,
+      "learning_rate": 2.9253548274049414e-05,
+      "loss": 1.6044,
+      "step": 8078
+    },
+    {
+      "epoch": 2.1230334679935625,
+      "grad_norm": 0.6076900362968445,
+      "learning_rate": 2.923602593306466e-05,
+      "loss": 1.5982,
+      "step": 8080
+    },
+    {
+      "epoch": 2.1235589713272245,
+      "grad_norm": 0.5850244760513306,
+      "learning_rate": 2.9218503592079906e-05,
+      "loss": 1.6248,
+      "step": 8082
+    },
+    {
+      "epoch": 2.124084474660886,
+      "grad_norm": 0.5558748841285706,
+      "learning_rate": 2.920098125109515e-05,
+      "loss": 1.6224,
+      "step": 8084
+    },
+    {
+      "epoch": 2.124609977994548,
+      "grad_norm": 0.5563886165618896,
+      "learning_rate": 2.918345891011039e-05,
+      "loss": 1.5805,
+      "step": 8086
+    },
+    {
+      "epoch": 2.1251354813282095,
+      "grad_norm": 0.6558648943901062,
+      "learning_rate": 2.9165936569125635e-05,
+      "loss": 1.6255,
+      "step": 8088
+    },
+    {
+      "epoch": 2.1256609846618715,
+      "grad_norm": 0.6515272259712219,
+      "learning_rate": 2.914841422814088e-05,
+      "loss": 1.6127,
+      "step": 8090
+    },
+    {
+      "epoch": 2.1261864879955334,
+      "grad_norm": 0.5626866221427917,
+      "learning_rate": 2.9130891887156127e-05,
+      "loss": 1.6125,
+      "step": 8092
+    },
+    {
+      "epoch": 2.126711991329195,
+      "grad_norm": 0.6942625045776367,
+      "learning_rate": 2.911336954617137e-05,
+      "loss": 1.5824,
+      "step": 8094
+    },
+    {
+      "epoch": 2.127237494662857,
+      "grad_norm": 0.5993350148200989,
+      "learning_rate": 2.9095847205186615e-05,
+      "loss": 1.6309,
+      "step": 8096
+    },
+    {
+      "epoch": 2.1277629979965185,
+      "grad_norm": 0.5620863437652588,
+      "learning_rate": 2.9078324864201863e-05,
+      "loss": 1.6287,
+      "step": 8098
+    },
+    {
+      "epoch": 2.1282885013301804,
+      "grad_norm": 0.5420330762863159,
+      "learning_rate": 2.90608025232171e-05,
+      "loss": 1.6263,
+      "step": 8100
+    },
+    {
+      "epoch": 2.128814004663842,
+      "grad_norm": 0.5831283330917358,
+      "learning_rate": 2.9043280182232348e-05,
+      "loss": 1.6221,
+      "step": 8102
+    },
+    {
+      "epoch": 2.129339507997504,
+      "grad_norm": 0.6942195296287537,
+      "learning_rate": 2.9025757841247592e-05,
+      "loss": 1.6411,
+      "step": 8104
+    },
+    {
+      "epoch": 2.1298650113311655,
+      "grad_norm": 0.5689584016799927,
+      "learning_rate": 2.9008235500262836e-05,
+      "loss": 1.6443,
+      "step": 8106
+    },
+    {
+      "epoch": 2.1303905146648274,
+      "grad_norm": 0.6463392376899719,
+      "learning_rate": 2.8990713159278084e-05,
+      "loss": 1.5987,
+      "step": 8108
+    },
+    {
+      "epoch": 2.130916017998489,
+      "grad_norm": 0.591022253036499,
+      "learning_rate": 2.8973190818293328e-05,
+      "loss": 1.6283,
+      "step": 8110
+    },
+    {
+      "epoch": 2.131441521332151,
+      "grad_norm": 0.6164734363555908,
+      "learning_rate": 2.895566847730857e-05,
+      "loss": 1.6188,
+      "step": 8112
+    },
+    {
+      "epoch": 2.131967024665813,
+      "grad_norm": 0.5488820672035217,
+      "learning_rate": 2.8938146136323813e-05,
+      "loss": 1.6562,
+      "step": 8114
+    },
+    {
+      "epoch": 2.1324925279994744,
+      "grad_norm": 0.7101133465766907,
+      "learning_rate": 2.8920623795339057e-05,
+      "loss": 1.6206,
+      "step": 8116
+    },
+    {
+      "epoch": 2.1330180313331364,
+      "grad_norm": 0.5032434463500977,
+      "learning_rate": 2.8903101454354304e-05,
+      "loss": 1.6042,
+      "step": 8118
+    },
+    {
+      "epoch": 2.133543534666798,
+      "grad_norm": 0.5224511623382568,
+      "learning_rate": 2.888557911336955e-05,
+      "loss": 1.5865,
+      "step": 8120
+    },
+    {
+      "epoch": 2.13406903800046,
+      "grad_norm": 0.5269969701766968,
+      "learning_rate": 2.8868056772384793e-05,
+      "loss": 1.6296,
+      "step": 8122
+    },
+    {
+      "epoch": 2.1345945413341214,
+      "grad_norm": 0.5620083212852478,
+      "learning_rate": 2.885053443140004e-05,
+      "loss": 1.6425,
+      "step": 8124
+    },
+    {
+      "epoch": 2.1351200446677834,
+      "grad_norm": 0.6436342597007751,
+      "learning_rate": 2.8833012090415278e-05,
+      "loss": 1.6053,
+      "step": 8126
+    },
+    {
+      "epoch": 2.1356455480014453,
+      "grad_norm": 0.6406291127204895,
+      "learning_rate": 2.8815489749430525e-05,
+      "loss": 1.6254,
+      "step": 8128
+    },
+    {
+      "epoch": 2.136171051335107,
+      "grad_norm": 0.551295280456543,
+      "learning_rate": 2.879796740844577e-05,
+      "loss": 1.621,
+      "step": 8130
+    },
+    {
+      "epoch": 2.136696554668769,
+      "grad_norm": 0.6591944098472595,
+      "learning_rate": 2.8780445067461014e-05,
+      "loss": 1.6578,
+      "step": 8132
+    },
+    {
+      "epoch": 2.1372220580024304,
+      "grad_norm": 0.7538178563117981,
+      "learning_rate": 2.876292272647626e-05,
+      "loss": 1.6057,
+      "step": 8134
+    },
+    {
+      "epoch": 2.1377475613360923,
+      "grad_norm": 0.5904437303543091,
+      "learning_rate": 2.8745400385491505e-05,
+      "loss": 1.5984,
+      "step": 8136
+    },
+    {
+      "epoch": 2.138273064669754,
+      "grad_norm": 0.5584096312522888,
+      "learning_rate": 2.8727878044506746e-05,
+      "loss": 1.5751,
+      "step": 8138
+    },
+    {
+      "epoch": 2.138798568003416,
+      "grad_norm": 0.6350988745689392,
+      "learning_rate": 2.871035570352199e-05,
+      "loss": 1.6102,
+      "step": 8140
+    },
+    {
+      "epoch": 2.1393240713370774,
+      "grad_norm": 0.8004497289657593,
+      "learning_rate": 2.8692833362537235e-05,
+      "loss": 1.6205,
+      "step": 8142
+    },
+    {
+      "epoch": 2.1398495746707393,
+      "grad_norm": 0.6031829118728638,
+      "learning_rate": 2.8675311021552482e-05,
+      "loss": 1.615,
+      "step": 8144
+    },
+    {
+      "epoch": 2.140375078004401,
+      "grad_norm": 0.5796357989311218,
+      "learning_rate": 2.8657788680567726e-05,
+      "loss": 1.6007,
+      "step": 8146
+    },
+    {
+      "epoch": 2.140900581338063,
+      "grad_norm": 0.5166471004486084,
+      "learning_rate": 2.864026633958297e-05,
+      "loss": 1.592,
+      "step": 8148
+    },
+    {
+      "epoch": 2.141426084671725,
+      "grad_norm": 0.6542901396751404,
+      "learning_rate": 2.862274399859821e-05,
+      "loss": 1.648,
+      "step": 8150
+    },
+    {
+      "epoch": 2.1419515880053863,
+      "grad_norm": 0.5643502473831177,
+      "learning_rate": 2.8605221657613455e-05,
+      "loss": 1.6187,
+      "step": 8152
+    },
+    {
+      "epoch": 2.1424770913390483,
+      "grad_norm": 0.5144004821777344,
+      "learning_rate": 2.8587699316628703e-05,
+      "loss": 1.5961,
+      "step": 8154
+    },
+    {
+      "epoch": 2.14300259467271,
+      "grad_norm": 0.5907447934150696,
+      "learning_rate": 2.8570176975643947e-05,
+      "loss": 1.6091,
+      "step": 8156
+    },
+    {
+      "epoch": 2.143528098006372,
+      "grad_norm": 0.6062257885932922,
+      "learning_rate": 2.855265463465919e-05,
+      "loss": 1.596,
+      "step": 8158
+    },
+    {
+      "epoch": 2.1440536013400333,
+      "grad_norm": 0.5850198268890381,
+      "learning_rate": 2.853513229367444e-05,
+      "loss": 1.6202,
+      "step": 8160
+    },
+    {
+      "epoch": 2.1445791046736953,
+      "grad_norm": 0.6049871444702148,
+      "learning_rate": 2.8517609952689683e-05,
+      "loss": 1.6151,
+      "step": 8162
+    },
+    {
+      "epoch": 2.1451046080073572,
+      "grad_norm": 0.6060041785240173,
+      "learning_rate": 2.8500087611704924e-05,
+      "loss": 1.566,
+      "step": 8164
+    },
+    {
+      "epoch": 2.1456301113410188,
+      "grad_norm": 0.5683371424674988,
+      "learning_rate": 2.8482565270720168e-05,
+      "loss": 1.6301,
+      "step": 8166
+    },
+    {
+      "epoch": 2.1461556146746807,
+      "grad_norm": 0.5292571783065796,
+      "learning_rate": 2.8465042929735412e-05,
+      "loss": 1.6151,
+      "step": 8168
+    },
+    {
+      "epoch": 2.1466811180083423,
+      "grad_norm": 0.6342765688896179,
+      "learning_rate": 2.844752058875066e-05,
+      "loss": 1.5728,
+      "step": 8170
+    },
+    {
+      "epoch": 2.1472066213420042,
+      "grad_norm": 0.551582932472229,
+      "learning_rate": 2.8429998247765904e-05,
+      "loss": 1.6054,
+      "step": 8172
+    },
+    {
+      "epoch": 2.1477321246756658,
+      "grad_norm": 0.548427939414978,
+      "learning_rate": 2.8412475906781148e-05,
+      "loss": 1.5991,
+      "step": 8174
+    },
+    {
+      "epoch": 2.1482576280093277,
+      "grad_norm": 0.6584466695785522,
+      "learning_rate": 2.839495356579639e-05,
+      "loss": 1.6213,
+      "step": 8176
+    },
+    {
+      "epoch": 2.1487831313429893,
+      "grad_norm": 0.6075412631034851,
+      "learning_rate": 2.8377431224811633e-05,
+      "loss": 1.6279,
+      "step": 8178
+    },
+    {
+      "epoch": 2.1493086346766512,
+      "grad_norm": 0.569524884223938,
+      "learning_rate": 2.835990888382688e-05,
+      "loss": 1.6053,
+      "step": 8180
+    },
+    {
+      "epoch": 2.149834138010313,
+      "grad_norm": 0.5383790731430054,
+      "learning_rate": 2.8342386542842125e-05,
+      "loss": 1.5938,
+      "step": 8182
+    },
+    {
+      "epoch": 2.1503596413439747,
+      "grad_norm": 0.5958060026168823,
+      "learning_rate": 2.832486420185737e-05,
+      "loss": 1.6272,
+      "step": 8184
+    },
+    {
+      "epoch": 2.1508851446776367,
+      "grad_norm": 0.5532635450363159,
+      "learning_rate": 2.8307341860872617e-05,
+      "loss": 1.6431,
+      "step": 8186
+    },
+    {
+      "epoch": 2.151410648011298,
+      "grad_norm": 0.5432900190353394,
+      "learning_rate": 2.8289819519887857e-05,
+      "loss": 1.6579,
+      "step": 8188
+    },
+    {
+      "epoch": 2.15193615134496,
+      "grad_norm": 0.6363177299499512,
+      "learning_rate": 2.82722971789031e-05,
+      "loss": 1.6349,
+      "step": 8190
+    },
+    {
+      "epoch": 2.1524616546786217,
+      "grad_norm": 0.6205646395683289,
+      "learning_rate": 2.8254774837918346e-05,
+      "loss": 1.6014,
+      "step": 8192
+    },
+    {
+      "epoch": 2.1529871580122837,
+      "grad_norm": 0.5696967244148254,
+      "learning_rate": 2.8237252496933593e-05,
+      "loss": 1.6086,
+      "step": 8194
+    },
+    {
+      "epoch": 2.153512661345945,
+      "grad_norm": 0.5662211775779724,
+      "learning_rate": 2.8219730155948837e-05,
+      "loss": 1.6246,
+      "step": 8196
+    },
+    {
+      "epoch": 2.154038164679607,
+      "grad_norm": 0.5794999599456787,
+      "learning_rate": 2.820220781496408e-05,
+      "loss": 1.6127,
+      "step": 8198
+    },
+    {
+      "epoch": 2.154563668013269,
+      "grad_norm": 0.6433011889457703,
+      "learning_rate": 2.818468547397933e-05,
+      "loss": 1.5973,
+      "step": 8200
+    },
+    {
+      "epoch": 2.1550891713469307,
+      "grad_norm": 0.5784979462623596,
+      "learning_rate": 2.8167163132994567e-05,
+      "loss": 1.6314,
+      "step": 8202
+    },
+    {
+      "epoch": 2.1556146746805926,
+      "grad_norm": 0.5415730476379395,
+      "learning_rate": 2.8149640792009814e-05,
+      "loss": 1.635,
+      "step": 8204
+    },
+    {
+      "epoch": 2.156140178014254,
+      "grad_norm": 0.5347517132759094,
+      "learning_rate": 2.8132118451025058e-05,
+      "loss": 1.5721,
+      "step": 8206
+    },
+    {
+      "epoch": 2.156665681347916,
+      "grad_norm": 0.5696678757667542,
+      "learning_rate": 2.8114596110040302e-05,
+      "loss": 1.6163,
+      "step": 8208
+    },
+    {
+      "epoch": 2.1571911846815777,
+      "grad_norm": 0.5461645126342773,
+      "learning_rate": 2.809707376905555e-05,
+      "loss": 1.6097,
+      "step": 8210
+    },
+    {
+      "epoch": 2.1577166880152396,
+      "grad_norm": 0.5799645781517029,
+      "learning_rate": 2.8079551428070794e-05,
+      "loss": 1.6526,
+      "step": 8212
+    },
+    {
+      "epoch": 2.158242191348901,
+      "grad_norm": 0.5340853333473206,
+      "learning_rate": 2.8062029087086035e-05,
+      "loss": 1.5865,
+      "step": 8214
+    },
+    {
+      "epoch": 2.158767694682563,
+      "grad_norm": 0.5770725607872009,
+      "learning_rate": 2.804450674610128e-05,
+      "loss": 1.6281,
+      "step": 8216
+    },
+    {
+      "epoch": 2.159293198016225,
+      "grad_norm": 0.5207443237304688,
+      "learning_rate": 2.8026984405116523e-05,
+      "loss": 1.6303,
+      "step": 8218
+    },
+    {
+      "epoch": 2.1598187013498866,
+      "grad_norm": 0.6137516498565674,
+      "learning_rate": 2.800946206413177e-05,
+      "loss": 1.639,
+      "step": 8220
+    },
+    {
+      "epoch": 2.1603442046835486,
+      "grad_norm": 0.5726540088653564,
+      "learning_rate": 2.7991939723147015e-05,
+      "loss": 1.6108,
+      "step": 8222
+    },
+    {
+      "epoch": 2.16086970801721,
+      "grad_norm": 0.5607831478118896,
+      "learning_rate": 2.797441738216226e-05,
+      "loss": 1.611,
+      "step": 8224
+    },
+    {
+      "epoch": 2.161395211350872,
+      "grad_norm": 0.6390299201011658,
+      "learning_rate": 2.79568950411775e-05,
+      "loss": 1.6325,
+      "step": 8226
+    },
+    {
+      "epoch": 2.1619207146845336,
+      "grad_norm": 0.548121452331543,
+      "learning_rate": 2.7939372700192744e-05,
+      "loss": 1.6311,
+      "step": 8228
+    },
+    {
+      "epoch": 2.1624462180181956,
+      "grad_norm": 0.6899086833000183,
+      "learning_rate": 2.7921850359207992e-05,
+      "loss": 1.5686,
+      "step": 8230
+    },
+    {
+      "epoch": 2.162971721351857,
+      "grad_norm": 0.6069139838218689,
+      "learning_rate": 2.7904328018223236e-05,
+      "loss": 1.603,
+      "step": 8232
+    },
+    {
+      "epoch": 2.163497224685519,
+      "grad_norm": 0.5628888010978699,
+      "learning_rate": 2.788680567723848e-05,
+      "loss": 1.616,
+      "step": 8234
+    },
+    {
+      "epoch": 2.164022728019181,
+      "grad_norm": 0.5229991674423218,
+      "learning_rate": 2.7869283336253728e-05,
+      "loss": 1.6172,
+      "step": 8236
+    },
+    {
+      "epoch": 2.1645482313528426,
+      "grad_norm": 0.5317997932434082,
+      "learning_rate": 2.7851760995268972e-05,
+      "loss": 1.5771,
+      "step": 8238
+    },
+    {
+      "epoch": 2.1650737346865045,
+      "grad_norm": 0.5842552781105042,
+      "learning_rate": 2.7834238654284213e-05,
+      "loss": 1.6199,
+      "step": 8240
+    },
+    {
+      "epoch": 2.165599238020166,
+      "grad_norm": 0.6741251349449158,
+      "learning_rate": 2.7816716313299457e-05,
+      "loss": 1.6156,
+      "step": 8242
+    },
+    {
+      "epoch": 2.166124741353828,
+      "grad_norm": 0.5447443723678589,
+      "learning_rate": 2.77991939723147e-05,
+      "loss": 1.5832,
+      "step": 8244
+    },
+    {
+      "epoch": 2.1666502446874896,
+      "grad_norm": 0.49737128615379333,
+      "learning_rate": 2.778167163132995e-05,
+      "loss": 1.6369,
+      "step": 8246
+    },
+    {
+      "epoch": 2.1671757480211515,
+      "grad_norm": 0.6779609322547913,
+      "learning_rate": 2.7764149290345193e-05,
+      "loss": 1.631,
+      "step": 8248
+    },
+    {
+      "epoch": 2.1677012513548135,
+      "grad_norm": 0.5730245113372803,
+      "learning_rate": 2.7746626949360437e-05,
+      "loss": 1.6004,
+      "step": 8250
+    },
+    {
+      "epoch": 2.168226754688475,
+      "grad_norm": 0.5977212190628052,
+      "learning_rate": 2.7729104608375678e-05,
+      "loss": 1.6173,
+      "step": 8252
+    },
+    {
+      "epoch": 2.168752258022137,
+      "grad_norm": 0.5202796459197998,
+      "learning_rate": 2.7711582267390922e-05,
+      "loss": 1.6027,
+      "step": 8254
+    },
+    {
+      "epoch": 2.1692777613557985,
+      "grad_norm": 0.7288246154785156,
+      "learning_rate": 2.769405992640617e-05,
+      "loss": 1.6007,
+      "step": 8256
+    },
+    {
+      "epoch": 2.1698032646894605,
+      "grad_norm": 0.5423540472984314,
+      "learning_rate": 2.7676537585421414e-05,
+      "loss": 1.6032,
+      "step": 8258
+    },
+    {
+      "epoch": 2.170328768023122,
+      "grad_norm": 0.6103919744491577,
+      "learning_rate": 2.7659015244436658e-05,
+      "loss": 1.6147,
+      "step": 8260
+    },
+    {
+      "epoch": 2.170854271356784,
+      "grad_norm": 0.5658676624298096,
+      "learning_rate": 2.7641492903451905e-05,
+      "loss": 1.6187,
+      "step": 8262
+    },
+    {
+      "epoch": 2.1713797746904455,
+      "grad_norm": 0.5302108526229858,
+      "learning_rate": 2.762397056246715e-05,
+      "loss": 1.5885,
+      "step": 8264
+    },
+    {
+      "epoch": 2.1719052780241075,
+      "grad_norm": 0.5685304999351501,
+      "learning_rate": 2.760644822148239e-05,
+      "loss": 1.5809,
+      "step": 8266
+    },
+    {
+      "epoch": 2.172430781357769,
+      "grad_norm": 0.6211028695106506,
+      "learning_rate": 2.7588925880497635e-05,
+      "loss": 1.5982,
+      "step": 8268
+    },
+    {
+      "epoch": 2.172956284691431,
+      "grad_norm": 0.5506071448326111,
+      "learning_rate": 2.757140353951288e-05,
+      "loss": 1.6059,
+      "step": 8270
+    },
+    {
+      "epoch": 2.173481788025093,
+      "grad_norm": 0.5578741431236267,
+      "learning_rate": 2.7553881198528126e-05,
+      "loss": 1.6065,
+      "step": 8272
+    },
+    {
+      "epoch": 2.1740072913587545,
+      "grad_norm": 0.6002737283706665,
+      "learning_rate": 2.753635885754337e-05,
+      "loss": 1.6622,
+      "step": 8274
+    },
+    {
+      "epoch": 2.1745327946924164,
+      "grad_norm": 0.5421825647354126,
+      "learning_rate": 2.7518836516558615e-05,
+      "loss": 1.594,
+      "step": 8276
+    },
+    {
+      "epoch": 2.175058298026078,
+      "grad_norm": 0.5473990440368652,
+      "learning_rate": 2.7501314175573855e-05,
+      "loss": 1.6006,
+      "step": 8278
+    },
+    {
+      "epoch": 2.17558380135974,
+      "grad_norm": 0.6513854265213013,
+      "learning_rate": 2.74837918345891e-05,
+      "loss": 1.6222,
+      "step": 8280
+    },
+    {
+      "epoch": 2.1761093046934015,
+      "grad_norm": 0.5265668034553528,
+      "learning_rate": 2.7466269493604347e-05,
+      "loss": 1.5859,
+      "step": 8282
+    },
+    {
+      "epoch": 2.1766348080270634,
+      "grad_norm": 0.5653666853904724,
+      "learning_rate": 2.744874715261959e-05,
+      "loss": 1.6145,
+      "step": 8284
+    },
+    {
+      "epoch": 2.1771603113607254,
+      "grad_norm": 0.6216524839401245,
+      "learning_rate": 2.7431224811634835e-05,
+      "loss": 1.5975,
+      "step": 8286
+    },
+    {
+      "epoch": 2.177685814694387,
+      "grad_norm": 0.5801669955253601,
+      "learning_rate": 2.7413702470650083e-05,
+      "loss": 1.6318,
+      "step": 8288
+    },
+    {
+      "epoch": 2.178211318028049,
+      "grad_norm": 0.6386235356330872,
+      "learning_rate": 2.739618012966532e-05,
+      "loss": 1.6193,
+      "step": 8290
+    },
+    {
+      "epoch": 2.1787368213617104,
+      "grad_norm": 0.5952389240264893,
+      "learning_rate": 2.7378657788680568e-05,
+      "loss": 1.5982,
+      "step": 8292
+    },
+    {
+      "epoch": 2.1792623246953724,
+      "grad_norm": 0.5205252170562744,
+      "learning_rate": 2.7361135447695812e-05,
+      "loss": 1.6088,
+      "step": 8294
+    },
+    {
+      "epoch": 2.179787828029034,
+      "grad_norm": 0.5996859669685364,
+      "learning_rate": 2.7343613106711056e-05,
+      "loss": 1.5929,
+      "step": 8296
+    },
+    {
+      "epoch": 2.180313331362696,
+      "grad_norm": 0.5440699458122253,
+      "learning_rate": 2.7326090765726304e-05,
+      "loss": 1.6266,
+      "step": 8298
+    },
+    {
+      "epoch": 2.1808388346963574,
+      "grad_norm": 0.5174278616905212,
+      "learning_rate": 2.7308568424741548e-05,
+      "loss": 1.6134,
+      "step": 8300
+    },
+    {
+      "epoch": 2.1813643380300194,
+      "grad_norm": 0.5336670875549316,
+      "learning_rate": 2.7291046083756792e-05,
+      "loss": 1.6026,
+      "step": 8302
+    },
+    {
+      "epoch": 2.181889841363681,
+      "grad_norm": 0.5199602246284485,
+      "learning_rate": 2.7273523742772033e-05,
+      "loss": 1.6117,
+      "step": 8304
+    },
+    {
+      "epoch": 2.182415344697343,
+      "grad_norm": 0.5980170369148254,
+      "learning_rate": 2.7256001401787277e-05,
+      "loss": 1.6027,
+      "step": 8306
+    },
+    {
+      "epoch": 2.182940848031005,
+      "grad_norm": 0.6026611924171448,
+      "learning_rate": 2.7238479060802525e-05,
+      "loss": 1.6326,
+      "step": 8308
+    },
+    {
+      "epoch": 2.1834663513646664,
+      "grad_norm": 0.54527747631073,
+      "learning_rate": 2.722095671981777e-05,
+      "loss": 1.587,
+      "step": 8310
+    },
+    {
+      "epoch": 2.1839918546983283,
+      "grad_norm": 0.6268514394760132,
+      "learning_rate": 2.7203434378833013e-05,
+      "loss": 1.6362,
+      "step": 8312
+    },
+    {
+      "epoch": 2.18451735803199,
+      "grad_norm": 0.6219541430473328,
+      "learning_rate": 2.718591203784826e-05,
+      "loss": 1.5783,
+      "step": 8314
+    },
+    {
+      "epoch": 2.185042861365652,
+      "grad_norm": 0.5530925393104553,
+      "learning_rate": 2.71683896968635e-05,
+      "loss": 1.6121,
+      "step": 8316
+    },
+    {
+      "epoch": 2.1855683646993134,
+      "grad_norm": 0.6584067940711975,
+      "learning_rate": 2.7150867355878746e-05,
+      "loss": 1.6114,
+      "step": 8318
+    },
+    {
+      "epoch": 2.1860938680329753,
+      "grad_norm": 0.5550515055656433,
+      "learning_rate": 2.713334501489399e-05,
+      "loss": 1.6087,
+      "step": 8320
+    },
+    {
+      "epoch": 2.1866193713666373,
+      "grad_norm": 0.595063328742981,
+      "learning_rate": 2.7115822673909237e-05,
+      "loss": 1.606,
+      "step": 8322
+    },
+    {
+      "epoch": 2.187144874700299,
+      "grad_norm": 0.6574273705482483,
+      "learning_rate": 2.709830033292448e-05,
+      "loss": 1.6054,
+      "step": 8324
+    },
+    {
+      "epoch": 2.187670378033961,
+      "grad_norm": 0.5191264748573303,
+      "learning_rate": 2.7080777991939726e-05,
+      "loss": 1.5956,
+      "step": 8326
+    },
+    {
+      "epoch": 2.1881958813676223,
+      "grad_norm": 0.6227658987045288,
+      "learning_rate": 2.7063255650954967e-05,
+      "loss": 1.6053,
+      "step": 8328
+    },
+    {
+      "epoch": 2.1887213847012843,
+      "grad_norm": 0.6377692222595215,
+      "learning_rate": 2.704573330997021e-05,
+      "loss": 1.6253,
+      "step": 8330
+    },
+    {
+      "epoch": 2.189246888034946,
+      "grad_norm": 0.5611161589622498,
+      "learning_rate": 2.7028210968985458e-05,
+      "loss": 1.5996,
+      "step": 8332
+    },
+    {
+      "epoch": 2.189772391368608,
+      "grad_norm": 0.628909170627594,
+      "learning_rate": 2.7010688628000702e-05,
+      "loss": 1.6267,
+      "step": 8334
+    },
+    {
+      "epoch": 2.1902978947022693,
+      "grad_norm": 0.5963556170463562,
+      "learning_rate": 2.6993166287015947e-05,
+      "loss": 1.5845,
+      "step": 8336
+    },
+    {
+      "epoch": 2.1908233980359313,
+      "grad_norm": 0.5764312744140625,
+      "learning_rate": 2.6975643946031194e-05,
+      "loss": 1.593,
+      "step": 8338
+    },
+    {
+      "epoch": 2.1913489013695933,
+      "grad_norm": 0.7820791006088257,
+      "learning_rate": 2.695812160504644e-05,
+      "loss": 1.5945,
+      "step": 8340
+    },
+    {
+      "epoch": 2.191874404703255,
+      "grad_norm": 0.542460560798645,
+      "learning_rate": 2.694059926406168e-05,
+      "loss": 1.6183,
+      "step": 8342
+    },
+    {
+      "epoch": 2.1923999080369168,
+      "grad_norm": 0.6432899832725525,
+      "learning_rate": 2.6923076923076923e-05,
+      "loss": 1.6584,
+      "step": 8344
+    },
+    {
+      "epoch": 2.1929254113705783,
+      "grad_norm": 0.5868924856185913,
+      "learning_rate": 2.6905554582092167e-05,
+      "loss": 1.6196,
+      "step": 8346
+    },
+    {
+      "epoch": 2.1934509147042403,
+      "grad_norm": 0.5434276461601257,
+      "learning_rate": 2.6888032241107415e-05,
+      "loss": 1.581,
+      "step": 8348
+    },
+    {
+      "epoch": 2.1939764180379018,
+      "grad_norm": 0.5824474692344666,
+      "learning_rate": 2.687050990012266e-05,
+      "loss": 1.5816,
+      "step": 8350
+    },
+    {
+      "epoch": 2.1945019213715637,
+      "grad_norm": 0.5336950421333313,
+      "learning_rate": 2.6852987559137903e-05,
+      "loss": 1.574,
+      "step": 8352
+    },
+    {
+      "epoch": 2.1950274247052253,
+      "grad_norm": 0.5753375291824341,
+      "learning_rate": 2.6835465218153144e-05,
+      "loss": 1.604,
+      "step": 8354
+    },
+    {
+      "epoch": 2.1955529280388872,
+      "grad_norm": 0.762776792049408,
+      "learning_rate": 2.681794287716839e-05,
+      "loss": 1.6355,
+      "step": 8356
+    },
+    {
+      "epoch": 2.196078431372549,
+      "grad_norm": 0.5597438812255859,
+      "learning_rate": 2.6800420536183636e-05,
+      "loss": 1.5729,
+      "step": 8358
+    },
+    {
+      "epoch": 2.1966039347062107,
+      "grad_norm": 0.5700204968452454,
+      "learning_rate": 2.678289819519888e-05,
+      "loss": 1.6107,
+      "step": 8360
+    },
+    {
+      "epoch": 2.1971294380398727,
+      "grad_norm": 0.5418670773506165,
+      "learning_rate": 2.6765375854214124e-05,
+      "loss": 1.6243,
+      "step": 8362
+    },
+    {
+      "epoch": 2.1976549413735342,
+      "grad_norm": 0.5794771909713745,
+      "learning_rate": 2.6747853513229372e-05,
+      "loss": 1.6099,
+      "step": 8364
+    },
+    {
+      "epoch": 2.198180444707196,
+      "grad_norm": 0.6689446568489075,
+      "learning_rate": 2.6730331172244616e-05,
+      "loss": 1.5979,
+      "step": 8366
+    },
+    {
+      "epoch": 2.1987059480408577,
+      "grad_norm": 0.6027274131774902,
+      "learning_rate": 2.6712808831259857e-05,
+      "loss": 1.6143,
+      "step": 8368
+    },
+    {
+      "epoch": 2.1992314513745197,
+      "grad_norm": 0.5987696051597595,
+      "learning_rate": 2.66952864902751e-05,
+      "loss": 1.5895,
+      "step": 8370
+    },
+    {
+      "epoch": 2.1997569547081812,
+      "grad_norm": 0.566920816898346,
+      "learning_rate": 2.6677764149290345e-05,
+      "loss": 1.6106,
+      "step": 8372
+    },
+    {
+      "epoch": 2.200282458041843,
+      "grad_norm": 0.611470639705658,
+      "learning_rate": 2.6660241808305593e-05,
+      "loss": 1.5572,
+      "step": 8374
+    },
+    {
+      "epoch": 2.200807961375505,
+      "grad_norm": 0.666123628616333,
+      "learning_rate": 2.6642719467320837e-05,
+      "loss": 1.6263,
+      "step": 8376
+    },
+    {
+      "epoch": 2.2013334647091667,
+      "grad_norm": 0.522304117679596,
+      "learning_rate": 2.662519712633608e-05,
+      "loss": 1.6031,
+      "step": 8378
+    },
+    {
+      "epoch": 2.2018589680428287,
+      "grad_norm": 0.518319845199585,
+      "learning_rate": 2.6607674785351322e-05,
+      "loss": 1.5743,
+      "step": 8380
+    },
+    {
+      "epoch": 2.20238447137649,
+      "grad_norm": 0.547029435634613,
+      "learning_rate": 2.6590152444366566e-05,
+      "loss": 1.6111,
+      "step": 8382
+    },
+    {
+      "epoch": 2.202909974710152,
+      "grad_norm": 0.5654579401016235,
+      "learning_rate": 2.6572630103381814e-05,
+      "loss": 1.6057,
+      "step": 8384
+    },
+    {
+      "epoch": 2.2034354780438137,
+      "grad_norm": 0.6351662874221802,
+      "learning_rate": 2.6555107762397058e-05,
+      "loss": 1.6298,
+      "step": 8386
+    },
+    {
+      "epoch": 2.2039609813774756,
+      "grad_norm": 0.5234793424606323,
+      "learning_rate": 2.6537585421412302e-05,
+      "loss": 1.5897,
+      "step": 8388
+    },
+    {
+      "epoch": 2.204486484711137,
+      "grad_norm": 0.6238772869110107,
+      "learning_rate": 2.652006308042755e-05,
+      "loss": 1.6321,
+      "step": 8390
+    },
+    {
+      "epoch": 2.205011988044799,
+      "grad_norm": 0.5562398433685303,
+      "learning_rate": 2.6502540739442787e-05,
+      "loss": 1.575,
+      "step": 8392
+    },
+    {
+      "epoch": 2.205537491378461,
+      "grad_norm": 0.5776354670524597,
+      "learning_rate": 2.6485018398458034e-05,
+      "loss": 1.5714,
+      "step": 8394
+    },
+    {
+      "epoch": 2.2060629947121226,
+      "grad_norm": 0.5097812414169312,
+      "learning_rate": 2.646749605747328e-05,
+      "loss": 1.6081,
+      "step": 8396
+    },
+    {
+      "epoch": 2.2065884980457846,
+      "grad_norm": 0.5820745825767517,
+      "learning_rate": 2.6449973716488523e-05,
+      "loss": 1.646,
+      "step": 8398
+    },
+    {
+      "epoch": 2.207114001379446,
+      "grad_norm": 0.5649163722991943,
+      "learning_rate": 2.643245137550377e-05,
+      "loss": 1.6181,
+      "step": 8400
+    },
+    {
+      "epoch": 2.207114001379446,
+      "eval_loss": 1.6614795923233032,
+      "eval_runtime": 487.2799,
+      "eval_samples_per_second": 249.936,
+      "eval_steps_per_second": 31.243,
+      "step": 8400
+    },
+    {
+      "epoch": 2.207639504713108,
+      "grad_norm": 0.5877487063407898,
+      "learning_rate": 2.6414929034519015e-05,
+      "loss": 1.6239,
+      "step": 8402
+    },
+    {
+      "epoch": 2.2081650080467696,
+      "grad_norm": 0.5796919465065002,
+      "learning_rate": 2.639740669353426e-05,
+      "loss": 1.6233,
+      "step": 8404
+    },
+    {
+      "epoch": 2.2086905113804316,
+      "grad_norm": 0.58669114112854,
+      "learning_rate": 2.63798843525495e-05,
+      "loss": 1.6164,
+      "step": 8406
+    },
+    {
+      "epoch": 2.2092160147140936,
+      "grad_norm": 0.6143302917480469,
+      "learning_rate": 2.6362362011564744e-05,
+      "loss": 1.6609,
+      "step": 8408
+    },
+    {
+      "epoch": 2.209741518047755,
+      "grad_norm": 0.6326837539672852,
+      "learning_rate": 2.634483967057999e-05,
+      "loss": 1.5887,
+      "step": 8410
+    },
+    {
+      "epoch": 2.210267021381417,
+      "grad_norm": 0.5528976917266846,
+      "learning_rate": 2.6327317329595235e-05,
+      "loss": 1.5859,
+      "step": 8412
+    },
+    {
+      "epoch": 2.2107925247150786,
+      "grad_norm": 0.654974102973938,
+      "learning_rate": 2.630979498861048e-05,
+      "loss": 1.5918,
+      "step": 8414
+    },
+    {
+      "epoch": 2.2113180280487406,
+      "grad_norm": 0.5469368100166321,
+      "learning_rate": 2.6292272647625727e-05,
+      "loss": 1.6064,
+      "step": 8416
+    },
+    {
+      "epoch": 2.211843531382402,
+      "grad_norm": 0.5513100028038025,
+      "learning_rate": 2.6274750306640965e-05,
+      "loss": 1.5964,
+      "step": 8418
+    },
+    {
+      "epoch": 2.212369034716064,
+      "grad_norm": 0.5723447799682617,
+      "learning_rate": 2.6257227965656212e-05,
+      "loss": 1.5847,
+      "step": 8420
+    },
+    {
+      "epoch": 2.2128945380497256,
+      "grad_norm": 0.5903245806694031,
+      "learning_rate": 2.6239705624671456e-05,
+      "loss": 1.6176,
+      "step": 8422
+    },
+    {
+      "epoch": 2.2134200413833875,
+      "grad_norm": 0.7093879580497742,
+      "learning_rate": 2.62221832836867e-05,
+      "loss": 1.627,
+      "step": 8424
+    },
+    {
+      "epoch": 2.213945544717049,
+      "grad_norm": 0.6263571977615356,
+      "learning_rate": 2.6204660942701948e-05,
+      "loss": 1.606,
+      "step": 8426
+    },
+    {
+      "epoch": 2.214471048050711,
+      "grad_norm": 0.5385801196098328,
+      "learning_rate": 2.6187138601717192e-05,
+      "loss": 1.6006,
+      "step": 8428
+    },
+    {
+      "epoch": 2.214996551384373,
+      "grad_norm": 0.5380145311355591,
+      "learning_rate": 2.6169616260732433e-05,
+      "loss": 1.6179,
+      "step": 8430
+    },
+    {
+      "epoch": 2.2155220547180345,
+      "grad_norm": 0.6820893287658691,
+      "learning_rate": 2.6152093919747677e-05,
+      "loss": 1.5878,
+      "step": 8432
+    },
+    {
+      "epoch": 2.2160475580516965,
+      "grad_norm": 0.6620394587516785,
+      "learning_rate": 2.613457157876292e-05,
+      "loss": 1.6155,
+      "step": 8434
+    },
+    {
+      "epoch": 2.216573061385358,
+      "grad_norm": 0.5877991914749146,
+      "learning_rate": 2.611704923777817e-05,
+      "loss": 1.6346,
+      "step": 8436
+    },
+    {
+      "epoch": 2.21709856471902,
+      "grad_norm": 0.7124659419059753,
+      "learning_rate": 2.6099526896793413e-05,
+      "loss": 1.624,
+      "step": 8438
+    },
+    {
+      "epoch": 2.2176240680526815,
+      "grad_norm": 0.6077954769134521,
+      "learning_rate": 2.608200455580866e-05,
+      "loss": 1.606,
+      "step": 8440
+    },
+    {
+      "epoch": 2.2181495713863435,
+      "grad_norm": 0.7110419869422913,
+      "learning_rate": 2.6064482214823905e-05,
+      "loss": 1.5749,
+      "step": 8442
+    },
+    {
+      "epoch": 2.2186750747200055,
+      "grad_norm": 0.5183662176132202,
+      "learning_rate": 2.6046959873839146e-05,
+      "loss": 1.6047,
+      "step": 8444
+    },
+    {
+      "epoch": 2.219200578053667,
+      "grad_norm": 0.6769959926605225,
+      "learning_rate": 2.602943753285439e-05,
+      "loss": 1.6635,
+      "step": 8446
+    },
+    {
+      "epoch": 2.219726081387329,
+      "grad_norm": 0.610445499420166,
+      "learning_rate": 2.6011915191869634e-05,
+      "loss": 1.6053,
+      "step": 8448
+    },
+    {
+      "epoch": 2.2202515847209905,
+      "grad_norm": 0.537892758846283,
+      "learning_rate": 2.599439285088488e-05,
+      "loss": 1.6046,
+      "step": 8450
+    },
+    {
+      "epoch": 2.2207770880546525,
+      "grad_norm": 0.5192505121231079,
+      "learning_rate": 2.5976870509900126e-05,
+      "loss": 1.5956,
+      "step": 8452
+    },
+    {
+      "epoch": 2.221302591388314,
+      "grad_norm": 0.5772652626037598,
+      "learning_rate": 2.595934816891537e-05,
+      "loss": 1.6013,
+      "step": 8454
+    },
+    {
+      "epoch": 2.221828094721976,
+      "grad_norm": 0.5051235556602478,
+      "learning_rate": 2.594182582793061e-05,
+      "loss": 1.5965,
+      "step": 8456
+    },
+    {
+      "epoch": 2.2223535980556375,
+      "grad_norm": 0.5779315233230591,
+      "learning_rate": 2.5924303486945855e-05,
+      "loss": 1.5888,
+      "step": 8458
+    },
+    {
+      "epoch": 2.2228791013892994,
+      "grad_norm": 0.5708422660827637,
+      "learning_rate": 2.5906781145961102e-05,
+      "loss": 1.638,
+      "step": 8460
+    },
+    {
+      "epoch": 2.223404604722961,
+      "grad_norm": 0.6042600274085999,
+      "learning_rate": 2.5889258804976347e-05,
+      "loss": 1.5939,
+      "step": 8462
+    },
+    {
+      "epoch": 2.223930108056623,
+      "grad_norm": 0.5264221429824829,
+      "learning_rate": 2.587173646399159e-05,
+      "loss": 1.6029,
+      "step": 8464
+    },
+    {
+      "epoch": 2.224455611390285,
+      "grad_norm": 0.5832411050796509,
+      "learning_rate": 2.585421412300684e-05,
+      "loss": 1.6332,
+      "step": 8466
+    },
+    {
+      "epoch": 2.2249811147239464,
+      "grad_norm": 0.5503547191619873,
+      "learning_rate": 2.5836691782022076e-05,
+      "loss": 1.6139,
+      "step": 8468
+    },
+    {
+      "epoch": 2.2255066180576084,
+      "grad_norm": 0.6002954244613647,
+      "learning_rate": 2.5819169441037323e-05,
+      "loss": 1.6085,
+      "step": 8470
+    },
+    {
+      "epoch": 2.22603212139127,
+      "grad_norm": 0.5220708250999451,
+      "learning_rate": 2.5801647100052567e-05,
+      "loss": 1.5779,
+      "step": 8472
+    },
+    {
+      "epoch": 2.226557624724932,
+      "grad_norm": 0.5423254370689392,
+      "learning_rate": 2.578412475906781e-05,
+      "loss": 1.5818,
+      "step": 8474
+    },
+    {
+      "epoch": 2.2270831280585934,
+      "grad_norm": 0.5738494396209717,
+      "learning_rate": 2.576660241808306e-05,
+      "loss": 1.6046,
+      "step": 8476
+    },
+    {
+      "epoch": 2.2276086313922554,
+      "grad_norm": 0.5414014458656311,
+      "learning_rate": 2.5749080077098303e-05,
+      "loss": 1.6238,
+      "step": 8478
+    },
+    {
+      "epoch": 2.2281341347259174,
+      "grad_norm": 0.5559380650520325,
+      "learning_rate": 2.5731557736113548e-05,
+      "loss": 1.6421,
+      "step": 8480
+    },
+    {
+      "epoch": 2.228659638059579,
+      "grad_norm": 0.5835655331611633,
+      "learning_rate": 2.571403539512879e-05,
+      "loss": 1.6208,
+      "step": 8482
+    },
+    {
+      "epoch": 2.229185141393241,
+      "grad_norm": 0.5489543676376343,
+      "learning_rate": 2.5696513054144033e-05,
+      "loss": 1.5749,
+      "step": 8484
+    },
+    {
+      "epoch": 2.2297106447269024,
+      "grad_norm": 0.5658118724822998,
+      "learning_rate": 2.567899071315928e-05,
+      "loss": 1.6382,
+      "step": 8486
+    },
+    {
+      "epoch": 2.2302361480605644,
+      "grad_norm": 0.5926935076713562,
+      "learning_rate": 2.5661468372174524e-05,
+      "loss": 1.6327,
+      "step": 8488
+    },
+    {
+      "epoch": 2.230761651394226,
+      "grad_norm": 0.5220993757247925,
+      "learning_rate": 2.564394603118977e-05,
+      "loss": 1.6266,
+      "step": 8490
+    },
+    {
+      "epoch": 2.231287154727888,
+      "grad_norm": 0.5491675138473511,
+      "learning_rate": 2.5626423690205016e-05,
+      "loss": 1.615,
+      "step": 8492
+    },
+    {
+      "epoch": 2.2318126580615494,
+      "grad_norm": 0.5789914131164551,
+      "learning_rate": 2.5608901349220253e-05,
+      "loss": 1.6568,
+      "step": 8494
+    },
+    {
+      "epoch": 2.2323381613952114,
+      "grad_norm": 0.5792714357376099,
+      "learning_rate": 2.55913790082355e-05,
+      "loss": 1.6344,
+      "step": 8496
+    },
+    {
+      "epoch": 2.2328636647288733,
+      "grad_norm": 0.4977031648159027,
+      "learning_rate": 2.5573856667250745e-05,
+      "loss": 1.6286,
+      "step": 8498
+    },
+    {
+      "epoch": 2.233389168062535,
+      "grad_norm": 0.6100000739097595,
+      "learning_rate": 2.555633432626599e-05,
+      "loss": 1.6234,
+      "step": 8500
+    },
+    {
+      "epoch": 2.233914671396197,
+      "grad_norm": 0.6163233518600464,
+      "learning_rate": 2.5538811985281237e-05,
+      "loss": 1.6096,
+      "step": 8502
+    },
+    {
+      "epoch": 2.2344401747298583,
+      "grad_norm": 0.5867898464202881,
+      "learning_rate": 2.552128964429648e-05,
+      "loss": 1.604,
+      "step": 8504
+    },
+    {
+      "epoch": 2.2349656780635203,
+      "grad_norm": 0.8214870095252991,
+      "learning_rate": 2.5503767303311725e-05,
+      "loss": 1.6476,
+      "step": 8506
+    },
+    {
+      "epoch": 2.235491181397182,
+      "grad_norm": 0.7553618550300598,
+      "learning_rate": 2.5486244962326966e-05,
+      "loss": 1.588,
+      "step": 8508
+    },
+    {
+      "epoch": 2.236016684730844,
+      "grad_norm": 0.5797671675682068,
+      "learning_rate": 2.546872262134221e-05,
+      "loss": 1.6125,
+      "step": 8510
+    },
+    {
+      "epoch": 2.2365421880645053,
+      "grad_norm": 0.6248254179954529,
+      "learning_rate": 2.5451200280357458e-05,
+      "loss": 1.6339,
+      "step": 8512
+    },
+    {
+      "epoch": 2.2370676913981673,
+      "grad_norm": 0.5321136116981506,
+      "learning_rate": 2.5433677939372702e-05,
+      "loss": 1.6058,
+      "step": 8514
+    },
+    {
+      "epoch": 2.2375931947318293,
+      "grad_norm": 0.685704231262207,
+      "learning_rate": 2.5416155598387946e-05,
+      "loss": 1.6305,
+      "step": 8516
+    },
+    {
+      "epoch": 2.238118698065491,
+      "grad_norm": 0.7109000086784363,
+      "learning_rate": 2.5398633257403194e-05,
+      "loss": 1.6176,
+      "step": 8518
+    },
+    {
+      "epoch": 2.2386442013991528,
+      "grad_norm": 0.5430434346199036,
+      "learning_rate": 2.538111091641843e-05,
+      "loss": 1.5943,
+      "step": 8520
+    },
+    {
+      "epoch": 2.2391697047328143,
+      "grad_norm": 0.5879561305046082,
+      "learning_rate": 2.536358857543368e-05,
+      "loss": 1.6418,
+      "step": 8522
+    },
+    {
+      "epoch": 2.2396952080664763,
+      "grad_norm": 0.5790303349494934,
+      "learning_rate": 2.5346066234448923e-05,
+      "loss": 1.6413,
+      "step": 8524
+    },
+    {
+      "epoch": 2.240220711400138,
+      "grad_norm": 0.5287625193595886,
+      "learning_rate": 2.5328543893464167e-05,
+      "loss": 1.5898,
+      "step": 8526
+    },
+    {
+      "epoch": 2.2407462147337998,
+      "grad_norm": 0.6279188394546509,
+      "learning_rate": 2.5311021552479415e-05,
+      "loss": 1.6066,
+      "step": 8528
+    },
+    {
+      "epoch": 2.2412717180674613,
+      "grad_norm": 0.5477473139762878,
+      "learning_rate": 2.529349921149466e-05,
+      "loss": 1.6004,
+      "step": 8530
+    },
+    {
+      "epoch": 2.2417972214011233,
+      "grad_norm": 0.5529081225395203,
+      "learning_rate": 2.52759768705099e-05,
+      "loss": 1.6265,
+      "step": 8532
+    },
+    {
+      "epoch": 2.242322724734785,
+      "grad_norm": 0.6740888357162476,
+      "learning_rate": 2.5258454529525144e-05,
+      "loss": 1.6015,
+      "step": 8534
+    },
+    {
+      "epoch": 2.2428482280684467,
+      "grad_norm": 0.5619795322418213,
+      "learning_rate": 2.5240932188540388e-05,
+      "loss": 1.6388,
+      "step": 8536
+    },
+    {
+      "epoch": 2.2433737314021087,
+      "grad_norm": 0.5655804872512817,
+      "learning_rate": 2.5223409847555635e-05,
+      "loss": 1.6052,
+      "step": 8538
+    },
+    {
+      "epoch": 2.2438992347357702,
+      "grad_norm": 0.5193714499473572,
+      "learning_rate": 2.520588750657088e-05,
+      "loss": 1.6223,
+      "step": 8540
+    },
+    {
+      "epoch": 2.244424738069432,
+      "grad_norm": 0.5750133991241455,
+      "learning_rate": 2.5188365165586124e-05,
+      "loss": 1.6025,
+      "step": 8542
+    },
+    {
+      "epoch": 2.2449502414030937,
+      "grad_norm": 0.673194169998169,
+      "learning_rate": 2.517084282460137e-05,
+      "loss": 1.6275,
+      "step": 8544
+    },
+    {
+      "epoch": 2.2454757447367557,
+      "grad_norm": 0.5762125849723816,
+      "learning_rate": 2.515332048361661e-05,
+      "loss": 1.6061,
+      "step": 8546
+    },
+    {
+      "epoch": 2.2460012480704172,
+      "grad_norm": 0.5318456292152405,
+      "learning_rate": 2.5135798142631856e-05,
+      "loss": 1.6048,
+      "step": 8548
+    },
+    {
+      "epoch": 2.246526751404079,
+      "grad_norm": 0.5658667683601379,
+      "learning_rate": 2.51182758016471e-05,
+      "loss": 1.5761,
+      "step": 8550
+    },
+    {
+      "epoch": 2.247052254737741,
+      "grad_norm": 0.5093829035758972,
+      "learning_rate": 2.5100753460662345e-05,
+      "loss": 1.5823,
+      "step": 8552
+    },
+    {
+      "epoch": 2.2475777580714027,
+      "grad_norm": 0.6146730780601501,
+      "learning_rate": 2.5083231119677592e-05,
+      "loss": 1.6037,
+      "step": 8554
+    },
+    {
+      "epoch": 2.2481032614050647,
+      "grad_norm": 0.5599241256713867,
+      "learning_rate": 2.5065708778692836e-05,
+      "loss": 1.6559,
+      "step": 8556
+    },
+    {
+      "epoch": 2.248628764738726,
+      "grad_norm": 0.6648929119110107,
+      "learning_rate": 2.5048186437708077e-05,
+      "loss": 1.6054,
+      "step": 8558
+    },
+    {
+      "epoch": 2.249154268072388,
+      "grad_norm": 0.6260311603546143,
+      "learning_rate": 2.503066409672332e-05,
+      "loss": 1.6113,
+      "step": 8560
+    },
+    {
+      "epoch": 2.2496797714060497,
+      "grad_norm": 0.6752457022666931,
+      "learning_rate": 2.5013141755738566e-05,
+      "loss": 1.6019,
+      "step": 8562
+    },
+    {
+      "epoch": 2.2502052747397117,
+      "grad_norm": 0.5815406441688538,
+      "learning_rate": 2.4995619414753813e-05,
+      "loss": 1.6158,
+      "step": 8564
+    },
+    {
+      "epoch": 2.2507307780733736,
+      "grad_norm": 0.7072321772575378,
+      "learning_rate": 2.4978097073769057e-05,
+      "loss": 1.6549,
+      "step": 8566
+    },
+    {
+      "epoch": 2.251256281407035,
+      "grad_norm": 0.5526368021965027,
+      "learning_rate": 2.49605747327843e-05,
+      "loss": 1.6122,
+      "step": 8568
+    },
+    {
+      "epoch": 2.251781784740697,
+      "grad_norm": 0.644477128982544,
+      "learning_rate": 2.4943052391799546e-05,
+      "loss": 1.5802,
+      "step": 8570
+    },
+    {
+      "epoch": 2.2523072880743586,
+      "grad_norm": 0.5599563717842102,
+      "learning_rate": 2.492553005081479e-05,
+      "loss": 1.6255,
+      "step": 8572
+    },
+    {
+      "epoch": 2.2528327914080206,
+      "grad_norm": 0.5871409177780151,
+      "learning_rate": 2.4908007709830034e-05,
+      "loss": 1.5929,
+      "step": 8574
+    },
+    {
+      "epoch": 2.253358294741682,
+      "grad_norm": 0.654251754283905,
+      "learning_rate": 2.4890485368845278e-05,
+      "loss": 1.6135,
+      "step": 8576
+    },
+    {
+      "epoch": 2.253883798075344,
+      "grad_norm": 0.5650330185890198,
+      "learning_rate": 2.4872963027860526e-05,
+      "loss": 1.5967,
+      "step": 8578
+    },
+    {
+      "epoch": 2.2544093014090056,
+      "grad_norm": 0.5818485021591187,
+      "learning_rate": 2.4855440686875766e-05,
+      "loss": 1.642,
+      "step": 8580
+    },
+    {
+      "epoch": 2.2549348047426676,
+      "grad_norm": 0.5410641431808472,
+      "learning_rate": 2.483791834589101e-05,
+      "loss": 1.6136,
+      "step": 8582
+    },
+    {
+      "epoch": 2.255460308076329,
+      "grad_norm": 0.5259566903114319,
+      "learning_rate": 2.4820396004906258e-05,
+      "loss": 1.6212,
+      "step": 8584
+    },
+    {
+      "epoch": 2.255985811409991,
+      "grad_norm": 0.5885828733444214,
+      "learning_rate": 2.48028736639215e-05,
+      "loss": 1.6015,
+      "step": 8586
+    },
+    {
+      "epoch": 2.256511314743653,
+      "grad_norm": 0.6429271697998047,
+      "learning_rate": 2.4785351322936747e-05,
+      "loss": 1.621,
+      "step": 8588
+    },
+    {
+      "epoch": 2.2570368180773146,
+      "grad_norm": 0.5706207156181335,
+      "learning_rate": 2.476782898195199e-05,
+      "loss": 1.5826,
+      "step": 8590
+    },
+    {
+      "epoch": 2.2575623214109766,
+      "grad_norm": 0.6020461320877075,
+      "learning_rate": 2.4750306640967235e-05,
+      "loss": 1.6296,
+      "step": 8592
+    },
+    {
+      "epoch": 2.258087824744638,
+      "grad_norm": 0.5262504816055298,
+      "learning_rate": 2.473278429998248e-05,
+      "loss": 1.6015,
+      "step": 8594
+    },
+    {
+      "epoch": 2.2586133280783,
+      "grad_norm": 0.6201303005218506,
+      "learning_rate": 2.4715261958997723e-05,
+      "loss": 1.5931,
+      "step": 8596
+    },
+    {
+      "epoch": 2.2591388314119616,
+      "grad_norm": 0.5339419841766357,
+      "learning_rate": 2.4697739618012967e-05,
+      "loss": 1.6369,
+      "step": 8598
+    },
+    {
+      "epoch": 2.2596643347456236,
+      "grad_norm": 0.6167606115341187,
+      "learning_rate": 2.468021727702821e-05,
+      "loss": 1.6056,
+      "step": 8600
+    },
+    {
+      "epoch": 2.2601898380792855,
+      "grad_norm": 0.6669736504554749,
+      "learning_rate": 2.4662694936043456e-05,
+      "loss": 1.6262,
+      "step": 8602
+    },
+    {
+      "epoch": 2.260715341412947,
+      "grad_norm": 0.6823409795761108,
+      "learning_rate": 2.4645172595058703e-05,
+      "loss": 1.5777,
+      "step": 8604
+    },
+    {
+      "epoch": 2.261240844746609,
+      "grad_norm": 0.5810272097587585,
+      "learning_rate": 2.4627650254073944e-05,
+      "loss": 1.6163,
+      "step": 8606
+    },
+    {
+      "epoch": 2.2617663480802706,
+      "grad_norm": 0.5693057179450989,
+      "learning_rate": 2.461012791308919e-05,
+      "loss": 1.6286,
+      "step": 8608
+    },
+    {
+      "epoch": 2.2622918514139325,
+      "grad_norm": 0.6601256132125854,
+      "learning_rate": 2.4592605572104436e-05,
+      "loss": 1.6179,
+      "step": 8610
+    },
+    {
+      "epoch": 2.262817354747594,
+      "grad_norm": 0.5781053304672241,
+      "learning_rate": 2.4575083231119677e-05,
+      "loss": 1.6364,
+      "step": 8612
+    },
+    {
+      "epoch": 2.263342858081256,
+      "grad_norm": 0.7442349195480347,
+      "learning_rate": 2.4557560890134924e-05,
+      "loss": 1.6105,
+      "step": 8614
+    },
+    {
+      "epoch": 2.2638683614149175,
+      "grad_norm": 0.6304720044136047,
+      "learning_rate": 2.454003854915017e-05,
+      "loss": 1.6368,
+      "step": 8616
+    },
+    {
+      "epoch": 2.2643938647485795,
+      "grad_norm": 0.8399868607521057,
+      "learning_rate": 2.452251620816541e-05,
+      "loss": 1.6196,
+      "step": 8618
+    },
+    {
+      "epoch": 2.264919368082241,
+      "grad_norm": 0.6608265042304993,
+      "learning_rate": 2.4504993867180657e-05,
+      "loss": 1.6327,
+      "step": 8620
+    },
+    {
+      "epoch": 2.265444871415903,
+      "grad_norm": 0.5849167108535767,
+      "learning_rate": 2.44874715261959e-05,
+      "loss": 1.6294,
+      "step": 8622
+    },
+    {
+      "epoch": 2.265970374749565,
+      "grad_norm": 0.5310716032981873,
+      "learning_rate": 2.4469949185211145e-05,
+      "loss": 1.5839,
+      "step": 8624
+    },
+    {
+      "epoch": 2.2664958780832265,
+      "grad_norm": 0.5808155536651611,
+      "learning_rate": 2.445242684422639e-05,
+      "loss": 1.6326,
+      "step": 8626
+    },
+    {
+      "epoch": 2.2670213814168885,
+      "grad_norm": 0.6455567479133606,
+      "learning_rate": 2.4434904503241633e-05,
+      "loss": 1.6199,
+      "step": 8628
+    },
+    {
+      "epoch": 2.26754688475055,
+      "grad_norm": 0.7273616790771484,
+      "learning_rate": 2.441738216225688e-05,
+      "loss": 1.6067,
+      "step": 8630
+    },
+    {
+      "epoch": 2.268072388084212,
+      "grad_norm": 0.7093321681022644,
+      "learning_rate": 2.4399859821272122e-05,
+      "loss": 1.6274,
+      "step": 8632
+    },
+    {
+      "epoch": 2.2685978914178735,
+      "grad_norm": 0.5976110696792603,
+      "learning_rate": 2.438233748028737e-05,
+      "loss": 1.6328,
+      "step": 8634
+    },
+    {
+      "epoch": 2.2691233947515355,
+      "grad_norm": 0.6373220682144165,
+      "learning_rate": 2.4364815139302614e-05,
+      "loss": 1.6248,
+      "step": 8636
+    },
+    {
+      "epoch": 2.2696488980851974,
+      "grad_norm": 0.5636451244354248,
+      "learning_rate": 2.4347292798317854e-05,
+      "loss": 1.6133,
+      "step": 8638
+    },
+    {
+      "epoch": 2.270174401418859,
+      "grad_norm": 0.6106368899345398,
+      "learning_rate": 2.4329770457333102e-05,
+      "loss": 1.6036,
+      "step": 8640
+    },
+    {
+      "epoch": 2.270699904752521,
+      "grad_norm": 0.526772677898407,
+      "learning_rate": 2.4312248116348346e-05,
+      "loss": 1.5967,
+      "step": 8642
+    },
+    {
+      "epoch": 2.2712254080861825,
+      "grad_norm": 0.5734396576881409,
+      "learning_rate": 2.429472577536359e-05,
+      "loss": 1.6056,
+      "step": 8644
+    },
+    {
+      "epoch": 2.2717509114198444,
+      "grad_norm": 0.5490990281105042,
+      "learning_rate": 2.4277203434378834e-05,
+      "loss": 1.6057,
+      "step": 8646
+    },
+    {
+      "epoch": 2.272276414753506,
+      "grad_norm": 0.5757094621658325,
+      "learning_rate": 2.425968109339408e-05,
+      "loss": 1.6127,
+      "step": 8648
+    },
+    {
+      "epoch": 2.272801918087168,
+      "grad_norm": 0.5849709510803223,
+      "learning_rate": 2.4242158752409323e-05,
+      "loss": 1.6182,
+      "step": 8650
+    },
+    {
+      "epoch": 2.27332742142083,
+      "grad_norm": 0.6258499026298523,
+      "learning_rate": 2.4224636411424567e-05,
+      "loss": 1.5819,
+      "step": 8652
+    },
+    {
+      "epoch": 2.2738529247544914,
+      "grad_norm": 0.5439596176147461,
+      "learning_rate": 2.420711407043981e-05,
+      "loss": 1.6197,
+      "step": 8654
+    },
+    {
+      "epoch": 2.274378428088153,
+      "grad_norm": 0.5642632246017456,
+      "learning_rate": 2.4189591729455055e-05,
+      "loss": 1.619,
+      "step": 8656
+    },
+    {
+      "epoch": 2.274903931421815,
+      "grad_norm": 0.5832663178443909,
+      "learning_rate": 2.41720693884703e-05,
+      "loss": 1.6291,
+      "step": 8658
+    },
+    {
+      "epoch": 2.275429434755477,
+      "grad_norm": 0.5677375197410583,
+      "learning_rate": 2.4154547047485547e-05,
+      "loss": 1.6087,
+      "step": 8660
+    },
+    {
+      "epoch": 2.2759549380891384,
+      "grad_norm": 0.5964575409889221,
+      "learning_rate": 2.413702470650079e-05,
+      "loss": 1.5785,
+      "step": 8662
+    },
+    {
+      "epoch": 2.2764804414228004,
+      "grad_norm": 0.6877849102020264,
+      "learning_rate": 2.4119502365516032e-05,
+      "loss": 1.6554,
+      "step": 8664
+    },
+    {
+      "epoch": 2.277005944756462,
+      "grad_norm": 0.6050941348075867,
+      "learning_rate": 2.410198002453128e-05,
+      "loss": 1.6324,
+      "step": 8666
+    },
+    {
+      "epoch": 2.277531448090124,
+      "grad_norm": 0.6498258113861084,
+      "learning_rate": 2.4084457683546524e-05,
+      "loss": 1.6018,
+      "step": 8668
+    },
+    {
+      "epoch": 2.2780569514237854,
+      "grad_norm": 0.7161571979522705,
+      "learning_rate": 2.4066935342561768e-05,
+      "loss": 1.6233,
+      "step": 8670
+    },
+    {
+      "epoch": 2.2785824547574474,
+      "grad_norm": 0.6215277910232544,
+      "learning_rate": 2.4049413001577012e-05,
+      "loss": 1.5664,
+      "step": 8672
+    },
+    {
+      "epoch": 2.2791079580911093,
+      "grad_norm": 0.6165092587471008,
+      "learning_rate": 2.4031890660592256e-05,
+      "loss": 1.6217,
+      "step": 8674
+    },
+    {
+      "epoch": 2.279633461424771,
+      "grad_norm": 0.5820355415344238,
+      "learning_rate": 2.40143683196075e-05,
+      "loss": 1.631,
+      "step": 8676
+    },
+    {
+      "epoch": 2.280158964758433,
+      "grad_norm": 0.5764251351356506,
+      "learning_rate": 2.3996845978622745e-05,
+      "loss": 1.6371,
+      "step": 8678
+    },
+    {
+      "epoch": 2.2806844680920944,
+      "grad_norm": 0.5712262988090515,
+      "learning_rate": 2.397932363763799e-05,
+      "loss": 1.6164,
+      "step": 8680
+    },
+    {
+      "epoch": 2.2812099714257563,
+      "grad_norm": 0.5973013639450073,
+      "learning_rate": 2.3961801296653233e-05,
+      "loss": 1.6543,
+      "step": 8682
+    },
+    {
+      "epoch": 2.281735474759418,
+      "grad_norm": 0.5426216721534729,
+      "learning_rate": 2.3944278955668477e-05,
+      "loss": 1.5761,
+      "step": 8684
+    },
+    {
+      "epoch": 2.28226097809308,
+      "grad_norm": 0.5732000470161438,
+      "learning_rate": 2.3926756614683725e-05,
+      "loss": 1.6052,
+      "step": 8686
+    },
+    {
+      "epoch": 2.282786481426742,
+      "grad_norm": 0.581996500492096,
+      "learning_rate": 2.3909234273698966e-05,
+      "loss": 1.6382,
+      "step": 8688
+    },
+    {
+      "epoch": 2.2833119847604033,
+      "grad_norm": 0.5811718702316284,
+      "learning_rate": 2.3891711932714213e-05,
+      "loss": 1.6127,
+      "step": 8690
+    },
+    {
+      "epoch": 2.2838374880940653,
+      "grad_norm": 0.59296053647995,
+      "learning_rate": 2.3874189591729457e-05,
+      "loss": 1.6458,
+      "step": 8692
+    },
+    {
+      "epoch": 2.284362991427727,
+      "grad_norm": 0.4986303746700287,
+      "learning_rate": 2.38566672507447e-05,
+      "loss": 1.6373,
+      "step": 8694
+    },
+    {
+      "epoch": 2.2848884947613888,
+      "grad_norm": 0.711003839969635,
+      "learning_rate": 2.3839144909759946e-05,
+      "loss": 1.6123,
+      "step": 8696
+    },
+    {
+      "epoch": 2.2854139980950503,
+      "grad_norm": 0.5871834754943848,
+      "learning_rate": 2.382162256877519e-05,
+      "loss": 1.6206,
+      "step": 8698
+    },
+    {
+      "epoch": 2.2859395014287123,
+      "grad_norm": 0.5751228928565979,
+      "learning_rate": 2.3804100227790434e-05,
+      "loss": 1.6307,
+      "step": 8700
+    },
+    {
+      "epoch": 2.286465004762374,
+      "grad_norm": 0.5439415574073792,
+      "learning_rate": 2.3786577886805678e-05,
+      "loss": 1.6081,
+      "step": 8702
+    },
+    {
+      "epoch": 2.2869905080960358,
+      "grad_norm": 0.6993831992149353,
+      "learning_rate": 2.3769055545820922e-05,
+      "loss": 1.6197,
+      "step": 8704
+    },
+    {
+      "epoch": 2.2875160114296973,
+      "grad_norm": 0.5735906958580017,
+      "learning_rate": 2.375153320483617e-05,
+      "loss": 1.6184,
+      "step": 8706
+    },
+    {
+      "epoch": 2.2880415147633593,
+      "grad_norm": 0.7263919115066528,
+      "learning_rate": 2.373401086385141e-05,
+      "loss": 1.626,
+      "step": 8708
+    },
+    {
+      "epoch": 2.2885670180970212,
+      "grad_norm": 0.6145097613334656,
+      "learning_rate": 2.3716488522866655e-05,
+      "loss": 1.5732,
+      "step": 8710
+    },
+    {
+      "epoch": 2.2890925214306828,
+      "grad_norm": 0.5744717717170715,
+      "learning_rate": 2.3698966181881902e-05,
+      "loss": 1.6301,
+      "step": 8712
+    },
+    {
+      "epoch": 2.2896180247643447,
+      "grad_norm": 0.5422975420951843,
+      "learning_rate": 2.3681443840897143e-05,
+      "loss": 1.6168,
+      "step": 8714
+    },
+    {
+      "epoch": 2.2901435280980063,
+      "grad_norm": 0.5488165020942688,
+      "learning_rate": 2.366392149991239e-05,
+      "loss": 1.6252,
+      "step": 8716
+    },
+    {
+      "epoch": 2.2906690314316682,
+      "grad_norm": 0.5232349634170532,
+      "learning_rate": 2.3646399158927635e-05,
+      "loss": 1.5893,
+      "step": 8718
+    },
+    {
+      "epoch": 2.2911945347653297,
+      "grad_norm": 0.5339794754981995,
+      "learning_rate": 2.3628876817942876e-05,
+      "loss": 1.6187,
+      "step": 8720
+    },
+    {
+      "epoch": 2.2917200380989917,
+      "grad_norm": 0.5548917651176453,
+      "learning_rate": 2.3611354476958123e-05,
+      "loss": 1.6204,
+      "step": 8722
+    },
+    {
+      "epoch": 2.2922455414326537,
+      "grad_norm": 0.5368824005126953,
+      "learning_rate": 2.3593832135973367e-05,
+      "loss": 1.6096,
+      "step": 8724
+    },
+    {
+      "epoch": 2.292771044766315,
+      "grad_norm": 0.581849217414856,
+      "learning_rate": 2.357630979498861e-05,
+      "loss": 1.6213,
+      "step": 8726
+    },
+    {
+      "epoch": 2.293296548099977,
+      "grad_norm": 0.5897862315177917,
+      "learning_rate": 2.3558787454003856e-05,
+      "loss": 1.6521,
+      "step": 8728
+    },
+    {
+      "epoch": 2.2938220514336387,
+      "grad_norm": 0.49220868945121765,
+      "learning_rate": 2.35412651130191e-05,
+      "loss": 1.5872,
+      "step": 8730
+    },
+    {
+      "epoch": 2.2943475547673007,
+      "grad_norm": 0.6717751622200012,
+      "learning_rate": 2.3523742772034348e-05,
+      "loss": 1.6153,
+      "step": 8732
+    },
+    {
+      "epoch": 2.294873058100962,
+      "grad_norm": 0.7588834166526794,
+      "learning_rate": 2.350622043104959e-05,
+      "loss": 1.6137,
+      "step": 8734
+    },
+    {
+      "epoch": 2.295398561434624,
+      "grad_norm": 0.6023603081703186,
+      "learning_rate": 2.3488698090064832e-05,
+      "loss": 1.5825,
+      "step": 8736
+    },
+    {
+      "epoch": 2.2959240647682857,
+      "grad_norm": 0.590947687625885,
+      "learning_rate": 2.347117574908008e-05,
+      "loss": 1.5805,
+      "step": 8738
+    },
+    {
+      "epoch": 2.2964495681019477,
+      "grad_norm": 0.5251069664955139,
+      "learning_rate": 2.345365340809532e-05,
+      "loss": 1.6111,
+      "step": 8740
+    },
+    {
+      "epoch": 2.296975071435609,
+      "grad_norm": 0.5737573504447937,
+      "learning_rate": 2.343613106711057e-05,
+      "loss": 1.5913,
+      "step": 8742
+    },
+    {
+      "epoch": 2.297500574769271,
+      "grad_norm": 0.5265782475471497,
+      "learning_rate": 2.3418608726125813e-05,
+      "loss": 1.6285,
+      "step": 8744
+    },
+    {
+      "epoch": 2.298026078102933,
+      "grad_norm": 0.6365170478820801,
+      "learning_rate": 2.3401086385141053e-05,
+      "loss": 1.6699,
+      "step": 8746
+    },
+    {
+      "epoch": 2.2985515814365947,
+      "grad_norm": 0.5703795552253723,
+      "learning_rate": 2.33835640441563e-05,
+      "loss": 1.6226,
+      "step": 8748
+    },
+    {
+      "epoch": 2.2990770847702566,
+      "grad_norm": 0.6636425852775574,
+      "learning_rate": 2.3366041703171545e-05,
+      "loss": 1.6502,
+      "step": 8750
+    },
+    {
+      "epoch": 2.299602588103918,
+      "grad_norm": 0.5792987942695618,
+      "learning_rate": 2.334851936218679e-05,
+      "loss": 1.6144,
+      "step": 8752
+    },
+    {
+      "epoch": 2.30012809143758,
+      "grad_norm": 0.5882440209388733,
+      "learning_rate": 2.3330997021202033e-05,
+      "loss": 1.6421,
+      "step": 8754
+    },
+    {
+      "epoch": 2.3006535947712417,
+      "grad_norm": 0.8438565135002136,
+      "learning_rate": 2.3313474680217278e-05,
+      "loss": 1.6205,
+      "step": 8756
+    },
+    {
+      "epoch": 2.3011790981049036,
+      "grad_norm": 0.5310785174369812,
+      "learning_rate": 2.3295952339232522e-05,
+      "loss": 1.6295,
+      "step": 8758
+    },
+    {
+      "epoch": 2.3017046014385656,
+      "grad_norm": 0.5153002738952637,
+      "learning_rate": 2.3278429998247766e-05,
+      "loss": 1.6023,
+      "step": 8760
+    },
+    {
+      "epoch": 2.302230104772227,
+      "grad_norm": 0.5312076807022095,
+      "learning_rate": 2.3260907657263014e-05,
+      "loss": 1.617,
+      "step": 8762
+    },
+    {
+      "epoch": 2.302755608105889,
+      "grad_norm": 0.554939329624176,
+      "learning_rate": 2.3243385316278258e-05,
+      "loss": 1.6376,
+      "step": 8764
+    },
+    {
+      "epoch": 2.3032811114395506,
+      "grad_norm": 0.5337182283401489,
+      "learning_rate": 2.32258629752935e-05,
+      "loss": 1.6258,
+      "step": 8766
+    },
+    {
+      "epoch": 2.3038066147732126,
+      "grad_norm": 0.5950394868850708,
+      "learning_rate": 2.3208340634308746e-05,
+      "loss": 1.5906,
+      "step": 8768
+    },
+    {
+      "epoch": 2.304332118106874,
+      "grad_norm": 0.5580046772956848,
+      "learning_rate": 2.319081829332399e-05,
+      "loss": 1.602,
+      "step": 8770
+    },
+    {
+      "epoch": 2.304857621440536,
+      "grad_norm": 0.5954751968383789,
+      "learning_rate": 2.3173295952339234e-05,
+      "loss": 1.6433,
+      "step": 8772
+    },
+    {
+      "epoch": 2.3053831247741976,
+      "grad_norm": 0.5827406644821167,
+      "learning_rate": 2.315577361135448e-05,
+      "loss": 1.6155,
+      "step": 8774
+    },
+    {
+      "epoch": 2.3059086281078596,
+      "grad_norm": 0.5616747140884399,
+      "learning_rate": 2.3138251270369723e-05,
+      "loss": 1.6427,
+      "step": 8776
+    },
+    {
+      "epoch": 2.306434131441521,
+      "grad_norm": 0.5560279488563538,
+      "learning_rate": 2.3120728929384967e-05,
+      "loss": 1.5679,
+      "step": 8778
+    },
+    {
+      "epoch": 2.306959634775183,
+      "grad_norm": 0.5234860777854919,
+      "learning_rate": 2.310320658840021e-05,
+      "loss": 1.6176,
+      "step": 8780
+    },
+    {
+      "epoch": 2.307485138108845,
+      "grad_norm": 0.5881795287132263,
+      "learning_rate": 2.3085684247415455e-05,
+      "loss": 1.635,
+      "step": 8782
+    },
+    {
+      "epoch": 2.3080106414425066,
+      "grad_norm": 0.6444844603538513,
+      "learning_rate": 2.30681619064307e-05,
+      "loss": 1.6151,
+      "step": 8784
+    },
+    {
+      "epoch": 2.3085361447761685,
+      "grad_norm": 0.5576640367507935,
+      "learning_rate": 2.3050639565445944e-05,
+      "loss": 1.6469,
+      "step": 8786
+    },
+    {
+      "epoch": 2.30906164810983,
+      "grad_norm": 0.6228268146514893,
+      "learning_rate": 2.303311722446119e-05,
+      "loss": 1.6463,
+      "step": 8788
+    },
+    {
+      "epoch": 2.309587151443492,
+      "grad_norm": 0.589922308921814,
+      "learning_rate": 2.3015594883476432e-05,
+      "loss": 1.6203,
+      "step": 8790
+    },
+    {
+      "epoch": 2.3101126547771536,
+      "grad_norm": 0.6492023468017578,
+      "learning_rate": 2.2998072542491676e-05,
+      "loss": 1.6371,
+      "step": 8792
+    },
+    {
+      "epoch": 2.3106381581108155,
+      "grad_norm": 0.5567497611045837,
+      "learning_rate": 2.2980550201506924e-05,
+      "loss": 1.6225,
+      "step": 8794
+    },
+    {
+      "epoch": 2.3111636614444775,
+      "grad_norm": 0.568630576133728,
+      "learning_rate": 2.2963027860522165e-05,
+      "loss": 1.6063,
+      "step": 8796
+    },
+    {
+      "epoch": 2.311689164778139,
+      "grad_norm": 0.5894288420677185,
+      "learning_rate": 2.2945505519537412e-05,
+      "loss": 1.6124,
+      "step": 8798
+    },
+    {
+      "epoch": 2.312214668111801,
+      "grad_norm": 0.5140713453292847,
+      "learning_rate": 2.2927983178552656e-05,
+      "loss": 1.6183,
+      "step": 8800
+    },
+    {
+      "epoch": 2.312214668111801,
+      "eval_loss": 1.6566152572631836,
+      "eval_runtime": 486.4646,
+      "eval_samples_per_second": 250.355,
+      "eval_steps_per_second": 31.295,
+      "step": 8800
+    },
+    {
+      "epoch": 2.3127401714454625,
+      "grad_norm": 0.5767953395843506,
+      "learning_rate": 2.29104608375679e-05,
+      "loss": 1.588,
+      "step": 8802
+    },
+    {
+      "epoch": 2.3132656747791245,
+      "grad_norm": 0.5397815704345703,
+      "learning_rate": 2.2892938496583145e-05,
+      "loss": 1.6311,
+      "step": 8804
+    },
+    {
+      "epoch": 2.313791178112786,
+      "grad_norm": 0.59912109375,
+      "learning_rate": 2.287541615559839e-05,
+      "loss": 1.6282,
+      "step": 8806
+    },
+    {
+      "epoch": 2.314316681446448,
+      "grad_norm": 0.5919666886329651,
+      "learning_rate": 2.2857893814613633e-05,
+      "loss": 1.5869,
+      "step": 8808
+    },
+    {
+      "epoch": 2.31484218478011,
+      "grad_norm": 0.6073286533355713,
+      "learning_rate": 2.2840371473628877e-05,
+      "loss": 1.6117,
+      "step": 8810
+    },
+    {
+      "epoch": 2.3153676881137715,
+      "grad_norm": 0.5764369368553162,
+      "learning_rate": 2.282284913264412e-05,
+      "loss": 1.6216,
+      "step": 8812
+    },
+    {
+      "epoch": 2.315893191447433,
+      "grad_norm": 0.5165627598762512,
+      "learning_rate": 2.280532679165937e-05,
+      "loss": 1.5775,
+      "step": 8814
+    },
+    {
+      "epoch": 2.316418694781095,
+      "grad_norm": 0.6328101754188538,
+      "learning_rate": 2.278780445067461e-05,
+      "loss": 1.6164,
+      "step": 8816
+    },
+    {
+      "epoch": 2.316944198114757,
+      "grad_norm": 0.5554302930831909,
+      "learning_rate": 2.2770282109689857e-05,
+      "loss": 1.5912,
+      "step": 8818
+    },
+    {
+      "epoch": 2.3174697014484185,
+      "grad_norm": 0.5721133947372437,
+      "learning_rate": 2.27527597687051e-05,
+      "loss": 1.5831,
+      "step": 8820
+    },
+    {
+      "epoch": 2.3179952047820804,
+      "grad_norm": 0.5364890694618225,
+      "learning_rate": 2.2735237427720342e-05,
+      "loss": 1.6334,
+      "step": 8822
+    },
+    {
+      "epoch": 2.318520708115742,
+      "grad_norm": 0.5981882214546204,
+      "learning_rate": 2.271771508673559e-05,
+      "loss": 1.5971,
+      "step": 8824
+    },
+    {
+      "epoch": 2.319046211449404,
+      "grad_norm": 0.5527270436286926,
+      "learning_rate": 2.2700192745750834e-05,
+      "loss": 1.5844,
+      "step": 8826
+    },
+    {
+      "epoch": 2.3195717147830655,
+      "grad_norm": 0.5365408062934875,
+      "learning_rate": 2.2682670404766078e-05,
+      "loss": 1.6016,
+      "step": 8828
+    },
+    {
+      "epoch": 2.3200972181167274,
+      "grad_norm": 0.5421087741851807,
+      "learning_rate": 2.2665148063781322e-05,
+      "loss": 1.6095,
+      "step": 8830
+    },
+    {
+      "epoch": 2.3206227214503894,
+      "grad_norm": 0.6584184765815735,
+      "learning_rate": 2.2647625722796566e-05,
+      "loss": 1.6167,
+      "step": 8832
+    },
+    {
+      "epoch": 2.321148224784051,
+      "grad_norm": 0.6406242251396179,
+      "learning_rate": 2.2630103381811814e-05,
+      "loss": 1.6178,
+      "step": 8834
+    },
+    {
+      "epoch": 2.321673728117713,
+      "grad_norm": 0.5838053226470947,
+      "learning_rate": 2.2612581040827055e-05,
+      "loss": 1.596,
+      "step": 8836
+    },
+    {
+      "epoch": 2.3221992314513744,
+      "grad_norm": 0.5230047702789307,
+      "learning_rate": 2.25950586998423e-05,
+      "loss": 1.6143,
+      "step": 8838
+    },
+    {
+      "epoch": 2.3227247347850364,
+      "grad_norm": 0.698841392993927,
+      "learning_rate": 2.2577536358857547e-05,
+      "loss": 1.5636,
+      "step": 8840
+    },
+    {
+      "epoch": 2.323250238118698,
+      "grad_norm": 0.6338639855384827,
+      "learning_rate": 2.2560014017872787e-05,
+      "loss": 1.6137,
+      "step": 8842
+    },
+    {
+      "epoch": 2.32377574145236,
+      "grad_norm": 0.5631201863288879,
+      "learning_rate": 2.2542491676888035e-05,
+      "loss": 1.6118,
+      "step": 8844
+    },
+    {
+      "epoch": 2.324301244786022,
+      "grad_norm": 0.5797097086906433,
+      "learning_rate": 2.252496933590328e-05,
+      "loss": 1.6117,
+      "step": 8846
+    },
+    {
+      "epoch": 2.3248267481196834,
+      "grad_norm": 0.55458003282547,
+      "learning_rate": 2.250744699491852e-05,
+      "loss": 1.6074,
+      "step": 8848
+    },
+    {
+      "epoch": 2.3253522514533453,
+      "grad_norm": 0.6182234883308411,
+      "learning_rate": 2.2489924653933767e-05,
+      "loss": 1.5855,
+      "step": 8850
+    },
+    {
+      "epoch": 2.325877754787007,
+      "grad_norm": 0.7982610464096069,
+      "learning_rate": 2.247240231294901e-05,
+      "loss": 1.6479,
+      "step": 8852
+    },
+    {
+      "epoch": 2.326403258120669,
+      "grad_norm": 0.5972685813903809,
+      "learning_rate": 2.2454879971964256e-05,
+      "loss": 1.6508,
+      "step": 8854
+    },
+    {
+      "epoch": 2.3269287614543304,
+      "grad_norm": 0.5340309739112854,
+      "learning_rate": 2.24373576309795e-05,
+      "loss": 1.587,
+      "step": 8856
+    },
+    {
+      "epoch": 2.3274542647879923,
+      "grad_norm": 0.598004162311554,
+      "learning_rate": 2.2419835289994744e-05,
+      "loss": 1.5807,
+      "step": 8858
+    },
+    {
+      "epoch": 2.327979768121654,
+      "grad_norm": 0.5330363512039185,
+      "learning_rate": 2.2402312949009988e-05,
+      "loss": 1.5864,
+      "step": 8860
+    },
+    {
+      "epoch": 2.328505271455316,
+      "grad_norm": 0.5319264531135559,
+      "learning_rate": 2.2384790608025232e-05,
+      "loss": 1.5815,
+      "step": 8862
+    },
+    {
+      "epoch": 2.3290307747889774,
+      "grad_norm": 0.7008064389228821,
+      "learning_rate": 2.2367268267040477e-05,
+      "loss": 1.5968,
+      "step": 8864
+    },
+    {
+      "epoch": 2.3295562781226393,
+      "grad_norm": 0.5392124056816101,
+      "learning_rate": 2.2349745926055724e-05,
+      "loss": 1.5889,
+      "step": 8866
+    },
+    {
+      "epoch": 2.3300817814563013,
+      "grad_norm": 0.5684927701950073,
+      "learning_rate": 2.2332223585070965e-05,
+      "loss": 1.5789,
+      "step": 8868
+    },
+    {
+      "epoch": 2.330607284789963,
+      "grad_norm": 0.5385169386863708,
+      "learning_rate": 2.2314701244086213e-05,
+      "loss": 1.6039,
+      "step": 8870
+    },
+    {
+      "epoch": 2.331132788123625,
+      "grad_norm": 0.5375089645385742,
+      "learning_rate": 2.2297178903101457e-05,
+      "loss": 1.6012,
+      "step": 8872
+    },
+    {
+      "epoch": 2.3316582914572863,
+      "grad_norm": 0.5316143035888672,
+      "learning_rate": 2.2279656562116698e-05,
+      "loss": 1.6277,
+      "step": 8874
+    },
+    {
+      "epoch": 2.3321837947909483,
+      "grad_norm": 0.5862188339233398,
+      "learning_rate": 2.2262134221131945e-05,
+      "loss": 1.5808,
+      "step": 8876
+    },
+    {
+      "epoch": 2.33270929812461,
+      "grad_norm": 0.5072478652000427,
+      "learning_rate": 2.224461188014719e-05,
+      "loss": 1.6257,
+      "step": 8878
+    },
+    {
+      "epoch": 2.333234801458272,
+      "grad_norm": 0.7240272164344788,
+      "learning_rate": 2.2227089539162433e-05,
+      "loss": 1.6551,
+      "step": 8880
+    },
+    {
+      "epoch": 2.3337603047919337,
+      "grad_norm": 0.5968764424324036,
+      "learning_rate": 2.2209567198177678e-05,
+      "loss": 1.6277,
+      "step": 8882
+    },
+    {
+      "epoch": 2.3342858081255953,
+      "grad_norm": 0.576076328754425,
+      "learning_rate": 2.2192044857192922e-05,
+      "loss": 1.6187,
+      "step": 8884
+    },
+    {
+      "epoch": 2.3348113114592572,
+      "grad_norm": 0.7722670435905457,
+      "learning_rate": 2.2174522516208166e-05,
+      "loss": 1.6401,
+      "step": 8886
+    },
+    {
+      "epoch": 2.3353368147929188,
+      "grad_norm": 0.5726610422134399,
+      "learning_rate": 2.215700017522341e-05,
+      "loss": 1.6182,
+      "step": 8888
+    },
+    {
+      "epoch": 2.3358623181265807,
+      "grad_norm": 0.6139241456985474,
+      "learning_rate": 2.2139477834238658e-05,
+      "loss": 1.6314,
+      "step": 8890
+    },
+    {
+      "epoch": 2.3363878214602423,
+      "grad_norm": 0.6056875586509705,
+      "learning_rate": 2.21219554932539e-05,
+      "loss": 1.603,
+      "step": 8892
+    },
+    {
+      "epoch": 2.3369133247939042,
+      "grad_norm": 0.6434742212295532,
+      "learning_rate": 2.2104433152269143e-05,
+      "loss": 1.6344,
+      "step": 8894
+    },
+    {
+      "epoch": 2.3374388281275658,
+      "grad_norm": 0.5169597268104553,
+      "learning_rate": 2.208691081128439e-05,
+      "loss": 1.5879,
+      "step": 8896
+    },
+    {
+      "epoch": 2.3379643314612277,
+      "grad_norm": 0.5403237342834473,
+      "learning_rate": 2.206938847029963e-05,
+      "loss": 1.5932,
+      "step": 8898
+    },
+    {
+      "epoch": 2.3384898347948893,
+      "grad_norm": 0.6257784962654114,
+      "learning_rate": 2.205186612931488e-05,
+      "loss": 1.6506,
+      "step": 8900
+    },
+    {
+      "epoch": 2.3390153381285512,
+      "grad_norm": 0.5608735084533691,
+      "learning_rate": 2.2034343788330123e-05,
+      "loss": 1.587,
+      "step": 8902
+    },
+    {
+      "epoch": 2.339540841462213,
+      "grad_norm": 0.6551423668861389,
+      "learning_rate": 2.2016821447345367e-05,
+      "loss": 1.5972,
+      "step": 8904
+    },
+    {
+      "epoch": 2.3400663447958747,
+      "grad_norm": 0.5031147599220276,
+      "learning_rate": 2.199929910636061e-05,
+      "loss": 1.5885,
+      "step": 8906
+    },
+    {
+      "epoch": 2.3405918481295367,
+      "grad_norm": 0.6937734484672546,
+      "learning_rate": 2.1981776765375855e-05,
+      "loss": 1.6201,
+      "step": 8908
+    },
+    {
+      "epoch": 2.341117351463198,
+      "grad_norm": 0.6449792385101318,
+      "learning_rate": 2.19642544243911e-05,
+      "loss": 1.6335,
+      "step": 8910
+    },
+    {
+      "epoch": 2.34164285479686,
+      "grad_norm": 0.8069351315498352,
+      "learning_rate": 2.1946732083406344e-05,
+      "loss": 1.6147,
+      "step": 8912
+    },
+    {
+      "epoch": 2.3421683581305217,
+      "grad_norm": 0.5487890839576721,
+      "learning_rate": 2.1929209742421588e-05,
+      "loss": 1.6011,
+      "step": 8914
+    },
+    {
+      "epoch": 2.3426938614641837,
+      "grad_norm": 0.5937405228614807,
+      "learning_rate": 2.1911687401436835e-05,
+      "loss": 1.6196,
+      "step": 8916
+    },
+    {
+      "epoch": 2.3432193647978456,
+      "grad_norm": 0.5741852521896362,
+      "learning_rate": 2.1894165060452076e-05,
+      "loss": 1.5813,
+      "step": 8918
+    },
+    {
+      "epoch": 2.343744868131507,
+      "grad_norm": 0.5851845741271973,
+      "learning_rate": 2.187664271946732e-05,
+      "loss": 1.5938,
+      "step": 8920
+    },
+    {
+      "epoch": 2.344270371465169,
+      "grad_norm": 0.5910730361938477,
+      "learning_rate": 2.1859120378482568e-05,
+      "loss": 1.6274,
+      "step": 8922
+    },
+    {
+      "epoch": 2.3447958747988307,
+      "grad_norm": 0.5576980710029602,
+      "learning_rate": 2.184159803749781e-05,
+      "loss": 1.6288,
+      "step": 8924
+    },
+    {
+      "epoch": 2.3453213781324926,
+      "grad_norm": 0.5467607975006104,
+      "learning_rate": 2.1824075696513056e-05,
+      "loss": 1.5965,
+      "step": 8926
+    },
+    {
+      "epoch": 2.345846881466154,
+      "grad_norm": 0.5948976874351501,
+      "learning_rate": 2.18065533555283e-05,
+      "loss": 1.6238,
+      "step": 8928
+    },
+    {
+      "epoch": 2.346372384799816,
+      "grad_norm": 0.6754350066184998,
+      "learning_rate": 2.178903101454354e-05,
+      "loss": 1.6123,
+      "step": 8930
+    },
+    {
+      "epoch": 2.3468978881334777,
+      "grad_norm": 0.5642646551132202,
+      "learning_rate": 2.177150867355879e-05,
+      "loss": 1.6032,
+      "step": 8932
+    },
+    {
+      "epoch": 2.3474233914671396,
+      "grad_norm": 0.5213932991027832,
+      "learning_rate": 2.1753986332574033e-05,
+      "loss": 1.6275,
+      "step": 8934
+    },
+    {
+      "epoch": 2.347948894800801,
+      "grad_norm": 0.5514481663703918,
+      "learning_rate": 2.1736463991589277e-05,
+      "loss": 1.6281,
+      "step": 8936
+    },
+    {
+      "epoch": 2.348474398134463,
+      "grad_norm": 0.6020217537879944,
+      "learning_rate": 2.171894165060452e-05,
+      "loss": 1.5991,
+      "step": 8938
+    },
+    {
+      "epoch": 2.348999901468125,
+      "grad_norm": 0.6889709830284119,
+      "learning_rate": 2.1701419309619765e-05,
+      "loss": 1.5949,
+      "step": 8940
+    },
+    {
+      "epoch": 2.3495254048017866,
+      "grad_norm": 0.5877057313919067,
+      "learning_rate": 2.1683896968635013e-05,
+      "loss": 1.5978,
+      "step": 8942
+    },
+    {
+      "epoch": 2.3500509081354486,
+      "grad_norm": 0.7173901200294495,
+      "learning_rate": 2.1666374627650254e-05,
+      "loss": 1.594,
+      "step": 8944
+    },
+    {
+      "epoch": 2.35057641146911,
+      "grad_norm": 0.5696520805358887,
+      "learning_rate": 2.16488522866655e-05,
+      "loss": 1.6291,
+      "step": 8946
+    },
+    {
+      "epoch": 2.351101914802772,
+      "grad_norm": 0.5779030919075012,
+      "learning_rate": 2.1631329945680746e-05,
+      "loss": 1.6113,
+      "step": 8948
+    },
+    {
+      "epoch": 2.3516274181364336,
+      "grad_norm": 0.5482315421104431,
+      "learning_rate": 2.1613807604695986e-05,
+      "loss": 1.5829,
+      "step": 8950
+    },
+    {
+      "epoch": 2.3521529214700956,
+      "grad_norm": 0.6139046549797058,
+      "learning_rate": 2.1596285263711234e-05,
+      "loss": 1.6207,
+      "step": 8952
+    },
+    {
+      "epoch": 2.3526784248037576,
+      "grad_norm": 0.569508969783783,
+      "learning_rate": 2.1578762922726478e-05,
+      "loss": 1.6108,
+      "step": 8954
+    },
+    {
+      "epoch": 2.353203928137419,
+      "grad_norm": 0.5926774740219116,
+      "learning_rate": 2.1561240581741722e-05,
+      "loss": 1.6094,
+      "step": 8956
+    },
+    {
+      "epoch": 2.353729431471081,
+      "grad_norm": 0.527396559715271,
+      "learning_rate": 2.1543718240756966e-05,
+      "loss": 1.6116,
+      "step": 8958
+    },
+    {
+      "epoch": 2.3542549348047426,
+      "grad_norm": 0.589083731174469,
+      "learning_rate": 2.152619589977221e-05,
+      "loss": 1.6393,
+      "step": 8960
+    },
+    {
+      "epoch": 2.3547804381384045,
+      "grad_norm": 0.5615043044090271,
+      "learning_rate": 2.1508673558787455e-05,
+      "loss": 1.5967,
+      "step": 8962
+    },
+    {
+      "epoch": 2.355305941472066,
+      "grad_norm": 0.5468254685401917,
+      "learning_rate": 2.14911512178027e-05,
+      "loss": 1.6055,
+      "step": 8964
+    },
+    {
+      "epoch": 2.355831444805728,
+      "grad_norm": 0.5892741084098816,
+      "learning_rate": 2.1473628876817943e-05,
+      "loss": 1.6191,
+      "step": 8966
+    },
+    {
+      "epoch": 2.35635694813939,
+      "grad_norm": 0.5186066627502441,
+      "learning_rate": 2.1456106535833187e-05,
+      "loss": 1.6039,
+      "step": 8968
+    },
+    {
+      "epoch": 2.3568824514730515,
+      "grad_norm": 0.5194823741912842,
+      "learning_rate": 2.143858419484843e-05,
+      "loss": 1.6223,
+      "step": 8970
+    },
+    {
+      "epoch": 2.357407954806713,
+      "grad_norm": 0.590596616268158,
+      "learning_rate": 2.142106185386368e-05,
+      "loss": 1.6085,
+      "step": 8972
+    },
+    {
+      "epoch": 2.357933458140375,
+      "grad_norm": 0.5250555872917175,
+      "learning_rate": 2.1403539512878923e-05,
+      "loss": 1.5753,
+      "step": 8974
+    },
+    {
+      "epoch": 2.358458961474037,
+      "grad_norm": 0.5621309280395508,
+      "learning_rate": 2.1386017171894164e-05,
+      "loss": 1.5993,
+      "step": 8976
+    },
+    {
+      "epoch": 2.3589844648076985,
+      "grad_norm": 0.5648888945579529,
+      "learning_rate": 2.136849483090941e-05,
+      "loss": 1.6124,
+      "step": 8978
+    },
+    {
+      "epoch": 2.3595099681413605,
+      "grad_norm": 0.6254675388336182,
+      "learning_rate": 2.1350972489924656e-05,
+      "loss": 1.6137,
+      "step": 8980
+    },
+    {
+      "epoch": 2.360035471475022,
+      "grad_norm": 0.5874773859977722,
+      "learning_rate": 2.13334501489399e-05,
+      "loss": 1.6192,
+      "step": 8982
+    },
+    {
+      "epoch": 2.360560974808684,
+      "grad_norm": 0.5968673229217529,
+      "learning_rate": 2.1315927807955144e-05,
+      "loss": 1.5798,
+      "step": 8984
+    },
+    {
+      "epoch": 2.3610864781423455,
+      "grad_norm": 0.5419182181358337,
+      "learning_rate": 2.1298405466970388e-05,
+      "loss": 1.5893,
+      "step": 8986
+    },
+    {
+      "epoch": 2.3616119814760075,
+      "grad_norm": 0.5870303511619568,
+      "learning_rate": 2.1280883125985632e-05,
+      "loss": 1.6178,
+      "step": 8988
+    },
+    {
+      "epoch": 2.3621374848096695,
+      "grad_norm": 0.5218934416770935,
+      "learning_rate": 2.1263360785000877e-05,
+      "loss": 1.5627,
+      "step": 8990
+    },
+    {
+      "epoch": 2.362662988143331,
+      "grad_norm": 0.7437348365783691,
+      "learning_rate": 2.124583844401612e-05,
+      "loss": 1.633,
+      "step": 8992
+    },
+    {
+      "epoch": 2.363188491476993,
+      "grad_norm": 0.6861118078231812,
+      "learning_rate": 2.1228316103031365e-05,
+      "loss": 1.6051,
+      "step": 8994
+    },
+    {
+      "epoch": 2.3637139948106545,
+      "grad_norm": 0.7474240660667419,
+      "learning_rate": 2.121079376204661e-05,
+      "loss": 1.6053,
+      "step": 8996
+    },
+    {
+      "epoch": 2.3642394981443164,
+      "grad_norm": 0.6993446946144104,
+      "learning_rate": 2.1193271421061857e-05,
+      "loss": 1.6337,
+      "step": 8998
+    },
+    {
+      "epoch": 2.364765001477978,
+      "grad_norm": 0.5786541104316711,
+      "learning_rate": 2.1175749080077097e-05,
+      "loss": 1.6258,
+      "step": 9000
+    },
+    {
+      "epoch": 2.36529050481164,
+      "grad_norm": 0.6487932205200195,
+      "learning_rate": 2.115822673909234e-05,
+      "loss": 1.5785,
+      "step": 9002
+    },
+    {
+      "epoch": 2.365816008145302,
+      "grad_norm": 0.6424734592437744,
+      "learning_rate": 2.114070439810759e-05,
+      "loss": 1.6047,
+      "step": 9004
+    },
+    {
+      "epoch": 2.3663415114789634,
+      "grad_norm": 0.5754727721214294,
+      "learning_rate": 2.1123182057122833e-05,
+      "loss": 1.6199,
+      "step": 9006
+    },
+    {
+      "epoch": 2.3668670148126254,
+      "grad_norm": 0.6415467858314514,
+      "learning_rate": 2.1105659716138078e-05,
+      "loss": 1.6112,
+      "step": 9008
+    },
+    {
+      "epoch": 2.367392518146287,
+      "grad_norm": 0.5382227897644043,
+      "learning_rate": 2.1088137375153322e-05,
+      "loss": 1.6199,
+      "step": 9010
+    },
+    {
+      "epoch": 2.367918021479949,
+      "grad_norm": 0.5709097385406494,
+      "learning_rate": 2.1070615034168566e-05,
+      "loss": 1.6274,
+      "step": 9012
+    },
+    {
+      "epoch": 2.3684435248136104,
+      "grad_norm": 0.584823489189148,
+      "learning_rate": 2.105309269318381e-05,
+      "loss": 1.6364,
+      "step": 9014
+    },
+    {
+      "epoch": 2.3689690281472724,
+      "grad_norm": 0.5769666433334351,
+      "learning_rate": 2.1035570352199054e-05,
+      "loss": 1.613,
+      "step": 9016
+    },
+    {
+      "epoch": 2.369494531480934,
+      "grad_norm": 0.6009288430213928,
+      "learning_rate": 2.1018048011214302e-05,
+      "loss": 1.6115,
+      "step": 9018
+    },
+    {
+      "epoch": 2.370020034814596,
+      "grad_norm": 0.5430034399032593,
+      "learning_rate": 2.1000525670229543e-05,
+      "loss": 1.5935,
+      "step": 9020
+    },
+    {
+      "epoch": 2.3705455381482574,
+      "grad_norm": 0.5354077816009521,
+      "learning_rate": 2.0983003329244787e-05,
+      "loss": 1.6303,
+      "step": 9022
+    },
+    {
+      "epoch": 2.3710710414819194,
+      "grad_norm": 0.5706160664558411,
+      "learning_rate": 2.0965480988260034e-05,
+      "loss": 1.5922,
+      "step": 9024
+    },
+    {
+      "epoch": 2.3715965448155814,
+      "grad_norm": 0.5503175854682922,
+      "learning_rate": 2.0947958647275275e-05,
+      "loss": 1.5995,
+      "step": 9026
+    },
+    {
+      "epoch": 2.372122048149243,
+      "grad_norm": 0.5860917568206787,
+      "learning_rate": 2.0930436306290523e-05,
+      "loss": 1.617,
+      "step": 9028
+    },
+    {
+      "epoch": 2.372647551482905,
+      "grad_norm": 0.5663593411445618,
+      "learning_rate": 2.0912913965305767e-05,
+      "loss": 1.6142,
+      "step": 9030
+    },
+    {
+      "epoch": 2.3731730548165664,
+      "grad_norm": 0.5283660888671875,
+      "learning_rate": 2.0895391624321008e-05,
+      "loss": 1.6111,
+      "step": 9032
+    },
+    {
+      "epoch": 2.3736985581502283,
+      "grad_norm": 0.6298090815544128,
+      "learning_rate": 2.0877869283336255e-05,
+      "loss": 1.6362,
+      "step": 9034
+    },
+    {
+      "epoch": 2.37422406148389,
+      "grad_norm": 0.6019343137741089,
+      "learning_rate": 2.08603469423515e-05,
+      "loss": 1.6493,
+      "step": 9036
+    },
+    {
+      "epoch": 2.374749564817552,
+      "grad_norm": 0.6397587656974792,
+      "learning_rate": 2.0842824601366744e-05,
+      "loss": 1.6063,
+      "step": 9038
+    },
+    {
+      "epoch": 2.375275068151214,
+      "grad_norm": 0.634833037853241,
+      "learning_rate": 2.0825302260381988e-05,
+      "loss": 1.6041,
+      "step": 9040
+    },
+    {
+      "epoch": 2.3758005714848753,
+      "grad_norm": 0.535504162311554,
+      "learning_rate": 2.0807779919397232e-05,
+      "loss": 1.6136,
+      "step": 9042
+    },
+    {
+      "epoch": 2.3763260748185373,
+      "grad_norm": 0.5898988246917725,
+      "learning_rate": 2.079025757841248e-05,
+      "loss": 1.5947,
+      "step": 9044
+    },
+    {
+      "epoch": 2.376851578152199,
+      "grad_norm": 0.6104636192321777,
+      "learning_rate": 2.077273523742772e-05,
+      "loss": 1.6149,
+      "step": 9046
+    },
+    {
+      "epoch": 2.377377081485861,
+      "grad_norm": 0.6539409756660461,
+      "learning_rate": 2.0755212896442964e-05,
+      "loss": 1.5934,
+      "step": 9048
+    },
+    {
+      "epoch": 2.3779025848195223,
+      "grad_norm": 0.6595098972320557,
+      "learning_rate": 2.0737690555458212e-05,
+      "loss": 1.588,
+      "step": 9050
+    },
+    {
+      "epoch": 2.3784280881531843,
+      "grad_norm": 0.5743740200996399,
+      "learning_rate": 2.0720168214473453e-05,
+      "loss": 1.6563,
+      "step": 9052
+    },
+    {
+      "epoch": 2.378953591486846,
+      "grad_norm": 0.7335343956947327,
+      "learning_rate": 2.07026458734887e-05,
+      "loss": 1.6172,
+      "step": 9054
+    },
+    {
+      "epoch": 2.379479094820508,
+      "grad_norm": 0.5948194861412048,
+      "learning_rate": 2.0685123532503945e-05,
+      "loss": 1.6038,
+      "step": 9056
+    },
+    {
+      "epoch": 2.3800045981541693,
+      "grad_norm": 0.6061370968818665,
+      "learning_rate": 2.0667601191519185e-05,
+      "loss": 1.6064,
+      "step": 9058
+    },
+    {
+      "epoch": 2.3805301014878313,
+      "grad_norm": 0.5088295936584473,
+      "learning_rate": 2.0650078850534433e-05,
+      "loss": 1.5974,
+      "step": 9060
+    },
+    {
+      "epoch": 2.3810556048214933,
+      "grad_norm": 0.5611675977706909,
+      "learning_rate": 2.0632556509549677e-05,
+      "loss": 1.6219,
+      "step": 9062
+    },
+    {
+      "epoch": 2.381581108155155,
+      "grad_norm": 0.7418251037597656,
+      "learning_rate": 2.061503416856492e-05,
+      "loss": 1.6091,
+      "step": 9064
+    },
+    {
+      "epoch": 2.3821066114888167,
+      "grad_norm": 0.729642927646637,
+      "learning_rate": 2.0597511827580165e-05,
+      "loss": 1.5959,
+      "step": 9066
+    },
+    {
+      "epoch": 2.3826321148224783,
+      "grad_norm": 0.564006507396698,
+      "learning_rate": 2.057998948659541e-05,
+      "loss": 1.6484,
+      "step": 9068
+    },
+    {
+      "epoch": 2.3831576181561402,
+      "grad_norm": 0.5700123906135559,
+      "learning_rate": 2.0562467145610654e-05,
+      "loss": 1.61,
+      "step": 9070
+    },
+    {
+      "epoch": 2.3836831214898018,
+      "grad_norm": 0.5242093801498413,
+      "learning_rate": 2.0544944804625898e-05,
+      "loss": 1.6011,
+      "step": 9072
+    },
+    {
+      "epoch": 2.3842086248234637,
+      "grad_norm": 0.5738998651504517,
+      "learning_rate": 2.0527422463641146e-05,
+      "loss": 1.6154,
+      "step": 9074
+    },
+    {
+      "epoch": 2.3847341281571257,
+      "grad_norm": 0.5821254253387451,
+      "learning_rate": 2.050990012265639e-05,
+      "loss": 1.6443,
+      "step": 9076
+    },
+    {
+      "epoch": 2.3852596314907872,
+      "grad_norm": 0.6593937873840332,
+      "learning_rate": 2.049237778167163e-05,
+      "loss": 1.6266,
+      "step": 9078
+    },
+    {
+      "epoch": 2.385785134824449,
+      "grad_norm": 0.5662251710891724,
+      "learning_rate": 2.0474855440686878e-05,
+      "loss": 1.6036,
+      "step": 9080
+    },
+    {
+      "epoch": 2.3863106381581107,
+      "grad_norm": 0.6064842939376831,
+      "learning_rate": 2.0457333099702122e-05,
+      "loss": 1.6067,
+      "step": 9082
+    },
+    {
+      "epoch": 2.3868361414917727,
+      "grad_norm": 0.5419683456420898,
+      "learning_rate": 2.0439810758717366e-05,
+      "loss": 1.6429,
+      "step": 9084
+    },
+    {
+      "epoch": 2.3873616448254342,
+      "grad_norm": 0.630354642868042,
+      "learning_rate": 2.042228841773261e-05,
+      "loss": 1.6354,
+      "step": 9086
+    },
+    {
+      "epoch": 2.387887148159096,
+      "grad_norm": 0.5797430276870728,
+      "learning_rate": 2.0404766076747855e-05,
+      "loss": 1.596,
+      "step": 9088
+    },
+    {
+      "epoch": 2.3884126514927577,
+      "grad_norm": 0.7524275183677673,
+      "learning_rate": 2.03872437357631e-05,
+      "loss": 1.6484,
+      "step": 9090
+    },
+    {
+      "epoch": 2.3889381548264197,
+      "grad_norm": 0.5289912819862366,
+      "learning_rate": 2.0369721394778343e-05,
+      "loss": 1.6089,
+      "step": 9092
+    },
+    {
+      "epoch": 2.389463658160081,
+      "grad_norm": 0.610967755317688,
+      "learning_rate": 2.0352199053793587e-05,
+      "loss": 1.5842,
+      "step": 9094
+    },
+    {
+      "epoch": 2.389989161493743,
+      "grad_norm": 0.5420004725456238,
+      "learning_rate": 2.033467671280883e-05,
+      "loss": 1.6497,
+      "step": 9096
+    },
+    {
+      "epoch": 2.390514664827405,
+      "grad_norm": 0.5605329275131226,
+      "learning_rate": 2.0317154371824076e-05,
+      "loss": 1.5927,
+      "step": 9098
+    },
+    {
+      "epoch": 2.3910401681610667,
+      "grad_norm": 0.7482327222824097,
+      "learning_rate": 2.0299632030839323e-05,
+      "loss": 1.6462,
+      "step": 9100
+    },
+    {
+      "epoch": 2.3915656714947287,
+      "grad_norm": 0.536505937576294,
+      "learning_rate": 2.0282109689854564e-05,
+      "loss": 1.5756,
+      "step": 9102
+    },
+    {
+      "epoch": 2.39209117482839,
+      "grad_norm": 0.6347717046737671,
+      "learning_rate": 2.0264587348869808e-05,
+      "loss": 1.5967,
+      "step": 9104
+    },
+    {
+      "epoch": 2.392616678162052,
+      "grad_norm": 0.8039228320121765,
+      "learning_rate": 2.0247065007885056e-05,
+      "loss": 1.589,
+      "step": 9106
+    },
+    {
+      "epoch": 2.3931421814957137,
+      "grad_norm": 0.5720674395561218,
+      "learning_rate": 2.02295426669003e-05,
+      "loss": 1.6207,
+      "step": 9108
+    },
+    {
+      "epoch": 2.3936676848293756,
+      "grad_norm": 0.538987398147583,
+      "learning_rate": 2.0212020325915544e-05,
+      "loss": 1.604,
+      "step": 9110
+    },
+    {
+      "epoch": 2.3941931881630376,
+      "grad_norm": 0.5741251111030579,
+      "learning_rate": 2.0194497984930788e-05,
+      "loss": 1.6345,
+      "step": 9112
+    },
+    {
+      "epoch": 2.394718691496699,
+      "grad_norm": 0.6911687254905701,
+      "learning_rate": 2.0176975643946032e-05,
+      "loss": 1.596,
+      "step": 9114
+    },
+    {
+      "epoch": 2.395244194830361,
+      "grad_norm": 0.5507877469062805,
+      "learning_rate": 2.0159453302961277e-05,
+      "loss": 1.5973,
+      "step": 9116
+    },
+    {
+      "epoch": 2.3957696981640226,
+      "grad_norm": 0.6160647869110107,
+      "learning_rate": 2.014193096197652e-05,
+      "loss": 1.6113,
+      "step": 9118
+    },
+    {
+      "epoch": 2.3962952014976846,
+      "grad_norm": 0.6313309669494629,
+      "learning_rate": 2.0124408620991765e-05,
+      "loss": 1.596,
+      "step": 9120
+    },
+    {
+      "epoch": 2.396820704831346,
+      "grad_norm": 0.6924790740013123,
+      "learning_rate": 2.010688628000701e-05,
+      "loss": 1.6417,
+      "step": 9122
+    },
+    {
+      "epoch": 2.397346208165008,
+      "grad_norm": 0.5447295904159546,
+      "learning_rate": 2.0089363939022253e-05,
+      "loss": 1.6247,
+      "step": 9124
+    },
+    {
+      "epoch": 2.39787171149867,
+      "grad_norm": 0.5692989230155945,
+      "learning_rate": 2.00718415980375e-05,
+      "loss": 1.5952,
+      "step": 9126
+    },
+    {
+      "epoch": 2.3983972148323316,
+      "grad_norm": 0.6437897682189941,
+      "learning_rate": 2.005431925705274e-05,
+      "loss": 1.6339,
+      "step": 9128
+    },
+    {
+      "epoch": 2.398922718165993,
+      "grad_norm": 0.6150215268135071,
+      "learning_rate": 2.0036796916067986e-05,
+      "loss": 1.6362,
+      "step": 9130
+    },
+    {
+      "epoch": 2.399448221499655,
+      "grad_norm": 0.6019598841667175,
+      "learning_rate": 2.0019274575083233e-05,
+      "loss": 1.6287,
+      "step": 9132
+    },
+    {
+      "epoch": 2.399973724833317,
+      "grad_norm": 0.578983724117279,
+      "learning_rate": 2.0001752234098474e-05,
+      "loss": 1.6199,
+      "step": 9134
+    },
+    {
+      "epoch": 2.4004992281669786,
+      "grad_norm": 0.5352420210838318,
+      "learning_rate": 1.9984229893113722e-05,
+      "loss": 1.6066,
+      "step": 9136
+    },
+    {
+      "epoch": 2.4010247315006406,
+      "grad_norm": 0.6357502341270447,
+      "learning_rate": 1.9966707552128966e-05,
+      "loss": 1.6451,
+      "step": 9138
+    },
+    {
+      "epoch": 2.401550234834302,
+      "grad_norm": 0.6802014708518982,
+      "learning_rate": 1.994918521114421e-05,
+      "loss": 1.6011,
+      "step": 9140
+    },
+    {
+      "epoch": 2.402075738167964,
+      "grad_norm": 0.6203175187110901,
+      "learning_rate": 1.9931662870159454e-05,
+      "loss": 1.6168,
+      "step": 9142
+    },
+    {
+      "epoch": 2.4026012415016256,
+      "grad_norm": 0.5582800507545471,
+      "learning_rate": 1.99141405291747e-05,
+      "loss": 1.6206,
+      "step": 9144
+    },
+    {
+      "epoch": 2.4031267448352875,
+      "grad_norm": 0.6481848955154419,
+      "learning_rate": 1.9896618188189946e-05,
+      "loss": 1.6198,
+      "step": 9146
+    },
+    {
+      "epoch": 2.4036522481689495,
+      "grad_norm": 0.5764234662055969,
+      "learning_rate": 1.9879095847205187e-05,
+      "loss": 1.6328,
+      "step": 9148
+    },
+    {
+      "epoch": 2.404177751502611,
+      "grad_norm": 0.6788586378097534,
+      "learning_rate": 1.986157350622043e-05,
+      "loss": 1.6166,
+      "step": 9150
+    },
+    {
+      "epoch": 2.404703254836273,
+      "grad_norm": 0.5868250727653503,
+      "learning_rate": 1.984405116523568e-05,
+      "loss": 1.5952,
+      "step": 9152
+    },
+    {
+      "epoch": 2.4052287581699345,
+      "grad_norm": 0.6610315442085266,
+      "learning_rate": 1.982652882425092e-05,
+      "loss": 1.597,
+      "step": 9154
+    },
+    {
+      "epoch": 2.4057542615035965,
+      "grad_norm": 0.7634231448173523,
+      "learning_rate": 1.9809006483266167e-05,
+      "loss": 1.6235,
+      "step": 9156
+    },
+    {
+      "epoch": 2.406279764837258,
+      "grad_norm": 0.5905351638793945,
+      "learning_rate": 1.979148414228141e-05,
+      "loss": 1.6214,
+      "step": 9158
+    },
+    {
+      "epoch": 2.40680526817092,
+      "grad_norm": 0.5715067982673645,
+      "learning_rate": 1.9773961801296652e-05,
+      "loss": 1.5922,
+      "step": 9160
+    },
+    {
+      "epoch": 2.407330771504582,
+      "grad_norm": 0.593390941619873,
+      "learning_rate": 1.97564394603119e-05,
+      "loss": 1.6207,
+      "step": 9162
+    },
+    {
+      "epoch": 2.4078562748382435,
+      "grad_norm": 0.5678249597549438,
+      "learning_rate": 1.9738917119327144e-05,
+      "loss": 1.6033,
+      "step": 9164
+    },
+    {
+      "epoch": 2.4083817781719055,
+      "grad_norm": 0.6406270861625671,
+      "learning_rate": 1.9721394778342388e-05,
+      "loss": 1.6363,
+      "step": 9166
+    },
+    {
+      "epoch": 2.408907281505567,
+      "grad_norm": 0.5460156202316284,
+      "learning_rate": 1.9703872437357632e-05,
+      "loss": 1.6043,
+      "step": 9168
+    },
+    {
+      "epoch": 2.409432784839229,
+      "grad_norm": 0.5826367735862732,
+      "learning_rate": 1.9686350096372876e-05,
+      "loss": 1.6166,
+      "step": 9170
+    },
+    {
+      "epoch": 2.4099582881728905,
+      "grad_norm": 0.5770626068115234,
+      "learning_rate": 1.966882775538812e-05,
+      "loss": 1.5933,
+      "step": 9172
+    },
+    {
+      "epoch": 2.4104837915065525,
+      "grad_norm": 0.5596659183502197,
+      "learning_rate": 1.9651305414403364e-05,
+      "loss": 1.6123,
+      "step": 9174
+    },
+    {
+      "epoch": 2.411009294840214,
+      "grad_norm": 0.6251547336578369,
+      "learning_rate": 1.963378307341861e-05,
+      "loss": 1.6105,
+      "step": 9176
+    },
+    {
+      "epoch": 2.411534798173876,
+      "grad_norm": 0.566064715385437,
+      "learning_rate": 1.9616260732433856e-05,
+      "loss": 1.6127,
+      "step": 9178
+    },
+    {
+      "epoch": 2.4120603015075375,
+      "grad_norm": 0.5645999908447266,
+      "learning_rate": 1.9598738391449097e-05,
+      "loss": 1.6105,
+      "step": 9180
+    },
+    {
+      "epoch": 2.4125858048411994,
+      "grad_norm": 0.5924414396286011,
+      "learning_rate": 1.9581216050464345e-05,
+      "loss": 1.6258,
+      "step": 9182
+    },
+    {
+      "epoch": 2.4131113081748614,
+      "grad_norm": 0.5936945080757141,
+      "learning_rate": 1.956369370947959e-05,
+      "loss": 1.6088,
+      "step": 9184
+    },
+    {
+      "epoch": 2.413636811508523,
+      "grad_norm": 0.711117148399353,
+      "learning_rate": 1.954617136849483e-05,
+      "loss": 1.5842,
+      "step": 9186
+    },
+    {
+      "epoch": 2.414162314842185,
+      "grad_norm": 0.5747193694114685,
+      "learning_rate": 1.9528649027510077e-05,
+      "loss": 1.5934,
+      "step": 9188
+    },
+    {
+      "epoch": 2.4146878181758464,
+      "grad_norm": 0.6013287901878357,
+      "learning_rate": 1.951112668652532e-05,
+      "loss": 1.5854,
+      "step": 9190
+    },
+    {
+      "epoch": 2.4152133215095084,
+      "grad_norm": 0.680145263671875,
+      "learning_rate": 1.9493604345540565e-05,
+      "loss": 1.6323,
+      "step": 9192
+    },
+    {
+      "epoch": 2.41573882484317,
+      "grad_norm": 0.6533603072166443,
+      "learning_rate": 1.947608200455581e-05,
+      "loss": 1.5944,
+      "step": 9194
+    },
+    {
+      "epoch": 2.416264328176832,
+      "grad_norm": 0.549355149269104,
+      "learning_rate": 1.9458559663571054e-05,
+      "loss": 1.628,
+      "step": 9196
+    },
+    {
+      "epoch": 2.416789831510494,
+      "grad_norm": 0.5181589722633362,
+      "learning_rate": 1.9441037322586298e-05,
+      "loss": 1.6072,
+      "step": 9198
+    },
+    {
+      "epoch": 2.4173153348441554,
+      "grad_norm": 0.5603959560394287,
+      "learning_rate": 1.9423514981601542e-05,
+      "loss": 1.6101,
+      "step": 9200
+    },
+    {
+      "epoch": 2.4173153348441554,
+      "eval_loss": 1.6572836637496948,
+      "eval_runtime": 486.1392,
+      "eval_samples_per_second": 250.523,
+      "eval_steps_per_second": 31.316,
+      "step": 9200
+    },
+    {
+      "epoch": 2.4178408381778174,
+      "grad_norm": 0.5568393468856812,
+      "learning_rate": 1.940599264061679e-05,
+      "loss": 1.5855,
+      "step": 9202
+    },
+    {
+      "epoch": 2.418366341511479,
+      "grad_norm": 0.6060105562210083,
+      "learning_rate": 1.938847029963203e-05,
+      "loss": 1.6208,
+      "step": 9204
+    },
+    {
+      "epoch": 2.418891844845141,
+      "grad_norm": 0.5444419384002686,
+      "learning_rate": 1.9370947958647275e-05,
+      "loss": 1.6137,
+      "step": 9206
+    },
+    {
+      "epoch": 2.4194173481788024,
+      "grad_norm": 0.7899800539016724,
+      "learning_rate": 1.9353425617662522e-05,
+      "loss": 1.6181,
+      "step": 9208
+    },
+    {
+      "epoch": 2.4199428515124644,
+      "grad_norm": 0.5947116017341614,
+      "learning_rate": 1.9335903276677763e-05,
+      "loss": 1.5974,
+      "step": 9210
+    },
+    {
+      "epoch": 2.420468354846126,
+      "grad_norm": 0.6116278171539307,
+      "learning_rate": 1.931838093569301e-05,
+      "loss": 1.6519,
+      "step": 9212
+    },
+    {
+      "epoch": 2.420993858179788,
+      "grad_norm": 0.645462691783905,
+      "learning_rate": 1.9300858594708255e-05,
+      "loss": 1.586,
+      "step": 9214
+    },
+    {
+      "epoch": 2.4215193615134494,
+      "grad_norm": 0.5454298853874207,
+      "learning_rate": 1.92833362537235e-05,
+      "loss": 1.5734,
+      "step": 9216
+    },
+    {
+      "epoch": 2.4220448648471113,
+      "grad_norm": 0.595712423324585,
+      "learning_rate": 1.9265813912738743e-05,
+      "loss": 1.6554,
+      "step": 9218
+    },
+    {
+      "epoch": 2.4225703681807733,
+      "grad_norm": 0.5496743321418762,
+      "learning_rate": 1.9248291571753987e-05,
+      "loss": 1.6148,
+      "step": 9220
+    },
+    {
+      "epoch": 2.423095871514435,
+      "grad_norm": 0.5601218342781067,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 1.59,
+      "step": 9222
+    },
+    {
+      "epoch": 2.423621374848097,
+      "grad_norm": 0.5808682441711426,
+      "learning_rate": 1.9213246889784476e-05,
+      "loss": 1.6106,
+      "step": 9224
+    },
+    {
+      "epoch": 2.4241468781817583,
+      "grad_norm": 0.6002793908119202,
+      "learning_rate": 1.919572454879972e-05,
+      "loss": 1.6091,
+      "step": 9226
+    },
+    {
+      "epoch": 2.4246723815154203,
+      "grad_norm": 0.5775554776191711,
+      "learning_rate": 1.9178202207814967e-05,
+      "loss": 1.5891,
+      "step": 9228
+    },
+    {
+      "epoch": 2.425197884849082,
+      "grad_norm": 0.6747956871986389,
+      "learning_rate": 1.9160679866830208e-05,
+      "loss": 1.5877,
+      "step": 9230
+    },
+    {
+      "epoch": 2.425723388182744,
+      "grad_norm": 0.549020528793335,
+      "learning_rate": 1.9143157525845452e-05,
+      "loss": 1.5931,
+      "step": 9232
+    },
+    {
+      "epoch": 2.4262488915164058,
+      "grad_norm": 0.6531383395195007,
+      "learning_rate": 1.91256351848607e-05,
+      "loss": 1.6393,
+      "step": 9234
+    },
+    {
+      "epoch": 2.4267743948500673,
+      "grad_norm": 0.7158095240592957,
+      "learning_rate": 1.910811284387594e-05,
+      "loss": 1.5868,
+      "step": 9236
+    },
+    {
+      "epoch": 2.4272998981837293,
+      "grad_norm": 0.5585731267929077,
+      "learning_rate": 1.9090590502891188e-05,
+      "loss": 1.6519,
+      "step": 9238
+    },
+    {
+      "epoch": 2.427825401517391,
+      "grad_norm": 0.6200599670410156,
+      "learning_rate": 1.9073068161906432e-05,
+      "loss": 1.6401,
+      "step": 9240
+    },
+    {
+      "epoch": 2.4283509048510528,
+      "grad_norm": 0.5844702124595642,
+      "learning_rate": 1.9055545820921673e-05,
+      "loss": 1.5893,
+      "step": 9242
+    },
+    {
+      "epoch": 2.4288764081847143,
+      "grad_norm": 0.5763403177261353,
+      "learning_rate": 1.903802347993692e-05,
+      "loss": 1.6056,
+      "step": 9244
+    },
+    {
+      "epoch": 2.4294019115183763,
+      "grad_norm": 0.6376482844352722,
+      "learning_rate": 1.9020501138952165e-05,
+      "loss": 1.6083,
+      "step": 9246
+    },
+    {
+      "epoch": 2.429927414852038,
+      "grad_norm": 0.6373746991157532,
+      "learning_rate": 1.900297879796741e-05,
+      "loss": 1.5721,
+      "step": 9248
+    },
+    {
+      "epoch": 2.4304529181856998,
+      "grad_norm": 0.572599470615387,
+      "learning_rate": 1.8985456456982653e-05,
+      "loss": 1.6161,
+      "step": 9250
+    },
+    {
+      "epoch": 2.4309784215193613,
+      "grad_norm": 0.5486949682235718,
+      "learning_rate": 1.8967934115997897e-05,
+      "loss": 1.591,
+      "step": 9252
+    },
+    {
+      "epoch": 2.4315039248530232,
+      "grad_norm": 0.6981235146522522,
+      "learning_rate": 1.8950411775013145e-05,
+      "loss": 1.6399,
+      "step": 9254
+    },
+    {
+      "epoch": 2.432029428186685,
+      "grad_norm": 0.7503864169120789,
+      "learning_rate": 1.8932889434028386e-05,
+      "loss": 1.6222,
+      "step": 9256
+    },
+    {
+      "epoch": 2.4325549315203467,
+      "grad_norm": 0.5889126062393188,
+      "learning_rate": 1.891536709304363e-05,
+      "loss": 1.6176,
+      "step": 9258
+    },
+    {
+      "epoch": 2.4330804348540087,
+      "grad_norm": 0.5790210962295532,
+      "learning_rate": 1.8897844752058878e-05,
+      "loss": 1.6321,
+      "step": 9260
+    },
+    {
+      "epoch": 2.4336059381876702,
+      "grad_norm": 0.5340055823326111,
+      "learning_rate": 1.888032241107412e-05,
+      "loss": 1.6009,
+      "step": 9262
+    },
+    {
+      "epoch": 2.434131441521332,
+      "grad_norm": 0.5527550578117371,
+      "learning_rate": 1.8862800070089366e-05,
+      "loss": 1.6173,
+      "step": 9264
+    },
+    {
+      "epoch": 2.4346569448549937,
+      "grad_norm": 0.6926127672195435,
+      "learning_rate": 1.884527772910461e-05,
+      "loss": 1.5944,
+      "step": 9266
+    },
+    {
+      "epoch": 2.4351824481886557,
+      "grad_norm": 0.5821343064308167,
+      "learning_rate": 1.8827755388119854e-05,
+      "loss": 1.6235,
+      "step": 9268
+    },
+    {
+      "epoch": 2.4357079515223177,
+      "grad_norm": 0.5754803419113159,
+      "learning_rate": 1.88102330471351e-05,
+      "loss": 1.6216,
+      "step": 9270
+    },
+    {
+      "epoch": 2.436233454855979,
+      "grad_norm": 0.5613455176353455,
+      "learning_rate": 1.8792710706150343e-05,
+      "loss": 1.607,
+      "step": 9272
+    },
+    {
+      "epoch": 2.436758958189641,
+      "grad_norm": 0.5742526650428772,
+      "learning_rate": 1.8775188365165587e-05,
+      "loss": 1.5704,
+      "step": 9274
+    },
+    {
+      "epoch": 2.4372844615233027,
+      "grad_norm": 0.5530628561973572,
+      "learning_rate": 1.875766602418083e-05,
+      "loss": 1.5849,
+      "step": 9276
+    },
+    {
+      "epoch": 2.4378099648569647,
+      "grad_norm": 0.5322666168212891,
+      "learning_rate": 1.8740143683196075e-05,
+      "loss": 1.5777,
+      "step": 9278
+    },
+    {
+      "epoch": 2.438335468190626,
+      "grad_norm": 0.5801582932472229,
+      "learning_rate": 1.8722621342211323e-05,
+      "loss": 1.6277,
+      "step": 9280
+    },
+    {
+      "epoch": 2.438860971524288,
+      "grad_norm": 0.6362050175666809,
+      "learning_rate": 1.8705099001226563e-05,
+      "loss": 1.6053,
+      "step": 9282
+    },
+    {
+      "epoch": 2.43938647485795,
+      "grad_norm": 0.6227912306785583,
+      "learning_rate": 1.868757666024181e-05,
+      "loss": 1.6558,
+      "step": 9284
+    },
+    {
+      "epoch": 2.4399119781916117,
+      "grad_norm": 0.5319087505340576,
+      "learning_rate": 1.8670054319257055e-05,
+      "loss": 1.6402,
+      "step": 9286
+    },
+    {
+      "epoch": 2.440437481525273,
+      "grad_norm": 0.5639674067497253,
+      "learning_rate": 1.8652531978272296e-05,
+      "loss": 1.6191,
+      "step": 9288
+    },
+    {
+      "epoch": 2.440962984858935,
+      "grad_norm": 0.5504940748214722,
+      "learning_rate": 1.8635009637287544e-05,
+      "loss": 1.6054,
+      "step": 9290
+    },
+    {
+      "epoch": 2.441488488192597,
+      "grad_norm": 0.535571277141571,
+      "learning_rate": 1.8617487296302788e-05,
+      "loss": 1.5875,
+      "step": 9292
+    },
+    {
+      "epoch": 2.4420139915262586,
+      "grad_norm": 0.5902593731880188,
+      "learning_rate": 1.8599964955318032e-05,
+      "loss": 1.5771,
+      "step": 9294
+    },
+    {
+      "epoch": 2.4425394948599206,
+      "grad_norm": 0.5818725824356079,
+      "learning_rate": 1.8582442614333276e-05,
+      "loss": 1.5895,
+      "step": 9296
+    },
+    {
+      "epoch": 2.443064998193582,
+      "grad_norm": 0.5598388314247131,
+      "learning_rate": 1.856492027334852e-05,
+      "loss": 1.619,
+      "step": 9298
+    },
+    {
+      "epoch": 2.443590501527244,
+      "grad_norm": 0.6018016934394836,
+      "learning_rate": 1.8547397932363764e-05,
+      "loss": 1.5975,
+      "step": 9300
+    },
+    {
+      "epoch": 2.4441160048609056,
+      "grad_norm": 0.5516964197158813,
+      "learning_rate": 1.852987559137901e-05,
+      "loss": 1.6073,
+      "step": 9302
+    },
+    {
+      "epoch": 2.4446415081945676,
+      "grad_norm": 0.5745545625686646,
+      "learning_rate": 1.8512353250394253e-05,
+      "loss": 1.6345,
+      "step": 9304
+    },
+    {
+      "epoch": 2.4451670115282296,
+      "grad_norm": 0.5729696750640869,
+      "learning_rate": 1.8494830909409497e-05,
+      "loss": 1.6356,
+      "step": 9306
+    },
+    {
+      "epoch": 2.445692514861891,
+      "grad_norm": 0.6375647783279419,
+      "learning_rate": 1.847730856842474e-05,
+      "loss": 1.6262,
+      "step": 9308
+    },
+    {
+      "epoch": 2.446218018195553,
+      "grad_norm": 0.599912166595459,
+      "learning_rate": 1.845978622743999e-05,
+      "loss": 1.5833,
+      "step": 9310
+    },
+    {
+      "epoch": 2.4467435215292146,
+      "grad_norm": 0.6265223026275635,
+      "learning_rate": 1.844226388645523e-05,
+      "loss": 1.5683,
+      "step": 9312
+    },
+    {
+      "epoch": 2.4472690248628766,
+      "grad_norm": 0.5674051642417908,
+      "learning_rate": 1.8424741545470474e-05,
+      "loss": 1.6312,
+      "step": 9314
+    },
+    {
+      "epoch": 2.447794528196538,
+      "grad_norm": 0.5264484286308289,
+      "learning_rate": 1.840721920448572e-05,
+      "loss": 1.6509,
+      "step": 9316
+    },
+    {
+      "epoch": 2.4483200315302,
+      "grad_norm": 0.5643804669380188,
+      "learning_rate": 1.8389696863500965e-05,
+      "loss": 1.6002,
+      "step": 9318
+    },
+    {
+      "epoch": 2.448845534863862,
+      "grad_norm": 0.5581643581390381,
+      "learning_rate": 1.837217452251621e-05,
+      "loss": 1.602,
+      "step": 9320
+    },
+    {
+      "epoch": 2.4493710381975236,
+      "grad_norm": 0.5524643063545227,
+      "learning_rate": 1.8354652181531454e-05,
+      "loss": 1.6092,
+      "step": 9322
+    },
+    {
+      "epoch": 2.4498965415311855,
+      "grad_norm": 0.7081454992294312,
+      "learning_rate": 1.8337129840546698e-05,
+      "loss": 1.5703,
+      "step": 9324
+    },
+    {
+      "epoch": 2.450422044864847,
+      "grad_norm": 0.6080058217048645,
+      "learning_rate": 1.8319607499561942e-05,
+      "loss": 1.6261,
+      "step": 9326
+    },
+    {
+      "epoch": 2.450947548198509,
+      "grad_norm": 0.5197411775588989,
+      "learning_rate": 1.8302085158577186e-05,
+      "loss": 1.6313,
+      "step": 9328
+    },
+    {
+      "epoch": 2.4514730515321705,
+      "grad_norm": 0.5948798060417175,
+      "learning_rate": 1.8284562817592434e-05,
+      "loss": 1.6415,
+      "step": 9330
+    },
+    {
+      "epoch": 2.4519985548658325,
+      "grad_norm": 0.4755839705467224,
+      "learning_rate": 1.8267040476607675e-05,
+      "loss": 1.5945,
+      "step": 9332
+    },
+    {
+      "epoch": 2.452524058199494,
+      "grad_norm": 0.577564537525177,
+      "learning_rate": 1.824951813562292e-05,
+      "loss": 1.6056,
+      "step": 9334
+    },
+    {
+      "epoch": 2.453049561533156,
+      "grad_norm": 0.5504733324050903,
+      "learning_rate": 1.8231995794638166e-05,
+      "loss": 1.5843,
+      "step": 9336
+    },
+    {
+      "epoch": 2.4535750648668175,
+      "grad_norm": 0.5584169626235962,
+      "learning_rate": 1.8214473453653407e-05,
+      "loss": 1.6318,
+      "step": 9338
+    },
+    {
+      "epoch": 2.4541005682004795,
+      "grad_norm": 0.5598637461662292,
+      "learning_rate": 1.8196951112668655e-05,
+      "loss": 1.5763,
+      "step": 9340
+    },
+    {
+      "epoch": 2.4546260715341415,
+      "grad_norm": 0.6208490133285522,
+      "learning_rate": 1.81794287716839e-05,
+      "loss": 1.6339,
+      "step": 9342
+    },
+    {
+      "epoch": 2.455151574867803,
+      "grad_norm": 0.5243476033210754,
+      "learning_rate": 1.816190643069914e-05,
+      "loss": 1.5918,
+      "step": 9344
+    },
+    {
+      "epoch": 2.455677078201465,
+      "grad_norm": 0.6066216826438904,
+      "learning_rate": 1.8144384089714387e-05,
+      "loss": 1.6419,
+      "step": 9346
+    },
+    {
+      "epoch": 2.4562025815351265,
+      "grad_norm": 0.5985385179519653,
+      "learning_rate": 1.812686174872963e-05,
+      "loss": 1.6074,
+      "step": 9348
+    },
+    {
+      "epoch": 2.4567280848687885,
+      "grad_norm": 0.6051018834114075,
+      "learning_rate": 1.8109339407744876e-05,
+      "loss": 1.6076,
+      "step": 9350
+    },
+    {
+      "epoch": 2.45725358820245,
+      "grad_norm": 0.5187597274780273,
+      "learning_rate": 1.809181706676012e-05,
+      "loss": 1.6128,
+      "step": 9352
+    },
+    {
+      "epoch": 2.457779091536112,
+      "grad_norm": 0.576802670955658,
+      "learning_rate": 1.8074294725775364e-05,
+      "loss": 1.6185,
+      "step": 9354
+    },
+    {
+      "epoch": 2.458304594869774,
+      "grad_norm": 0.6253780722618103,
+      "learning_rate": 1.805677238479061e-05,
+      "loss": 1.5805,
+      "step": 9356
+    },
+    {
+      "epoch": 2.4588300982034355,
+      "grad_norm": 0.5701897740364075,
+      "learning_rate": 1.8039250043805852e-05,
+      "loss": 1.622,
+      "step": 9358
+    },
+    {
+      "epoch": 2.4593556015370974,
+      "grad_norm": 0.5329028367996216,
+      "learning_rate": 1.8021727702821096e-05,
+      "loss": 1.616,
+      "step": 9360
+    },
+    {
+      "epoch": 2.459881104870759,
+      "grad_norm": 0.6378101706504822,
+      "learning_rate": 1.8004205361836344e-05,
+      "loss": 1.6283,
+      "step": 9362
+    },
+    {
+      "epoch": 2.460406608204421,
+      "grad_norm": 0.5629276633262634,
+      "learning_rate": 1.7986683020851585e-05,
+      "loss": 1.5953,
+      "step": 9364
+    },
+    {
+      "epoch": 2.4609321115380824,
+      "grad_norm": 0.55552738904953,
+      "learning_rate": 1.7969160679866832e-05,
+      "loss": 1.588,
+      "step": 9366
+    },
+    {
+      "epoch": 2.4614576148717444,
+      "grad_norm": 0.6844131350517273,
+      "learning_rate": 1.7951638338882077e-05,
+      "loss": 1.6477,
+      "step": 9368
+    },
+    {
+      "epoch": 2.461983118205406,
+      "grad_norm": 0.5406688451766968,
+      "learning_rate": 1.7934115997897317e-05,
+      "loss": 1.6109,
+      "step": 9370
+    },
+    {
+      "epoch": 2.462508621539068,
+      "grad_norm": 0.5940281748771667,
+      "learning_rate": 1.7916593656912565e-05,
+      "loss": 1.5788,
+      "step": 9372
+    },
+    {
+      "epoch": 2.4630341248727294,
+      "grad_norm": 0.5564101338386536,
+      "learning_rate": 1.789907131592781e-05,
+      "loss": 1.5828,
+      "step": 9374
+    },
+    {
+      "epoch": 2.4635596282063914,
+      "grad_norm": 0.8060266971588135,
+      "learning_rate": 1.7881548974943053e-05,
+      "loss": 1.5922,
+      "step": 9376
+    },
+    {
+      "epoch": 2.4640851315400534,
+      "grad_norm": 0.6775248646736145,
+      "learning_rate": 1.7864026633958297e-05,
+      "loss": 1.6292,
+      "step": 9378
+    },
+    {
+      "epoch": 2.464610634873715,
+      "grad_norm": 0.5210546851158142,
+      "learning_rate": 1.784650429297354e-05,
+      "loss": 1.5921,
+      "step": 9380
+    },
+    {
+      "epoch": 2.465136138207377,
+      "grad_norm": 0.5210246443748474,
+      "learning_rate": 1.7828981951988786e-05,
+      "loss": 1.5947,
+      "step": 9382
+    },
+    {
+      "epoch": 2.4656616415410384,
+      "grad_norm": 0.5449514985084534,
+      "learning_rate": 1.781145961100403e-05,
+      "loss": 1.5716,
+      "step": 9384
+    },
+    {
+      "epoch": 2.4661871448747004,
+      "grad_norm": 0.5264145135879517,
+      "learning_rate": 1.7793937270019274e-05,
+      "loss": 1.5807,
+      "step": 9386
+    },
+    {
+      "epoch": 2.466712648208362,
+      "grad_norm": 0.6389797329902649,
+      "learning_rate": 1.777641492903452e-05,
+      "loss": 1.5949,
+      "step": 9388
+    },
+    {
+      "epoch": 2.467238151542024,
+      "grad_norm": 0.5606738924980164,
+      "learning_rate": 1.7758892588049762e-05,
+      "loss": 1.5973,
+      "step": 9390
+    },
+    {
+      "epoch": 2.467763654875686,
+      "grad_norm": 0.634388267993927,
+      "learning_rate": 1.774137024706501e-05,
+      "loss": 1.6466,
+      "step": 9392
+    },
+    {
+      "epoch": 2.4682891582093474,
+      "grad_norm": 0.6049827337265015,
+      "learning_rate": 1.7723847906080254e-05,
+      "loss": 1.6121,
+      "step": 9394
+    },
+    {
+      "epoch": 2.4688146615430093,
+      "grad_norm": 0.6313089728355408,
+      "learning_rate": 1.77063255650955e-05,
+      "loss": 1.6176,
+      "step": 9396
+    },
+    {
+      "epoch": 2.469340164876671,
+      "grad_norm": 0.6693941950798035,
+      "learning_rate": 1.7688803224110743e-05,
+      "loss": 1.6162,
+      "step": 9398
+    },
+    {
+      "epoch": 2.469865668210333,
+      "grad_norm": 0.5389882326126099,
+      "learning_rate": 1.7671280883125987e-05,
+      "loss": 1.5935,
+      "step": 9400
+    },
+    {
+      "epoch": 2.4703911715439943,
+      "grad_norm": 0.6095534563064575,
+      "learning_rate": 1.765375854214123e-05,
+      "loss": 1.624,
+      "step": 9402
+    },
+    {
+      "epoch": 2.4709166748776563,
+      "grad_norm": 0.5612223744392395,
+      "learning_rate": 1.7636236201156475e-05,
+      "loss": 1.64,
+      "step": 9404
+    },
+    {
+      "epoch": 2.471442178211318,
+      "grad_norm": 0.58417147397995,
+      "learning_rate": 1.761871386017172e-05,
+      "loss": 1.5979,
+      "step": 9406
+    },
+    {
+      "epoch": 2.47196768154498,
+      "grad_norm": 0.5258997082710266,
+      "learning_rate": 1.7601191519186963e-05,
+      "loss": 1.5915,
+      "step": 9408
+    },
+    {
+      "epoch": 2.4724931848786413,
+      "grad_norm": 0.5420629382133484,
+      "learning_rate": 1.7583669178202208e-05,
+      "loss": 1.6292,
+      "step": 9410
+    },
+    {
+      "epoch": 2.4730186882123033,
+      "grad_norm": 0.5879486799240112,
+      "learning_rate": 1.7566146837217455e-05,
+      "loss": 1.5718,
+      "step": 9412
+    },
+    {
+      "epoch": 2.4735441915459653,
+      "grad_norm": 0.5970991253852844,
+      "learning_rate": 1.7548624496232696e-05,
+      "loss": 1.6217,
+      "step": 9414
+    },
+    {
+      "epoch": 2.474069694879627,
+      "grad_norm": 0.5610321760177612,
+      "learning_rate": 1.753110215524794e-05,
+      "loss": 1.6234,
+      "step": 9416
+    },
+    {
+      "epoch": 2.4745951982132888,
+      "grad_norm": 0.6712002158164978,
+      "learning_rate": 1.7513579814263188e-05,
+      "loss": 1.5815,
+      "step": 9418
+    },
+    {
+      "epoch": 2.4751207015469503,
+      "grad_norm": 0.5543602108955383,
+      "learning_rate": 1.7496057473278432e-05,
+      "loss": 1.616,
+      "step": 9420
+    },
+    {
+      "epoch": 2.4756462048806123,
+      "grad_norm": 0.6275373101234436,
+      "learning_rate": 1.7478535132293676e-05,
+      "loss": 1.61,
+      "step": 9422
+    },
+    {
+      "epoch": 2.476171708214274,
+      "grad_norm": 0.6385626792907715,
+      "learning_rate": 1.746101279130892e-05,
+      "loss": 1.5826,
+      "step": 9424
+    },
+    {
+      "epoch": 2.4766972115479358,
+      "grad_norm": 0.5762410759925842,
+      "learning_rate": 1.7443490450324164e-05,
+      "loss": 1.5961,
+      "step": 9426
+    },
+    {
+      "epoch": 2.4772227148815977,
+      "grad_norm": 0.5379599332809448,
+      "learning_rate": 1.742596810933941e-05,
+      "loss": 1.6381,
+      "step": 9428
+    },
+    {
+      "epoch": 2.4777482182152593,
+      "grad_norm": 0.5608949661254883,
+      "learning_rate": 1.7408445768354653e-05,
+      "loss": 1.6073,
+      "step": 9430
+    },
+    {
+      "epoch": 2.4782737215489212,
+      "grad_norm": 0.5401448607444763,
+      "learning_rate": 1.7390923427369897e-05,
+      "loss": 1.6178,
+      "step": 9432
+    },
+    {
+      "epoch": 2.4787992248825828,
+      "grad_norm": 0.648591935634613,
+      "learning_rate": 1.737340108638514e-05,
+      "loss": 1.5938,
+      "step": 9434
+    },
+    {
+      "epoch": 2.4793247282162447,
+      "grad_norm": 0.710050106048584,
+      "learning_rate": 1.7355878745400385e-05,
+      "loss": 1.6104,
+      "step": 9436
+    },
+    {
+      "epoch": 2.4798502315499062,
+      "grad_norm": 0.5923548340797424,
+      "learning_rate": 1.7338356404415633e-05,
+      "loss": 1.6235,
+      "step": 9438
+    },
+    {
+      "epoch": 2.480375734883568,
+      "grad_norm": 0.6905649900436401,
+      "learning_rate": 1.7320834063430874e-05,
+      "loss": 1.5891,
+      "step": 9440
+    },
+    {
+      "epoch": 2.48090123821723,
+      "grad_norm": 0.6398130059242249,
+      "learning_rate": 1.7303311722446118e-05,
+      "loss": 1.5641,
+      "step": 9442
+    },
+    {
+      "epoch": 2.4814267415508917,
+      "grad_norm": 0.6153111457824707,
+      "learning_rate": 1.7285789381461365e-05,
+      "loss": 1.6005,
+      "step": 9444
+    },
+    {
+      "epoch": 2.4819522448845532,
+      "grad_norm": 0.5550106763839722,
+      "learning_rate": 1.7268267040476606e-05,
+      "loss": 1.5992,
+      "step": 9446
+    },
+    {
+      "epoch": 2.482477748218215,
+      "grad_norm": 0.5491114258766174,
+      "learning_rate": 1.7250744699491854e-05,
+      "loss": 1.5753,
+      "step": 9448
+    },
+    {
+      "epoch": 2.483003251551877,
+      "grad_norm": 0.5267966985702515,
+      "learning_rate": 1.7233222358507098e-05,
+      "loss": 1.5962,
+      "step": 9450
+    },
+    {
+      "epoch": 2.4835287548855387,
+      "grad_norm": 0.6979205012321472,
+      "learning_rate": 1.7215700017522342e-05,
+      "loss": 1.5966,
+      "step": 9452
+    },
+    {
+      "epoch": 2.4840542582192007,
+      "grad_norm": 0.5285736918449402,
+      "learning_rate": 1.7198177676537586e-05,
+      "loss": 1.5854,
+      "step": 9454
+    },
+    {
+      "epoch": 2.484579761552862,
+      "grad_norm": 0.5110975503921509,
+      "learning_rate": 1.718065533555283e-05,
+      "loss": 1.6251,
+      "step": 9456
+    },
+    {
+      "epoch": 2.485105264886524,
+      "grad_norm": 0.5918470025062561,
+      "learning_rate": 1.7163132994568078e-05,
+      "loss": 1.5997,
+      "step": 9458
+    },
+    {
+      "epoch": 2.4856307682201857,
+      "grad_norm": 0.5723440051078796,
+      "learning_rate": 1.714561065358332e-05,
+      "loss": 1.5997,
+      "step": 9460
+    },
+    {
+      "epoch": 2.4861562715538477,
+      "grad_norm": 0.7100077271461487,
+      "learning_rate": 1.7128088312598563e-05,
+      "loss": 1.5781,
+      "step": 9462
+    },
+    {
+      "epoch": 2.4866817748875096,
+      "grad_norm": 0.6034718155860901,
+      "learning_rate": 1.711056597161381e-05,
+      "loss": 1.6122,
+      "step": 9464
+    },
+    {
+      "epoch": 2.487207278221171,
+      "grad_norm": 0.6209121346473694,
+      "learning_rate": 1.709304363062905e-05,
+      "loss": 1.6385,
+      "step": 9466
+    },
+    {
+      "epoch": 2.487732781554833,
+      "grad_norm": 0.6605438590049744,
+      "learning_rate": 1.70755212896443e-05,
+      "loss": 1.5764,
+      "step": 9468
+    },
+    {
+      "epoch": 2.4882582848884947,
+      "grad_norm": 0.5605977773666382,
+      "learning_rate": 1.7057998948659543e-05,
+      "loss": 1.5951,
+      "step": 9470
+    },
+    {
+      "epoch": 2.4887837882221566,
+      "grad_norm": 0.5933828949928284,
+      "learning_rate": 1.7040476607674784e-05,
+      "loss": 1.6321,
+      "step": 9472
+    },
+    {
+      "epoch": 2.489309291555818,
+      "grad_norm": 0.6970359683036804,
+      "learning_rate": 1.702295426669003e-05,
+      "loss": 1.6462,
+      "step": 9474
+    },
+    {
+      "epoch": 2.48983479488948,
+      "grad_norm": 0.5877122282981873,
+      "learning_rate": 1.7005431925705276e-05,
+      "loss": 1.6058,
+      "step": 9476
+    },
+    {
+      "epoch": 2.490360298223142,
+      "grad_norm": 0.8348531126976013,
+      "learning_rate": 1.698790958472052e-05,
+      "loss": 1.6066,
+      "step": 9478
+    },
+    {
+      "epoch": 2.4908858015568036,
+      "grad_norm": 0.5285534262657166,
+      "learning_rate": 1.6970387243735764e-05,
+      "loss": 1.6145,
+      "step": 9480
+    },
+    {
+      "epoch": 2.4914113048904656,
+      "grad_norm": 0.5586251020431519,
+      "learning_rate": 1.6952864902751008e-05,
+      "loss": 1.618,
+      "step": 9482
+    },
+    {
+      "epoch": 2.491936808224127,
+      "grad_norm": 0.6798741817474365,
+      "learning_rate": 1.6935342561766252e-05,
+      "loss": 1.5923,
+      "step": 9484
+    },
+    {
+      "epoch": 2.492462311557789,
+      "grad_norm": 0.6360746622085571,
+      "learning_rate": 1.6917820220781496e-05,
+      "loss": 1.5958,
+      "step": 9486
+    },
+    {
+      "epoch": 2.4929878148914506,
+      "grad_norm": 0.6008504629135132,
+      "learning_rate": 1.690029787979674e-05,
+      "loss": 1.6299,
+      "step": 9488
+    },
+    {
+      "epoch": 2.4935133182251126,
+      "grad_norm": 0.6286047697067261,
+      "learning_rate": 1.6882775538811988e-05,
+      "loss": 1.6249,
+      "step": 9490
+    },
+    {
+      "epoch": 2.494038821558774,
+      "grad_norm": 0.5566932559013367,
+      "learning_rate": 1.686525319782723e-05,
+      "loss": 1.6175,
+      "step": 9492
+    },
+    {
+      "epoch": 2.494564324892436,
+      "grad_norm": 0.5241519212722778,
+      "learning_rate": 1.6847730856842477e-05,
+      "loss": 1.5978,
+      "step": 9494
+    },
+    {
+      "epoch": 2.4950898282260976,
+      "grad_norm": 0.5487810969352722,
+      "learning_rate": 1.683020851585772e-05,
+      "loss": 1.6114,
+      "step": 9496
+    },
+    {
+      "epoch": 2.4956153315597596,
+      "grad_norm": 0.518281877040863,
+      "learning_rate": 1.681268617487296e-05,
+      "loss": 1.5993,
+      "step": 9498
+    },
+    {
+      "epoch": 2.4961408348934215,
+      "grad_norm": 0.5262603163719177,
+      "learning_rate": 1.679516383388821e-05,
+      "loss": 1.5982,
+      "step": 9500
+    },
+    {
+      "epoch": 2.496666338227083,
+      "grad_norm": 0.560440182685852,
+      "learning_rate": 1.6777641492903453e-05,
+      "loss": 1.6281,
+      "step": 9502
+    },
+    {
+      "epoch": 2.497191841560745,
+      "grad_norm": 0.6901085376739502,
+      "learning_rate": 1.6760119151918697e-05,
+      "loss": 1.6069,
+      "step": 9504
+    },
+    {
+      "epoch": 2.4977173448944066,
+      "grad_norm": 0.6319709420204163,
+      "learning_rate": 1.674259681093394e-05,
+      "loss": 1.6053,
+      "step": 9506
+    },
+    {
+      "epoch": 2.4982428482280685,
+      "grad_norm": 0.6285496354103088,
+      "learning_rate": 1.6725074469949186e-05,
+      "loss": 1.6269,
+      "step": 9508
+    },
+    {
+      "epoch": 2.49876835156173,
+      "grad_norm": 0.5803909301757812,
+      "learning_rate": 1.670755212896443e-05,
+      "loss": 1.5917,
+      "step": 9510
+    },
+    {
+      "epoch": 2.499293854895392,
+      "grad_norm": 0.519927978515625,
+      "learning_rate": 1.6690029787979674e-05,
+      "loss": 1.6247,
+      "step": 9512
+    },
+    {
+      "epoch": 2.499819358229054,
+      "grad_norm": 0.5453500747680664,
+      "learning_rate": 1.6672507446994918e-05,
+      "loss": 1.582,
+      "step": 9514
+    },
+    {
+      "epoch": 2.5003448615627155,
+      "grad_norm": 0.7243382930755615,
+      "learning_rate": 1.6654985106010162e-05,
+      "loss": 1.5941,
+      "step": 9516
+    },
+    {
+      "epoch": 2.500870364896377,
+      "grad_norm": 0.6101195812225342,
+      "learning_rate": 1.6637462765025407e-05,
+      "loss": 1.5646,
+      "step": 9518
+    },
+    {
+      "epoch": 2.501395868230039,
+      "grad_norm": 0.5308881402015686,
+      "learning_rate": 1.6619940424040654e-05,
+      "loss": 1.5863,
+      "step": 9520
+    },
+    {
+      "epoch": 2.501921371563701,
+      "grad_norm": 0.624251663684845,
+      "learning_rate": 1.66024180830559e-05,
+      "loss": 1.5914,
+      "step": 9522
+    },
+    {
+      "epoch": 2.5024468748973625,
+      "grad_norm": 0.6056520938873291,
+      "learning_rate": 1.6584895742071143e-05,
+      "loss": 1.5891,
+      "step": 9524
+    },
+    {
+      "epoch": 2.5029723782310245,
+      "grad_norm": 0.6396739482879639,
+      "learning_rate": 1.6567373401086387e-05,
+      "loss": 1.6248,
+      "step": 9526
+    },
+    {
+      "epoch": 2.5034978815646864,
+      "grad_norm": 0.5661067962646484,
+      "learning_rate": 1.654985106010163e-05,
+      "loss": 1.5923,
+      "step": 9528
+    },
+    {
+      "epoch": 2.504023384898348,
+      "grad_norm": 0.601337194442749,
+      "learning_rate": 1.6532328719116875e-05,
+      "loss": 1.6074,
+      "step": 9530
+    },
+    {
+      "epoch": 2.5045488882320095,
+      "grad_norm": 0.603577733039856,
+      "learning_rate": 1.651480637813212e-05,
+      "loss": 1.6247,
+      "step": 9532
+    },
+    {
+      "epoch": 2.5050743915656715,
+      "grad_norm": 0.5598317980766296,
+      "learning_rate": 1.6497284037147363e-05,
+      "loss": 1.6353,
+      "step": 9534
+    },
+    {
+      "epoch": 2.5055998948993334,
+      "grad_norm": 0.6507828831672668,
+      "learning_rate": 1.6479761696162608e-05,
+      "loss": 1.6215,
+      "step": 9536
+    },
+    {
+      "epoch": 2.506125398232995,
+      "grad_norm": 0.5373896360397339,
+      "learning_rate": 1.6462239355177852e-05,
+      "loss": 1.5898,
+      "step": 9538
+    },
+    {
+      "epoch": 2.506650901566657,
+      "grad_norm": 0.6569491028785706,
+      "learning_rate": 1.64447170141931e-05,
+      "loss": 1.6061,
+      "step": 9540
+    },
+    {
+      "epoch": 2.5071764049003185,
+      "grad_norm": 0.6949899196624756,
+      "learning_rate": 1.642719467320834e-05,
+      "loss": 1.5774,
+      "step": 9542
+    },
+    {
+      "epoch": 2.5077019082339804,
+      "grad_norm": 0.6420979499816895,
+      "learning_rate": 1.6409672332223584e-05,
+      "loss": 1.5848,
+      "step": 9544
+    },
+    {
+      "epoch": 2.508227411567642,
+      "grad_norm": 0.5891046524047852,
+      "learning_rate": 1.6392149991238832e-05,
+      "loss": 1.6175,
+      "step": 9546
+    },
+    {
+      "epoch": 2.508752914901304,
+      "grad_norm": 0.5385305285453796,
+      "learning_rate": 1.6374627650254073e-05,
+      "loss": 1.6017,
+      "step": 9548
+    },
+    {
+      "epoch": 2.509278418234966,
+      "grad_norm": 0.5676510334014893,
+      "learning_rate": 1.635710530926932e-05,
+      "loss": 1.5919,
+      "step": 9550
+    },
+    {
+      "epoch": 2.5098039215686274,
+      "grad_norm": 0.6153396964073181,
+      "learning_rate": 1.6339582968284564e-05,
+      "loss": 1.6062,
+      "step": 9552
+    },
+    {
+      "epoch": 2.5103294249022894,
+      "grad_norm": 0.625535786151886,
+      "learning_rate": 1.6322060627299805e-05,
+      "loss": 1.6017,
+      "step": 9554
+    },
+    {
+      "epoch": 2.510854928235951,
+      "grad_norm": 0.5933648347854614,
+      "learning_rate": 1.6304538286315053e-05,
+      "loss": 1.6464,
+      "step": 9556
+    },
+    {
+      "epoch": 2.511380431569613,
+      "grad_norm": 0.5222533941268921,
+      "learning_rate": 1.6287015945330297e-05,
+      "loss": 1.6486,
+      "step": 9558
+    },
+    {
+      "epoch": 2.5119059349032744,
+      "grad_norm": 0.511766791343689,
+      "learning_rate": 1.626949360434554e-05,
+      "loss": 1.6111,
+      "step": 9560
+    },
+    {
+      "epoch": 2.5124314382369364,
+      "grad_norm": 0.6270935535430908,
+      "learning_rate": 1.6251971263360785e-05,
+      "loss": 1.6497,
+      "step": 9562
+    },
+    {
+      "epoch": 2.5129569415705983,
+      "grad_norm": 0.6762127876281738,
+      "learning_rate": 1.623444892237603e-05,
+      "loss": 1.5915,
+      "step": 9564
+    },
+    {
+      "epoch": 2.51348244490426,
+      "grad_norm": 0.7019566297531128,
+      "learning_rate": 1.6216926581391277e-05,
+      "loss": 1.5817,
+      "step": 9566
+    },
+    {
+      "epoch": 2.5140079482379214,
+      "grad_norm": 0.6118047833442688,
+      "learning_rate": 1.6199404240406518e-05,
+      "loss": 1.6241,
+      "step": 9568
+    },
+    {
+      "epoch": 2.5145334515715834,
+      "grad_norm": 0.5920790433883667,
+      "learning_rate": 1.6181881899421762e-05,
+      "loss": 1.607,
+      "step": 9570
+    },
+    {
+      "epoch": 2.5150589549052453,
+      "grad_norm": 0.5979636311531067,
+      "learning_rate": 1.616435955843701e-05,
+      "loss": 1.5874,
+      "step": 9572
+    },
+    {
+      "epoch": 2.515584458238907,
+      "grad_norm": 0.6232622265815735,
+      "learning_rate": 1.614683721745225e-05,
+      "loss": 1.5878,
+      "step": 9574
+    },
+    {
+      "epoch": 2.516109961572569,
+      "grad_norm": 0.5250893235206604,
+      "learning_rate": 1.6129314876467498e-05,
+      "loss": 1.5994,
+      "step": 9576
+    },
+    {
+      "epoch": 2.5166354649062304,
+      "grad_norm": 0.5512763857841492,
+      "learning_rate": 1.6111792535482742e-05,
+      "loss": 1.6041,
+      "step": 9578
+    },
+    {
+      "epoch": 2.5171609682398923,
+      "grad_norm": 0.5368601083755493,
+      "learning_rate": 1.6094270194497986e-05,
+      "loss": 1.6131,
+      "step": 9580
+    },
+    {
+      "epoch": 2.517686471573554,
+      "grad_norm": 0.6187223792076111,
+      "learning_rate": 1.607674785351323e-05,
+      "loss": 1.6238,
+      "step": 9582
+    },
+    {
+      "epoch": 2.518211974907216,
+      "grad_norm": 0.5821846127510071,
+      "learning_rate": 1.6059225512528475e-05,
+      "loss": 1.6262,
+      "step": 9584
+    },
+    {
+      "epoch": 2.518737478240878,
+      "grad_norm": 0.5627323985099792,
+      "learning_rate": 1.604170317154372e-05,
+      "loss": 1.6379,
+      "step": 9586
+    },
+    {
+      "epoch": 2.5192629815745393,
+      "grad_norm": 0.5581722259521484,
+      "learning_rate": 1.6024180830558963e-05,
+      "loss": 1.5849,
+      "step": 9588
+    },
+    {
+      "epoch": 2.5197884849082013,
+      "grad_norm": 0.677148163318634,
+      "learning_rate": 1.6006658489574207e-05,
+      "loss": 1.6116,
+      "step": 9590
+    },
+    {
+      "epoch": 2.520313988241863,
+      "grad_norm": 0.5418302416801453,
+      "learning_rate": 1.5989136148589455e-05,
+      "loss": 1.5981,
+      "step": 9592
+    },
+    {
+      "epoch": 2.520839491575525,
+      "grad_norm": 0.5710042715072632,
+      "learning_rate": 1.5971613807604695e-05,
+      "loss": 1.5515,
+      "step": 9594
+    },
+    {
+      "epoch": 2.5213649949091863,
+      "grad_norm": 0.5451944470405579,
+      "learning_rate": 1.5954091466619943e-05,
+      "loss": 1.603,
+      "step": 9596
+    },
+    {
+      "epoch": 2.5218904982428483,
+      "grad_norm": 0.7278261780738831,
+      "learning_rate": 1.5936569125635187e-05,
+      "loss": 1.6079,
+      "step": 9598
+    },
+    {
+      "epoch": 2.5224160015765102,
+      "grad_norm": 0.5918084979057312,
+      "learning_rate": 1.5919046784650428e-05,
+      "loss": 1.6009,
+      "step": 9600
+    },
+    {
+      "epoch": 2.5224160015765102,
+      "eval_loss": 1.6515436172485352,
+      "eval_runtime": 487.0879,
+      "eval_samples_per_second": 250.035,
+      "eval_steps_per_second": 31.255,
+      "step": 9600
+    },
+    {
+      "epoch": 2.5229415049101718,
+      "grad_norm": 0.6685560345649719,
+      "learning_rate": 1.5901524443665676e-05,
+      "loss": 1.6337,
+      "step": 9602
+    },
+    {
+      "epoch": 2.5234670082438333,
+      "grad_norm": 0.579550564289093,
+      "learning_rate": 1.588400210268092e-05,
+      "loss": 1.5911,
+      "step": 9604
+    },
+    {
+      "epoch": 2.5239925115774953,
+      "grad_norm": 0.522769570350647,
+      "learning_rate": 1.5866479761696164e-05,
+      "loss": 1.612,
+      "step": 9606
+    },
+    {
+      "epoch": 2.5245180149111572,
+      "grad_norm": 0.6259097456932068,
+      "learning_rate": 1.5848957420711408e-05,
+      "loss": 1.6076,
+      "step": 9608
+    },
+    {
+      "epoch": 2.5250435182448188,
+      "grad_norm": 0.5262504816055298,
+      "learning_rate": 1.5831435079726652e-05,
+      "loss": 1.5794,
+      "step": 9610
+    },
+    {
+      "epoch": 2.5255690215784807,
+      "grad_norm": 0.6325481534004211,
+      "learning_rate": 1.5813912738741896e-05,
+      "loss": 1.5997,
+      "step": 9612
+    },
+    {
+      "epoch": 2.5260945249121423,
+      "grad_norm": 0.7307614088058472,
+      "learning_rate": 1.579639039775714e-05,
+      "loss": 1.624,
+      "step": 9614
+    },
+    {
+      "epoch": 2.5266200282458042,
+      "grad_norm": 0.5761781334877014,
+      "learning_rate": 1.5778868056772385e-05,
+      "loss": 1.6469,
+      "step": 9616
+    },
+    {
+      "epoch": 2.5271455315794658,
+      "grad_norm": 0.543783962726593,
+      "learning_rate": 1.576134571578763e-05,
+      "loss": 1.613,
+      "step": 9618
+    },
+    {
+      "epoch": 2.5276710349131277,
+      "grad_norm": 0.5635935068130493,
+      "learning_rate": 1.5743823374802873e-05,
+      "loss": 1.6249,
+      "step": 9620
+    },
+    {
+      "epoch": 2.5281965382467897,
+      "grad_norm": 0.5564760565757751,
+      "learning_rate": 1.572630103381812e-05,
+      "loss": 1.5941,
+      "step": 9622
+    },
+    {
+      "epoch": 2.528722041580451,
+      "grad_norm": 0.7148152589797974,
+      "learning_rate": 1.5708778692833365e-05,
+      "loss": 1.6357,
+      "step": 9624
+    },
+    {
+      "epoch": 2.529247544914113,
+      "grad_norm": 0.7149800062179565,
+      "learning_rate": 1.5691256351848606e-05,
+      "loss": 1.6298,
+      "step": 9626
+    },
+    {
+      "epoch": 2.5297730482477747,
+      "grad_norm": 0.6550901532173157,
+      "learning_rate": 1.5673734010863853e-05,
+      "loss": 1.6008,
+      "step": 9628
+    },
+    {
+      "epoch": 2.5302985515814367,
+      "grad_norm": 0.5404581427574158,
+      "learning_rate": 1.5656211669879097e-05,
+      "loss": 1.5608,
+      "step": 9630
+    },
+    {
+      "epoch": 2.530824054915098,
+      "grad_norm": 0.674629807472229,
+      "learning_rate": 1.563868932889434e-05,
+      "loss": 1.6215,
+      "step": 9632
+    },
+    {
+      "epoch": 2.53134955824876,
+      "grad_norm": 0.6244335770606995,
+      "learning_rate": 1.5621166987909586e-05,
+      "loss": 1.6011,
+      "step": 9634
+    },
+    {
+      "epoch": 2.531875061582422,
+      "grad_norm": 0.6089237332344055,
+      "learning_rate": 1.560364464692483e-05,
+      "loss": 1.5962,
+      "step": 9636
+    },
+    {
+      "epoch": 2.5324005649160837,
+      "grad_norm": 0.5669505000114441,
+      "learning_rate": 1.5586122305940074e-05,
+      "loss": 1.5876,
+      "step": 9638
+    },
+    {
+      "epoch": 2.532926068249745,
+      "grad_norm": 0.5484370589256287,
+      "learning_rate": 1.5568599964955318e-05,
+      "loss": 1.5994,
+      "step": 9640
+    },
+    {
+      "epoch": 2.533451571583407,
+      "grad_norm": 0.6164976954460144,
+      "learning_rate": 1.5551077623970562e-05,
+      "loss": 1.5965,
+      "step": 9642
+    },
+    {
+      "epoch": 2.533977074917069,
+      "grad_norm": 0.6072930693626404,
+      "learning_rate": 1.5533555282985807e-05,
+      "loss": 1.6167,
+      "step": 9644
+    },
+    {
+      "epoch": 2.5345025782507307,
+      "grad_norm": 0.6487789750099182,
+      "learning_rate": 1.551603294200105e-05,
+      "loss": 1.6376,
+      "step": 9646
+    },
+    {
+      "epoch": 2.5350280815843926,
+      "grad_norm": 0.641263484954834,
+      "learning_rate": 1.54985106010163e-05,
+      "loss": 1.6516,
+      "step": 9648
+    },
+    {
+      "epoch": 2.535553584918054,
+      "grad_norm": 0.6348699927330017,
+      "learning_rate": 1.548098826003154e-05,
+      "loss": 1.5979,
+      "step": 9650
+    },
+    {
+      "epoch": 2.536079088251716,
+      "grad_norm": 0.5915436744689941,
+      "learning_rate": 1.5463465919046787e-05,
+      "loss": 1.6025,
+      "step": 9652
+    },
+    {
+      "epoch": 2.5366045915853777,
+      "grad_norm": 0.6235045194625854,
+      "learning_rate": 1.544594357806203e-05,
+      "loss": 1.6093,
+      "step": 9654
+    },
+    {
+      "epoch": 2.5371300949190396,
+      "grad_norm": 0.5824310183525085,
+      "learning_rate": 1.542842123707727e-05,
+      "loss": 1.6317,
+      "step": 9656
+    },
+    {
+      "epoch": 2.5376555982527016,
+      "grad_norm": 0.7158166766166687,
+      "learning_rate": 1.541089889609252e-05,
+      "loss": 1.609,
+      "step": 9658
+    },
+    {
+      "epoch": 2.538181101586363,
+      "grad_norm": 0.6311430335044861,
+      "learning_rate": 1.5393376555107763e-05,
+      "loss": 1.6262,
+      "step": 9660
+    },
+    {
+      "epoch": 2.538706604920025,
+      "grad_norm": 0.6356306076049805,
+      "learning_rate": 1.5375854214123008e-05,
+      "loss": 1.5805,
+      "step": 9662
+    },
+    {
+      "epoch": 2.5392321082536866,
+      "grad_norm": 0.57065749168396,
+      "learning_rate": 1.5358331873138252e-05,
+      "loss": 1.6181,
+      "step": 9664
+    },
+    {
+      "epoch": 2.5397576115873486,
+      "grad_norm": 0.7448508143424988,
+      "learning_rate": 1.5340809532153496e-05,
+      "loss": 1.6235,
+      "step": 9666
+    },
+    {
+      "epoch": 2.54028311492101,
+      "grad_norm": 0.5627465844154358,
+      "learning_rate": 1.5323287191168743e-05,
+      "loss": 1.6273,
+      "step": 9668
+    },
+    {
+      "epoch": 2.540808618254672,
+      "grad_norm": 0.6133062839508057,
+      "learning_rate": 1.5305764850183984e-05,
+      "loss": 1.5949,
+      "step": 9670
+    },
+    {
+      "epoch": 2.541334121588334,
+      "grad_norm": 0.5348150134086609,
+      "learning_rate": 1.528824250919923e-05,
+      "loss": 1.6181,
+      "step": 9672
+    },
+    {
+      "epoch": 2.5418596249219956,
+      "grad_norm": 0.5559690594673157,
+      "learning_rate": 1.5270720168214476e-05,
+      "loss": 1.6224,
+      "step": 9674
+    },
+    {
+      "epoch": 2.542385128255657,
+      "grad_norm": 0.612876832485199,
+      "learning_rate": 1.5253197827229718e-05,
+      "loss": 1.6112,
+      "step": 9676
+    },
+    {
+      "epoch": 2.542910631589319,
+      "grad_norm": 0.697488009929657,
+      "learning_rate": 1.5235675486244963e-05,
+      "loss": 1.598,
+      "step": 9678
+    },
+    {
+      "epoch": 2.543436134922981,
+      "grad_norm": 0.6256018280982971,
+      "learning_rate": 1.5218153145260209e-05,
+      "loss": 1.6273,
+      "step": 9680
+    },
+    {
+      "epoch": 2.5439616382566426,
+      "grad_norm": 0.583524763584137,
+      "learning_rate": 1.5200630804275451e-05,
+      "loss": 1.648,
+      "step": 9682
+    },
+    {
+      "epoch": 2.5444871415903045,
+      "grad_norm": 0.682237446308136,
+      "learning_rate": 1.5183108463290697e-05,
+      "loss": 1.5871,
+      "step": 9684
+    },
+    {
+      "epoch": 2.5450126449239665,
+      "grad_norm": 0.5728941559791565,
+      "learning_rate": 1.5165586122305941e-05,
+      "loss": 1.5953,
+      "step": 9686
+    },
+    {
+      "epoch": 2.545538148257628,
+      "grad_norm": 0.5490486025810242,
+      "learning_rate": 1.5148063781321184e-05,
+      "loss": 1.6289,
+      "step": 9688
+    },
+    {
+      "epoch": 2.5460636515912896,
+      "grad_norm": 0.5297114849090576,
+      "learning_rate": 1.513054144033643e-05,
+      "loss": 1.6445,
+      "step": 9690
+    },
+    {
+      "epoch": 2.5465891549249515,
+      "grad_norm": 0.5523902177810669,
+      "learning_rate": 1.5113019099351675e-05,
+      "loss": 1.5715,
+      "step": 9692
+    },
+    {
+      "epoch": 2.5471146582586135,
+      "grad_norm": 0.5788924694061279,
+      "learning_rate": 1.509549675836692e-05,
+      "loss": 1.6178,
+      "step": 9694
+    },
+    {
+      "epoch": 2.547640161592275,
+      "grad_norm": 0.5816612839698792,
+      "learning_rate": 1.5077974417382162e-05,
+      "loss": 1.5855,
+      "step": 9696
+    },
+    {
+      "epoch": 2.548165664925937,
+      "grad_norm": 0.6361863017082214,
+      "learning_rate": 1.5060452076397408e-05,
+      "loss": 1.612,
+      "step": 9698
+    },
+    {
+      "epoch": 2.5486911682595985,
+      "grad_norm": 0.6106998324394226,
+      "learning_rate": 1.5042929735412654e-05,
+      "loss": 1.6482,
+      "step": 9700
+    },
+    {
+      "epoch": 2.5492166715932605,
+      "grad_norm": 0.623910665512085,
+      "learning_rate": 1.5025407394427896e-05,
+      "loss": 1.5743,
+      "step": 9702
+    },
+    {
+      "epoch": 2.549742174926922,
+      "grad_norm": 0.7251524925231934,
+      "learning_rate": 1.500788505344314e-05,
+      "loss": 1.6176,
+      "step": 9704
+    },
+    {
+      "epoch": 2.550267678260584,
+      "grad_norm": 0.5366904735565186,
+      "learning_rate": 1.4990362712458386e-05,
+      "loss": 1.5826,
+      "step": 9706
+    },
+    {
+      "epoch": 2.550793181594246,
+      "grad_norm": 0.5754824876785278,
+      "learning_rate": 1.4972840371473629e-05,
+      "loss": 1.6006,
+      "step": 9708
+    },
+    {
+      "epoch": 2.5513186849279075,
+      "grad_norm": 0.6944889426231384,
+      "learning_rate": 1.4955318030488875e-05,
+      "loss": 1.6318,
+      "step": 9710
+    },
+    {
+      "epoch": 2.5518441882615694,
+      "grad_norm": 0.6008621454238892,
+      "learning_rate": 1.4937795689504119e-05,
+      "loss": 1.6363,
+      "step": 9712
+    },
+    {
+      "epoch": 2.552369691595231,
+      "grad_norm": 0.6449068784713745,
+      "learning_rate": 1.4920273348519361e-05,
+      "loss": 1.6009,
+      "step": 9714
+    },
+    {
+      "epoch": 2.552895194928893,
+      "grad_norm": 0.687323272228241,
+      "learning_rate": 1.4902751007534607e-05,
+      "loss": 1.6247,
+      "step": 9716
+    },
+    {
+      "epoch": 2.5534206982625545,
+      "grad_norm": 0.6217597126960754,
+      "learning_rate": 1.4885228666549853e-05,
+      "loss": 1.6128,
+      "step": 9718
+    },
+    {
+      "epoch": 2.5539462015962164,
+      "grad_norm": 0.6320381760597229,
+      "learning_rate": 1.4867706325565095e-05,
+      "loss": 1.643,
+      "step": 9720
+    },
+    {
+      "epoch": 2.5544717049298784,
+      "grad_norm": 0.604827344417572,
+      "learning_rate": 1.485018398458034e-05,
+      "loss": 1.6073,
+      "step": 9722
+    },
+    {
+      "epoch": 2.55499720826354,
+      "grad_norm": 0.5245053768157959,
+      "learning_rate": 1.4832661643595585e-05,
+      "loss": 1.5963,
+      "step": 9724
+    },
+    {
+      "epoch": 2.5555227115972015,
+      "grad_norm": 0.6272046566009521,
+      "learning_rate": 1.4815139302610828e-05,
+      "loss": 1.5939,
+      "step": 9726
+    },
+    {
+      "epoch": 2.5560482149308634,
+      "grad_norm": 0.5605988502502441,
+      "learning_rate": 1.4797616961626074e-05,
+      "loss": 1.6283,
+      "step": 9728
+    },
+    {
+      "epoch": 2.5565737182645254,
+      "grad_norm": 0.5442174077033997,
+      "learning_rate": 1.4780094620641318e-05,
+      "loss": 1.5942,
+      "step": 9730
+    },
+    {
+      "epoch": 2.557099221598187,
+      "grad_norm": 0.806302011013031,
+      "learning_rate": 1.4762572279656564e-05,
+      "loss": 1.6132,
+      "step": 9732
+    },
+    {
+      "epoch": 2.557624724931849,
+      "grad_norm": 0.6235170960426331,
+      "learning_rate": 1.4745049938671806e-05,
+      "loss": 1.6062,
+      "step": 9734
+    },
+    {
+      "epoch": 2.5581502282655104,
+      "grad_norm": 0.6286174654960632,
+      "learning_rate": 1.4727527597687052e-05,
+      "loss": 1.6112,
+      "step": 9736
+    },
+    {
+      "epoch": 2.5586757315991724,
+      "grad_norm": 0.5533096194267273,
+      "learning_rate": 1.4710005256702296e-05,
+      "loss": 1.5906,
+      "step": 9738
+    },
+    {
+      "epoch": 2.559201234932834,
+      "grad_norm": 0.5867462754249573,
+      "learning_rate": 1.469248291571754e-05,
+      "loss": 1.5778,
+      "step": 9740
+    },
+    {
+      "epoch": 2.559726738266496,
+      "grad_norm": 0.5377023816108704,
+      "learning_rate": 1.4674960574732785e-05,
+      "loss": 1.5883,
+      "step": 9742
+    },
+    {
+      "epoch": 2.560252241600158,
+      "grad_norm": 0.5589543581008911,
+      "learning_rate": 1.465743823374803e-05,
+      "loss": 1.602,
+      "step": 9744
+    },
+    {
+      "epoch": 2.5607777449338194,
+      "grad_norm": 0.6195635795593262,
+      "learning_rate": 1.4639915892763273e-05,
+      "loss": 1.622,
+      "step": 9746
+    },
+    {
+      "epoch": 2.5613032482674813,
+      "grad_norm": 0.5796042680740356,
+      "learning_rate": 1.4622393551778519e-05,
+      "loss": 1.6333,
+      "step": 9748
+    },
+    {
+      "epoch": 2.561828751601143,
+      "grad_norm": 0.5899613499641418,
+      "learning_rate": 1.4604871210793763e-05,
+      "loss": 1.5945,
+      "step": 9750
+    },
+    {
+      "epoch": 2.562354254934805,
+      "grad_norm": 0.5779309272766113,
+      "learning_rate": 1.4587348869809006e-05,
+      "loss": 1.59,
+      "step": 9752
+    },
+    {
+      "epoch": 2.5628797582684664,
+      "grad_norm": 0.6350328326225281,
+      "learning_rate": 1.4569826528824251e-05,
+      "loss": 1.6087,
+      "step": 9754
+    },
+    {
+      "epoch": 2.5634052616021283,
+      "grad_norm": 0.5645972490310669,
+      "learning_rate": 1.4552304187839497e-05,
+      "loss": 1.6058,
+      "step": 9756
+    },
+    {
+      "epoch": 2.5639307649357903,
+      "grad_norm": 0.7305320501327515,
+      "learning_rate": 1.453478184685474e-05,
+      "loss": 1.5944,
+      "step": 9758
+    },
+    {
+      "epoch": 2.564456268269452,
+      "grad_norm": 0.600288987159729,
+      "learning_rate": 1.4517259505869984e-05,
+      "loss": 1.589,
+      "step": 9760
+    },
+    {
+      "epoch": 2.5649817716031134,
+      "grad_norm": 0.6395924091339111,
+      "learning_rate": 1.449973716488523e-05,
+      "loss": 1.5878,
+      "step": 9762
+    },
+    {
+      "epoch": 2.5655072749367753,
+      "grad_norm": 0.6453663110733032,
+      "learning_rate": 1.4482214823900476e-05,
+      "loss": 1.5795,
+      "step": 9764
+    },
+    {
+      "epoch": 2.5660327782704373,
+      "grad_norm": 0.6681911945343018,
+      "learning_rate": 1.4464692482915718e-05,
+      "loss": 1.5715,
+      "step": 9766
+    },
+    {
+      "epoch": 2.566558281604099,
+      "grad_norm": 0.8908346891403198,
+      "learning_rate": 1.4447170141930962e-05,
+      "loss": 1.604,
+      "step": 9768
+    },
+    {
+      "epoch": 2.567083784937761,
+      "grad_norm": 0.5479845404624939,
+      "learning_rate": 1.4429647800946208e-05,
+      "loss": 1.5835,
+      "step": 9770
+    },
+    {
+      "epoch": 2.5676092882714223,
+      "grad_norm": 0.5948062539100647,
+      "learning_rate": 1.441212545996145e-05,
+      "loss": 1.5969,
+      "step": 9772
+    },
+    {
+      "epoch": 2.5681347916050843,
+      "grad_norm": 0.5489128232002258,
+      "learning_rate": 1.4394603118976697e-05,
+      "loss": 1.6349,
+      "step": 9774
+    },
+    {
+      "epoch": 2.568660294938746,
+      "grad_norm": 0.6163507103919983,
+      "learning_rate": 1.437708077799194e-05,
+      "loss": 1.5721,
+      "step": 9776
+    },
+    {
+      "epoch": 2.569185798272408,
+      "grad_norm": 0.6882917881011963,
+      "learning_rate": 1.4359558437007183e-05,
+      "loss": 1.6179,
+      "step": 9778
+    },
+    {
+      "epoch": 2.5697113016060698,
+      "grad_norm": 0.5954879522323608,
+      "learning_rate": 1.434203609602243e-05,
+      "loss": 1.6136,
+      "step": 9780
+    },
+    {
+      "epoch": 2.5702368049397313,
+      "grad_norm": 0.6170343160629272,
+      "learning_rate": 1.4324513755037675e-05,
+      "loss": 1.585,
+      "step": 9782
+    },
+    {
+      "epoch": 2.5707623082733932,
+      "grad_norm": 0.5451639890670776,
+      "learning_rate": 1.4306991414052918e-05,
+      "loss": 1.5914,
+      "step": 9784
+    },
+    {
+      "epoch": 2.5712878116070548,
+      "grad_norm": 0.5690667629241943,
+      "learning_rate": 1.4289469073068162e-05,
+      "loss": 1.6272,
+      "step": 9786
+    },
+    {
+      "epoch": 2.5718133149407167,
+      "grad_norm": 0.7590607404708862,
+      "learning_rate": 1.4271946732083408e-05,
+      "loss": 1.6112,
+      "step": 9788
+    },
+    {
+      "epoch": 2.5723388182743783,
+      "grad_norm": 0.55152827501297,
+      "learning_rate": 1.425442439109865e-05,
+      "loss": 1.6291,
+      "step": 9790
+    },
+    {
+      "epoch": 2.5728643216080402,
+      "grad_norm": 0.5713075399398804,
+      "learning_rate": 1.4236902050113896e-05,
+      "loss": 1.6375,
+      "step": 9792
+    },
+    {
+      "epoch": 2.573389824941702,
+      "grad_norm": 0.5247568488121033,
+      "learning_rate": 1.421937970912914e-05,
+      "loss": 1.5643,
+      "step": 9794
+    },
+    {
+      "epoch": 2.5739153282753637,
+      "grad_norm": 0.7281646132469177,
+      "learning_rate": 1.4201857368144383e-05,
+      "loss": 1.5993,
+      "step": 9796
+    },
+    {
+      "epoch": 2.5744408316090253,
+      "grad_norm": 0.690531849861145,
+      "learning_rate": 1.4184335027159628e-05,
+      "loss": 1.6062,
+      "step": 9798
+    },
+    {
+      "epoch": 2.5749663349426872,
+      "grad_norm": 0.5356507897377014,
+      "learning_rate": 1.4166812686174874e-05,
+      "loss": 1.6073,
+      "step": 9800
+    },
+    {
+      "epoch": 2.575491838276349,
+      "grad_norm": 0.5589247941970825,
+      "learning_rate": 1.4149290345190118e-05,
+      "loss": 1.5874,
+      "step": 9802
+    },
+    {
+      "epoch": 2.5760173416100107,
+      "grad_norm": 0.570418119430542,
+      "learning_rate": 1.4131768004205363e-05,
+      "loss": 1.6107,
+      "step": 9804
+    },
+    {
+      "epoch": 2.5765428449436727,
+      "grad_norm": 0.6067774295806885,
+      "learning_rate": 1.4114245663220607e-05,
+      "loss": 1.6309,
+      "step": 9806
+    },
+    {
+      "epoch": 2.577068348277334,
+      "grad_norm": 0.5615005493164062,
+      "learning_rate": 1.4096723322235853e-05,
+      "loss": 1.6374,
+      "step": 9808
+    },
+    {
+      "epoch": 2.577593851610996,
+      "grad_norm": 0.5466208457946777,
+      "learning_rate": 1.4079200981251095e-05,
+      "loss": 1.6007,
+      "step": 9810
+    },
+    {
+      "epoch": 2.5781193549446577,
+      "grad_norm": 0.6845700740814209,
+      "learning_rate": 1.4061678640266341e-05,
+      "loss": 1.6121,
+      "step": 9812
+    },
+    {
+      "epoch": 2.5786448582783197,
+      "grad_norm": 0.6770169734954834,
+      "learning_rate": 1.4044156299281585e-05,
+      "loss": 1.6089,
+      "step": 9814
+    },
+    {
+      "epoch": 2.5791703616119817,
+      "grad_norm": 0.5839431881904602,
+      "learning_rate": 1.4026633958296828e-05,
+      "loss": 1.6216,
+      "step": 9816
+    },
+    {
+      "epoch": 2.579695864945643,
+      "grad_norm": 0.5246254205703735,
+      "learning_rate": 1.4009111617312074e-05,
+      "loss": 1.6482,
+      "step": 9818
+    },
+    {
+      "epoch": 2.580221368279305,
+      "grad_norm": 0.5945092439651489,
+      "learning_rate": 1.399158927632732e-05,
+      "loss": 1.6152,
+      "step": 9820
+    },
+    {
+      "epoch": 2.5807468716129667,
+      "grad_norm": 0.5551746487617493,
+      "learning_rate": 1.3974066935342562e-05,
+      "loss": 1.6123,
+      "step": 9822
+    },
+    {
+      "epoch": 2.5812723749466286,
+      "grad_norm": 0.5317911505699158,
+      "learning_rate": 1.3956544594357806e-05,
+      "loss": 1.6127,
+      "step": 9824
+    },
+    {
+      "epoch": 2.58179787828029,
+      "grad_norm": 0.5956730842590332,
+      "learning_rate": 1.3939022253373052e-05,
+      "loss": 1.6179,
+      "step": 9826
+    },
+    {
+      "epoch": 2.582323381613952,
+      "grad_norm": 0.5398740768432617,
+      "learning_rate": 1.3921499912388294e-05,
+      "loss": 1.6346,
+      "step": 9828
+    },
+    {
+      "epoch": 2.582848884947614,
+      "grad_norm": 0.5810662508010864,
+      "learning_rate": 1.390397757140354e-05,
+      "loss": 1.6218,
+      "step": 9830
+    },
+    {
+      "epoch": 2.5833743882812756,
+      "grad_norm": 0.5635197162628174,
+      "learning_rate": 1.3886455230418784e-05,
+      "loss": 1.5876,
+      "step": 9832
+    },
+    {
+      "epoch": 2.583899891614937,
+      "grad_norm": 0.626593291759491,
+      "learning_rate": 1.386893288943403e-05,
+      "loss": 1.596,
+      "step": 9834
+    },
+    {
+      "epoch": 2.584425394948599,
+      "grad_norm": 0.528767466545105,
+      "learning_rate": 1.3851410548449273e-05,
+      "loss": 1.6265,
+      "step": 9836
+    },
+    {
+      "epoch": 2.584950898282261,
+      "grad_norm": 0.5446393489837646,
+      "learning_rate": 1.3833888207464519e-05,
+      "loss": 1.6101,
+      "step": 9838
+    },
+    {
+      "epoch": 2.5854764016159226,
+      "grad_norm": 0.5221058130264282,
+      "learning_rate": 1.3816365866479763e-05,
+      "loss": 1.6393,
+      "step": 9840
+    },
+    {
+      "epoch": 2.5860019049495846,
+      "grad_norm": 0.721125602722168,
+      "learning_rate": 1.3798843525495005e-05,
+      "loss": 1.5934,
+      "step": 9842
+    },
+    {
+      "epoch": 2.5865274082832466,
+      "grad_norm": 0.6360064744949341,
+      "learning_rate": 1.3781321184510251e-05,
+      "loss": 1.613,
+      "step": 9844
+    },
+    {
+      "epoch": 2.587052911616908,
+      "grad_norm": 0.5796793699264526,
+      "learning_rate": 1.3763798843525497e-05,
+      "loss": 1.6069,
+      "step": 9846
+    },
+    {
+      "epoch": 2.5875784149505696,
+      "grad_norm": 0.6104748845100403,
+      "learning_rate": 1.374627650254074e-05,
+      "loss": 1.6083,
+      "step": 9848
+    },
+    {
+      "epoch": 2.5881039182842316,
+      "grad_norm": 0.5367170572280884,
+      "learning_rate": 1.3728754161555984e-05,
+      "loss": 1.6034,
+      "step": 9850
+    },
+    {
+      "epoch": 2.5886294216178936,
+      "grad_norm": 0.6482873558998108,
+      "learning_rate": 1.371123182057123e-05,
+      "loss": 1.5852,
+      "step": 9852
+    },
+    {
+      "epoch": 2.589154924951555,
+      "grad_norm": 0.5794985294342041,
+      "learning_rate": 1.3693709479586472e-05,
+      "loss": 1.5974,
+      "step": 9854
+    },
+    {
+      "epoch": 2.589680428285217,
+      "grad_norm": 0.6778134703636169,
+      "learning_rate": 1.3676187138601718e-05,
+      "loss": 1.5959,
+      "step": 9856
+    },
+    {
+      "epoch": 2.5902059316188786,
+      "grad_norm": 0.5274595618247986,
+      "learning_rate": 1.3658664797616962e-05,
+      "loss": 1.5888,
+      "step": 9858
+    },
+    {
+      "epoch": 2.5907314349525405,
+      "grad_norm": 0.648197591304779,
+      "learning_rate": 1.3641142456632205e-05,
+      "loss": 1.6369,
+      "step": 9860
+    },
+    {
+      "epoch": 2.591256938286202,
+      "grad_norm": 0.5107806324958801,
+      "learning_rate": 1.362362011564745e-05,
+      "loss": 1.5997,
+      "step": 9862
+    },
+    {
+      "epoch": 2.591782441619864,
+      "grad_norm": 0.5690180659294128,
+      "learning_rate": 1.3606097774662696e-05,
+      "loss": 1.6161,
+      "step": 9864
+    },
+    {
+      "epoch": 2.592307944953526,
+      "grad_norm": 0.5976451635360718,
+      "learning_rate": 1.358857543367794e-05,
+      "loss": 1.5956,
+      "step": 9866
+    },
+    {
+      "epoch": 2.5928334482871875,
+      "grad_norm": 0.5506587624549866,
+      "learning_rate": 1.3571053092693185e-05,
+      "loss": 1.5734,
+      "step": 9868
+    },
+    {
+      "epoch": 2.5933589516208495,
+      "grad_norm": 0.5882534384727478,
+      "learning_rate": 1.3553530751708429e-05,
+      "loss": 1.5834,
+      "step": 9870
+    },
+    {
+      "epoch": 2.593884454954511,
+      "grad_norm": 0.5685223340988159,
+      "learning_rate": 1.3536008410723675e-05,
+      "loss": 1.5997,
+      "step": 9872
+    },
+    {
+      "epoch": 2.594409958288173,
+      "grad_norm": 0.725277304649353,
+      "learning_rate": 1.3518486069738917e-05,
+      "loss": 1.573,
+      "step": 9874
+    },
+    {
+      "epoch": 2.5949354616218345,
+      "grad_norm": 0.563489556312561,
+      "learning_rate": 1.3500963728754163e-05,
+      "loss": 1.6113,
+      "step": 9876
+    },
+    {
+      "epoch": 2.5954609649554965,
+      "grad_norm": 0.5687708258628845,
+      "learning_rate": 1.3483441387769407e-05,
+      "loss": 1.6314,
+      "step": 9878
+    },
+    {
+      "epoch": 2.5959864682891585,
+      "grad_norm": 0.6753512620925903,
+      "learning_rate": 1.346591904678465e-05,
+      "loss": 1.5998,
+      "step": 9880
+    },
+    {
+      "epoch": 2.59651197162282,
+      "grad_norm": 0.7767539620399475,
+      "learning_rate": 1.3448396705799896e-05,
+      "loss": 1.5696,
+      "step": 9882
+    },
+    {
+      "epoch": 2.5970374749564815,
+      "grad_norm": 0.6438019275665283,
+      "learning_rate": 1.3430874364815142e-05,
+      "loss": 1.6511,
+      "step": 9884
+    },
+    {
+      "epoch": 2.5975629782901435,
+      "grad_norm": 0.673339307308197,
+      "learning_rate": 1.3413352023830384e-05,
+      "loss": 1.6375,
+      "step": 9886
+    },
+    {
+      "epoch": 2.5980884816238055,
+      "grad_norm": 0.6414830684661865,
+      "learning_rate": 1.3395829682845628e-05,
+      "loss": 1.6215,
+      "step": 9888
+    },
+    {
+      "epoch": 2.598613984957467,
+      "grad_norm": 0.5403038263320923,
+      "learning_rate": 1.3378307341860874e-05,
+      "loss": 1.5912,
+      "step": 9890
+    },
+    {
+      "epoch": 2.599139488291129,
+      "grad_norm": 0.6485872268676758,
+      "learning_rate": 1.3360785000876117e-05,
+      "loss": 1.6385,
+      "step": 9892
+    },
+    {
+      "epoch": 2.5996649916247905,
+      "grad_norm": 0.5818743109703064,
+      "learning_rate": 1.3343262659891362e-05,
+      "loss": 1.6295,
+      "step": 9894
+    },
+    {
+      "epoch": 2.6001904949584524,
+      "grad_norm": 0.6335439682006836,
+      "learning_rate": 1.3325740318906607e-05,
+      "loss": 1.5726,
+      "step": 9896
+    },
+    {
+      "epoch": 2.600715998292114,
+      "grad_norm": 0.5946658253669739,
+      "learning_rate": 1.3308217977921849e-05,
+      "loss": 1.6393,
+      "step": 9898
+    },
+    {
+      "epoch": 2.601241501625776,
+      "grad_norm": 0.5518163442611694,
+      "learning_rate": 1.3290695636937095e-05,
+      "loss": 1.5873,
+      "step": 9900
+    },
+    {
+      "epoch": 2.601767004959438,
+      "grad_norm": 0.5967461466789246,
+      "learning_rate": 1.327317329595234e-05,
+      "loss": 1.596,
+      "step": 9902
+    },
+    {
+      "epoch": 2.6022925082930994,
+      "grad_norm": 0.613856852054596,
+      "learning_rate": 1.3255650954967585e-05,
+      "loss": 1.6073,
+      "step": 9904
+    },
+    {
+      "epoch": 2.6028180116267614,
+      "grad_norm": 0.5403199791908264,
+      "learning_rate": 1.3238128613982827e-05,
+      "loss": 1.623,
+      "step": 9906
+    },
+    {
+      "epoch": 2.603343514960423,
+      "grad_norm": 0.6116963624954224,
+      "learning_rate": 1.3220606272998073e-05,
+      "loss": 1.6139,
+      "step": 9908
+    },
+    {
+      "epoch": 2.603869018294085,
+      "grad_norm": 0.6207565069198608,
+      "learning_rate": 1.320308393201332e-05,
+      "loss": 1.602,
+      "step": 9910
+    },
+    {
+      "epoch": 2.6043945216277464,
+      "grad_norm": 0.6587697863578796,
+      "learning_rate": 1.3185561591028562e-05,
+      "loss": 1.6115,
+      "step": 9912
+    },
+    {
+      "epoch": 2.6049200249614084,
+      "grad_norm": 0.5676011443138123,
+      "learning_rate": 1.3168039250043806e-05,
+      "loss": 1.6086,
+      "step": 9914
+    },
+    {
+      "epoch": 2.6054455282950704,
+      "grad_norm": 0.5473529100418091,
+      "learning_rate": 1.3150516909059052e-05,
+      "loss": 1.6345,
+      "step": 9916
+    },
+    {
+      "epoch": 2.605971031628732,
+      "grad_norm": 0.5622715950012207,
+      "learning_rate": 1.3132994568074294e-05,
+      "loss": 1.5891,
+      "step": 9918
+    },
+    {
+      "epoch": 2.6064965349623934,
+      "grad_norm": 0.6009332537651062,
+      "learning_rate": 1.311547222708954e-05,
+      "loss": 1.6095,
+      "step": 9920
+    },
+    {
+      "epoch": 2.6070220382960554,
+      "grad_norm": 0.746606707572937,
+      "learning_rate": 1.3097949886104784e-05,
+      "loss": 1.6328,
+      "step": 9922
+    },
+    {
+      "epoch": 2.6075475416297174,
+      "grad_norm": 0.6133710741996765,
+      "learning_rate": 1.3080427545120027e-05,
+      "loss": 1.6085,
+      "step": 9924
+    },
+    {
+      "epoch": 2.608073044963379,
+      "grad_norm": 0.6758058667182922,
+      "learning_rate": 1.3062905204135273e-05,
+      "loss": 1.62,
+      "step": 9926
+    },
+    {
+      "epoch": 2.608598548297041,
+      "grad_norm": 0.5908955335617065,
+      "learning_rate": 1.3045382863150518e-05,
+      "loss": 1.5991,
+      "step": 9928
+    },
+    {
+      "epoch": 2.6091240516307024,
+      "grad_norm": 0.5754519104957581,
+      "learning_rate": 1.3027860522165761e-05,
+      "loss": 1.6316,
+      "step": 9930
+    },
+    {
+      "epoch": 2.6096495549643643,
+      "grad_norm": 0.5225171446800232,
+      "learning_rate": 1.3010338181181007e-05,
+      "loss": 1.5902,
+      "step": 9932
+    },
+    {
+      "epoch": 2.610175058298026,
+      "grad_norm": 0.6654541492462158,
+      "learning_rate": 1.2992815840196251e-05,
+      "loss": 1.588,
+      "step": 9934
+    },
+    {
+      "epoch": 2.610700561631688,
+      "grad_norm": 0.592370331287384,
+      "learning_rate": 1.2975293499211497e-05,
+      "loss": 1.6396,
+      "step": 9936
+    },
+    {
+      "epoch": 2.61122606496535,
+      "grad_norm": 0.6326411366462708,
+      "learning_rate": 1.295777115822674e-05,
+      "loss": 1.6201,
+      "step": 9938
+    },
+    {
+      "epoch": 2.6117515682990113,
+      "grad_norm": 0.6815057396888733,
+      "learning_rate": 1.2940248817241985e-05,
+      "loss": 1.6337,
+      "step": 9940
+    },
+    {
+      "epoch": 2.6122770716326733,
+      "grad_norm": 0.6094640493392944,
+      "learning_rate": 1.292272647625723e-05,
+      "loss": 1.6009,
+      "step": 9942
+    },
+    {
+      "epoch": 2.612802574966335,
+      "grad_norm": 0.6490272283554077,
+      "learning_rate": 1.2905204135272472e-05,
+      "loss": 1.6249,
+      "step": 9944
+    },
+    {
+      "epoch": 2.613328078299997,
+      "grad_norm": 0.6855607032775879,
+      "learning_rate": 1.2887681794287718e-05,
+      "loss": 1.5982,
+      "step": 9946
+    },
+    {
+      "epoch": 2.6138535816336583,
+      "grad_norm": 0.6429476141929626,
+      "learning_rate": 1.2870159453302964e-05,
+      "loss": 1.6119,
+      "step": 9948
+    },
+    {
+      "epoch": 2.6143790849673203,
+      "grad_norm": 0.5955906510353088,
+      "learning_rate": 1.2852637112318206e-05,
+      "loss": 1.604,
+      "step": 9950
+    },
+    {
+      "epoch": 2.6149045883009823,
+      "grad_norm": 0.7396878600120544,
+      "learning_rate": 1.283511477133345e-05,
+      "loss": 1.6084,
+      "step": 9952
+    },
+    {
+      "epoch": 2.615430091634644,
+      "grad_norm": 0.631075382232666,
+      "learning_rate": 1.2817592430348696e-05,
+      "loss": 1.579,
+      "step": 9954
+    },
+    {
+      "epoch": 2.6159555949683053,
+      "grad_norm": 0.5280560255050659,
+      "learning_rate": 1.2800070089363939e-05,
+      "loss": 1.5952,
+      "step": 9956
+    },
+    {
+      "epoch": 2.6164810983019673,
+      "grad_norm": 0.5811305046081543,
+      "learning_rate": 1.2782547748379184e-05,
+      "loss": 1.5981,
+      "step": 9958
+    },
+    {
+      "epoch": 2.6170066016356293,
+      "grad_norm": 0.616111159324646,
+      "learning_rate": 1.2765025407394429e-05,
+      "loss": 1.6146,
+      "step": 9960
+    },
+    {
+      "epoch": 2.617532104969291,
+      "grad_norm": 0.6001870036125183,
+      "learning_rate": 1.2747503066409671e-05,
+      "loss": 1.612,
+      "step": 9962
+    },
+    {
+      "epoch": 2.6180576083029528,
+      "grad_norm": 0.5698304176330566,
+      "learning_rate": 1.2729980725424917e-05,
+      "loss": 1.5854,
+      "step": 9964
+    },
+    {
+      "epoch": 2.6185831116366143,
+      "grad_norm": 0.5722531676292419,
+      "learning_rate": 1.2712458384440163e-05,
+      "loss": 1.595,
+      "step": 9966
+    },
+    {
+      "epoch": 2.6191086149702762,
+      "grad_norm": 0.556814968585968,
+      "learning_rate": 1.2694936043455405e-05,
+      "loss": 1.6079,
+      "step": 9968
+    },
+    {
+      "epoch": 2.6196341183039378,
+      "grad_norm": 0.7477023601531982,
+      "learning_rate": 1.267741370247065e-05,
+      "loss": 1.6101,
+      "step": 9970
+    },
+    {
+      "epoch": 2.6201596216375997,
+      "grad_norm": 0.5011134743690491,
+      "learning_rate": 1.2659891361485895e-05,
+      "loss": 1.5684,
+      "step": 9972
+    },
+    {
+      "epoch": 2.6206851249712617,
+      "grad_norm": 0.5759318470954895,
+      "learning_rate": 1.2642369020501141e-05,
+      "loss": 1.5726,
+      "step": 9974
+    },
+    {
+      "epoch": 2.6212106283049232,
+      "grad_norm": 0.5837996006011963,
+      "learning_rate": 1.2624846679516384e-05,
+      "loss": 1.5975,
+      "step": 9976
+    },
+    {
+      "epoch": 2.621736131638585,
+      "grad_norm": 0.6139369606971741,
+      "learning_rate": 1.2607324338531628e-05,
+      "loss": 1.604,
+      "step": 9978
+    },
+    {
+      "epoch": 2.6222616349722467,
+      "grad_norm": 0.59381103515625,
+      "learning_rate": 1.2589801997546874e-05,
+      "loss": 1.631,
+      "step": 9980
+    },
+    {
+      "epoch": 2.6227871383059087,
+      "grad_norm": 0.59110027551651,
+      "learning_rate": 1.2572279656562116e-05,
+      "loss": 1.6342,
+      "step": 9982
+    },
+    {
+      "epoch": 2.6233126416395702,
+      "grad_norm": 0.6388522386550903,
+      "learning_rate": 1.2554757315577362e-05,
+      "loss": 1.602,
+      "step": 9984
+    },
+    {
+      "epoch": 2.623838144973232,
+      "grad_norm": 0.5485536456108093,
+      "learning_rate": 1.2537234974592606e-05,
+      "loss": 1.5904,
+      "step": 9986
+    },
+    {
+      "epoch": 2.624363648306894,
+      "grad_norm": 0.6147143840789795,
+      "learning_rate": 1.2519712633607849e-05,
+      "loss": 1.6193,
+      "step": 9988
+    },
+    {
+      "epoch": 2.6248891516405557,
+      "grad_norm": 0.6774529814720154,
+      "learning_rate": 1.2502190292623095e-05,
+      "loss": 1.5894,
+      "step": 9990
+    },
+    {
+      "epoch": 2.625414654974217,
+      "grad_norm": 0.5123162269592285,
+      "learning_rate": 1.2484667951638339e-05,
+      "loss": 1.5871,
+      "step": 9992
+    },
+    {
+      "epoch": 2.625940158307879,
+      "grad_norm": 0.5840069651603699,
+      "learning_rate": 1.2467145610653585e-05,
+      "loss": 1.6093,
+      "step": 9994
+    },
+    {
+      "epoch": 2.626465661641541,
+      "grad_norm": 0.6273425817489624,
+      "learning_rate": 1.2449623269668829e-05,
+      "loss": 1.5992,
+      "step": 9996
+    },
+    {
+      "epoch": 2.6269911649752027,
+      "grad_norm": 0.5548253655433655,
+      "learning_rate": 1.2432100928684073e-05,
+      "loss": 1.5966,
+      "step": 9998
+    },
+    {
+      "epoch": 2.6275166683088647,
+      "grad_norm": 0.6084941625595093,
+      "learning_rate": 1.2414578587699317e-05,
+      "loss": 1.6002,
+      "step": 10000
+    },
+    {
+      "epoch": 2.6275166683088647,
+      "eval_loss": 1.6519938707351685,
+      "eval_runtime": 487.1623,
+      "eval_samples_per_second": 249.997,
+      "eval_steps_per_second": 31.25,
+      "step": 10000
+    },
+    {
+      "epoch": 2.6280421716425266,
+      "grad_norm": 0.5661596655845642,
+      "learning_rate": 1.2397056246714561e-05,
+      "loss": 1.6081,
+      "step": 10002
+    },
+    {
+      "epoch": 2.628567674976188,
+      "grad_norm": 0.6092149615287781,
+      "learning_rate": 1.2379533905729807e-05,
+      "loss": 1.5984,
+      "step": 10004
+    },
+    {
+      "epoch": 2.6290931783098497,
+      "grad_norm": 0.6022893786430359,
+      "learning_rate": 1.236201156474505e-05,
+      "loss": 1.594,
+      "step": 10006
+    },
+    {
+      "epoch": 2.6296186816435116,
+      "grad_norm": 0.6382527351379395,
+      "learning_rate": 1.2344489223760294e-05,
+      "loss": 1.6076,
+      "step": 10008
+    },
+    {
+      "epoch": 2.6301441849771736,
+      "grad_norm": 0.6449958682060242,
+      "learning_rate": 1.232696688277554e-05,
+      "loss": 1.6056,
+      "step": 10010
+    },
+    {
+      "epoch": 2.630669688310835,
+      "grad_norm": 0.6634608507156372,
+      "learning_rate": 1.2309444541790784e-05,
+      "loss": 1.5862,
+      "step": 10012
+    },
+    {
+      "epoch": 2.631195191644497,
+      "grad_norm": 0.5740031003952026,
+      "learning_rate": 1.2291922200806028e-05,
+      "loss": 1.5988,
+      "step": 10014
+    },
+    {
+      "epoch": 2.6317206949781586,
+      "grad_norm": 0.7303963899612427,
+      "learning_rate": 1.2274399859821272e-05,
+      "loss": 1.6342,
+      "step": 10016
+    },
+    {
+      "epoch": 2.6322461983118206,
+      "grad_norm": 0.5871624946594238,
+      "learning_rate": 1.2256877518836516e-05,
+      "loss": 1.5836,
+      "step": 10018
+    },
+    {
+      "epoch": 2.632771701645482,
+      "grad_norm": 0.5941208600997925,
+      "learning_rate": 1.2239355177851762e-05,
+      "loss": 1.5655,
+      "step": 10020
+    },
+    {
+      "epoch": 2.633297204979144,
+      "grad_norm": 0.5943103432655334,
+      "learning_rate": 1.2221832836867007e-05,
+      "loss": 1.6448,
+      "step": 10022
+    },
+    {
+      "epoch": 2.633822708312806,
+      "grad_norm": 0.6347323060035706,
+      "learning_rate": 1.220431049588225e-05,
+      "loss": 1.621,
+      "step": 10024
+    },
+    {
+      "epoch": 2.6343482116464676,
+      "grad_norm": Infinity,
+      "learning_rate": 1.2195549325389873e-05,
+      "loss": 1.6178,
+      "step": 10026
+    },
+    {
+      "epoch": 2.6348737149801296,
+      "grad_norm": 0.6692628264427185,
+      "learning_rate": 1.2178026984405117e-05,
+      "loss": 1.6063,
+      "step": 10028
+    },
+    {
+      "epoch": 2.635399218313791,
+      "grad_norm": 0.5815702080726624,
+      "learning_rate": 1.2160504643420361e-05,
+      "loss": 1.5884,
+      "step": 10030
+    },
+    {
+      "epoch": 2.635924721647453,
+      "grad_norm": 0.5981874465942383,
+      "learning_rate": 1.2142982302435605e-05,
+      "loss": 1.614,
+      "step": 10032
+    },
+    {
+      "epoch": 2.6364502249811146,
+      "grad_norm": 0.5921944975852966,
+      "learning_rate": 1.2125459961450851e-05,
+      "loss": 1.6172,
+      "step": 10034
+    },
+    {
+      "epoch": 2.6369757283147766,
+      "grad_norm": 0.5707213878631592,
+      "learning_rate": 1.2107937620466095e-05,
+      "loss": 1.6205,
+      "step": 10036
+    },
+    {
+      "epoch": 2.6375012316484385,
+      "grad_norm": 0.5834996700286865,
+      "learning_rate": 1.209041527948134e-05,
+      "loss": 1.598,
+      "step": 10038
+    },
+    {
+      "epoch": 2.6380267349821,
+      "grad_norm": 0.568755567073822,
+      "learning_rate": 1.2072892938496584e-05,
+      "loss": 1.5819,
+      "step": 10040
+    },
+    {
+      "epoch": 2.6385522383157616,
+      "grad_norm": 0.6087754368782043,
+      "learning_rate": 1.2055370597511828e-05,
+      "loss": 1.6223,
+      "step": 10042
+    },
+    {
+      "epoch": 2.6390777416494235,
+      "grad_norm": 0.5805062651634216,
+      "learning_rate": 1.2037848256527072e-05,
+      "loss": 1.6055,
+      "step": 10044
+    },
+    {
+      "epoch": 2.6396032449830855,
+      "grad_norm": 0.5679923892021179,
+      "learning_rate": 1.2020325915542318e-05,
+      "loss": 1.6388,
+      "step": 10046
+    },
+    {
+      "epoch": 2.640128748316747,
+      "grad_norm": 0.7020934820175171,
+      "learning_rate": 1.200280357455756e-05,
+      "loss": 1.6076,
+      "step": 10048
+    },
+    {
+      "epoch": 2.640654251650409,
+      "grad_norm": 0.5716947913169861,
+      "learning_rate": 1.1985281233572806e-05,
+      "loss": 1.6035,
+      "step": 10050
+    },
+    {
+      "epoch": 2.6411797549840705,
+      "grad_norm": 0.6460137963294983,
+      "learning_rate": 1.196775889258805e-05,
+      "loss": 1.5981,
+      "step": 10052
+    },
+    {
+      "epoch": 2.6417052583177325,
+      "grad_norm": 0.5903118848800659,
+      "learning_rate": 1.1950236551603295e-05,
+      "loss": 1.5844,
+      "step": 10054
+    },
+    {
+      "epoch": 2.642230761651394,
+      "grad_norm": 0.5728901028633118,
+      "learning_rate": 1.193271421061854e-05,
+      "loss": 1.5811,
+      "step": 10056
+    },
+    {
+      "epoch": 2.642756264985056,
+      "grad_norm": 0.6795145273208618,
+      "learning_rate": 1.1915191869633783e-05,
+      "loss": 1.6334,
+      "step": 10058
+    },
+    {
+      "epoch": 2.643281768318718,
+      "grad_norm": 0.552899956703186,
+      "learning_rate": 1.1897669528649027e-05,
+      "loss": 1.5956,
+      "step": 10060
+    },
+    {
+      "epoch": 2.6438072716523795,
+      "grad_norm": 0.5750575065612793,
+      "learning_rate": 1.1880147187664273e-05,
+      "loss": 1.5974,
+      "step": 10062
+    },
+    {
+      "epoch": 2.6443327749860415,
+      "grad_norm": 0.6233214735984802,
+      "learning_rate": 1.1862624846679517e-05,
+      "loss": 1.6177,
+      "step": 10064
+    },
+    {
+      "epoch": 2.644858278319703,
+      "grad_norm": 0.7260066270828247,
+      "learning_rate": 1.1845102505694761e-05,
+      "loss": 1.6024,
+      "step": 10066
+    },
+    {
+      "epoch": 2.645383781653365,
+      "grad_norm": 0.5926297903060913,
+      "learning_rate": 1.1827580164710006e-05,
+      "loss": 1.6003,
+      "step": 10068
+    },
+    {
+      "epoch": 2.6459092849870265,
+      "grad_norm": 0.565710723400116,
+      "learning_rate": 1.181005782372525e-05,
+      "loss": 1.5879,
+      "step": 10070
+    },
+    {
+      "epoch": 2.6464347883206885,
+      "grad_norm": 0.5911100506782532,
+      "learning_rate": 1.1792535482740496e-05,
+      "loss": 1.5862,
+      "step": 10072
+    },
+    {
+      "epoch": 2.6469602916543504,
+      "grad_norm": 0.6018568277359009,
+      "learning_rate": 1.177501314175574e-05,
+      "loss": 1.6274,
+      "step": 10074
+    },
+    {
+      "epoch": 2.647485794988012,
+      "grad_norm": 0.6663408875465393,
+      "learning_rate": 1.1757490800770982e-05,
+      "loss": 1.6331,
+      "step": 10076
+    },
+    {
+      "epoch": 2.6480112983216735,
+      "grad_norm": 0.6028339266777039,
+      "learning_rate": 1.1739968459786228e-05,
+      "loss": 1.6208,
+      "step": 10078
+    },
+    {
+      "epoch": 2.6485368016553354,
+      "grad_norm": 0.5986884236335754,
+      "learning_rate": 1.1722446118801472e-05,
+      "loss": 1.5974,
+      "step": 10080
+    },
+    {
+      "epoch": 2.6490623049889974,
+      "grad_norm": 0.7179824113845825,
+      "learning_rate": 1.1704923777816718e-05,
+      "loss": 1.5959,
+      "step": 10082
+    },
+    {
+      "epoch": 2.649587808322659,
+      "grad_norm": 0.6211957931518555,
+      "learning_rate": 1.168740143683196e-05,
+      "loss": 1.6052,
+      "step": 10084
+    },
+    {
+      "epoch": 2.650113311656321,
+      "grad_norm": 0.5563115477561951,
+      "learning_rate": 1.1669879095847205e-05,
+      "loss": 1.6047,
+      "step": 10086
+    },
+    {
+      "epoch": 2.6506388149899824,
+      "grad_norm": 0.6166182160377502,
+      "learning_rate": 1.165235675486245e-05,
+      "loss": 1.591,
+      "step": 10088
+    },
+    {
+      "epoch": 2.6511643183236444,
+      "grad_norm": 0.5224390625953674,
+      "learning_rate": 1.1634834413877695e-05,
+      "loss": 1.6147,
+      "step": 10090
+    },
+    {
+      "epoch": 2.651689821657306,
+      "grad_norm": 0.6834142208099365,
+      "learning_rate": 1.1617312072892939e-05,
+      "loss": 1.6194,
+      "step": 10092
+    },
+    {
+      "epoch": 2.652215324990968,
+      "grad_norm": 0.5866041779518127,
+      "learning_rate": 1.1599789731908183e-05,
+      "loss": 1.6286,
+      "step": 10094
+    },
+    {
+      "epoch": 2.65274082832463,
+      "grad_norm": 0.6454026699066162,
+      "learning_rate": 1.1582267390923427e-05,
+      "loss": 1.6181,
+      "step": 10096
+    },
+    {
+      "epoch": 2.6532663316582914,
+      "grad_norm": 0.5898521542549133,
+      "learning_rate": 1.1564745049938672e-05,
+      "loss": 1.6157,
+      "step": 10098
+    },
+    {
+      "epoch": 2.6537918349919534,
+      "grad_norm": 0.5379465222358704,
+      "learning_rate": 1.1547222708953917e-05,
+      "loss": 1.6015,
+      "step": 10100
+    },
+    {
+      "epoch": 2.654317338325615,
+      "grad_norm": 0.624577522277832,
+      "learning_rate": 1.1529700367969162e-05,
+      "loss": 1.6161,
+      "step": 10102
+    },
+    {
+      "epoch": 2.654842841659277,
+      "grad_norm": 0.5811554193496704,
+      "learning_rate": 1.1512178026984406e-05,
+      "loss": 1.598,
+      "step": 10104
+    },
+    {
+      "epoch": 2.6553683449929384,
+      "grad_norm": 0.5827350616455078,
+      "learning_rate": 1.149465568599965e-05,
+      "loss": 1.625,
+      "step": 10106
+    },
+    {
+      "epoch": 2.6558938483266004,
+      "grad_norm": 0.6221204996109009,
+      "learning_rate": 1.1477133345014894e-05,
+      "loss": 1.6011,
+      "step": 10108
+    },
+    {
+      "epoch": 2.6564193516602623,
+      "grad_norm": 0.6008239984512329,
+      "learning_rate": 1.145961100403014e-05,
+      "loss": 1.6048,
+      "step": 10110
+    },
+    {
+      "epoch": 2.656944854993924,
+      "grad_norm": 0.5547776222229004,
+      "learning_rate": 1.1442088663045382e-05,
+      "loss": 1.6149,
+      "step": 10112
+    },
+    {
+      "epoch": 2.6574703583275854,
+      "grad_norm": 0.5586524605751038,
+      "learning_rate": 1.1424566322060627e-05,
+      "loss": 1.576,
+      "step": 10114
+    },
+    {
+      "epoch": 2.6579958616612473,
+      "grad_norm": 0.6657798886299133,
+      "learning_rate": 1.1407043981075873e-05,
+      "loss": 1.5939,
+      "step": 10116
+    },
+    {
+      "epoch": 2.6585213649949093,
+      "grad_norm": 0.6012902855873108,
+      "learning_rate": 1.1389521640091117e-05,
+      "loss": 1.6089,
+      "step": 10118
+    },
+    {
+      "epoch": 2.659046868328571,
+      "grad_norm": 0.6174466013908386,
+      "learning_rate": 1.1371999299106363e-05,
+      "loss": 1.606,
+      "step": 10120
+    },
+    {
+      "epoch": 2.659572371662233,
+      "grad_norm": 0.5988878011703491,
+      "learning_rate": 1.1354476958121605e-05,
+      "loss": 1.6162,
+      "step": 10122
+    },
+    {
+      "epoch": 2.6600978749958943,
+      "grad_norm": 0.6152504086494446,
+      "learning_rate": 1.133695461713685e-05,
+      "loss": 1.6265,
+      "step": 10124
+    },
+    {
+      "epoch": 2.6606233783295563,
+      "grad_norm": 0.5866356492042542,
+      "learning_rate": 1.1319432276152095e-05,
+      "loss": 1.5945,
+      "step": 10126
+    },
+    {
+      "epoch": 2.661148881663218,
+      "grad_norm": 0.6397704482078552,
+      "learning_rate": 1.130190993516734e-05,
+      "loss": 1.6141,
+      "step": 10128
+    },
+    {
+      "epoch": 2.66167438499688,
+      "grad_norm": 0.5501188039779663,
+      "learning_rate": 1.1284387594182583e-05,
+      "loss": 1.6161,
+      "step": 10130
+    },
+    {
+      "epoch": 2.6621998883305418,
+      "grad_norm": 0.5947812795639038,
+      "learning_rate": 1.1266865253197828e-05,
+      "loss": 1.6168,
+      "step": 10132
+    },
+    {
+      "epoch": 2.6627253916642033,
+      "grad_norm": 0.571863055229187,
+      "learning_rate": 1.1249342912213072e-05,
+      "loss": 1.6012,
+      "step": 10134
+    },
+    {
+      "epoch": 2.6632508949978653,
+      "grad_norm": 0.5414539575576782,
+      "learning_rate": 1.1231820571228318e-05,
+      "loss": 1.5672,
+      "step": 10136
+    },
+    {
+      "epoch": 2.663776398331527,
+      "grad_norm": 0.5631955862045288,
+      "learning_rate": 1.1214298230243562e-05,
+      "loss": 1.6291,
+      "step": 10138
+    },
+    {
+      "epoch": 2.6643019016651888,
+      "grad_norm": 0.6036393642425537,
+      "learning_rate": 1.1196775889258804e-05,
+      "loss": 1.6131,
+      "step": 10140
+    },
+    {
+      "epoch": 2.6648274049988503,
+      "grad_norm": 0.5736912488937378,
+      "learning_rate": 1.117925354827405e-05,
+      "loss": 1.595,
+      "step": 10142
+    },
+    {
+      "epoch": 2.6653529083325123,
+      "grad_norm": 0.6014968156814575,
+      "learning_rate": 1.1161731207289294e-05,
+      "loss": 1.6042,
+      "step": 10144
+    },
+    {
+      "epoch": 2.6658784116661742,
+      "grad_norm": 0.564576268196106,
+      "learning_rate": 1.1144208866304539e-05,
+      "loss": 1.6199,
+      "step": 10146
+    },
+    {
+      "epoch": 2.6664039149998358,
+      "grad_norm": 0.5501058101654053,
+      "learning_rate": 1.1126686525319783e-05,
+      "loss": 1.6063,
+      "step": 10148
+    },
+    {
+      "epoch": 2.6669294183334973,
+      "grad_norm": 0.6601042151451111,
+      "learning_rate": 1.1109164184335027e-05,
+      "loss": 1.5838,
+      "step": 10150
+    },
+    {
+      "epoch": 2.6674549216671593,
+      "grad_norm": 0.6125220656394958,
+      "learning_rate": 1.1091641843350273e-05,
+      "loss": 1.6151,
+      "step": 10152
+    },
+    {
+      "epoch": 2.667980425000821,
+      "grad_norm": 0.6514031291007996,
+      "learning_rate": 1.1074119502365517e-05,
+      "loss": 1.5978,
+      "step": 10154
+    },
+    {
+      "epoch": 2.6685059283344827,
+      "grad_norm": 0.5755125284194946,
+      "learning_rate": 1.1056597161380761e-05,
+      "loss": 1.5971,
+      "step": 10156
+    },
+    {
+      "epoch": 2.6690314316681447,
+      "grad_norm": 0.6414878368377686,
+      "learning_rate": 1.1039074820396005e-05,
+      "loss": 1.6212,
+      "step": 10158
+    },
+    {
+      "epoch": 2.6695569350018067,
+      "grad_norm": 0.6092830300331116,
+      "learning_rate": 1.102155247941125e-05,
+      "loss": 1.6006,
+      "step": 10160
+    },
+    {
+      "epoch": 2.670082438335468,
+      "grad_norm": 0.6845732927322388,
+      "learning_rate": 1.1004030138426494e-05,
+      "loss": 1.6473,
+      "step": 10162
+    },
+    {
+      "epoch": 2.6706079416691297,
+      "grad_norm": 0.5648114681243896,
+      "learning_rate": 1.098650779744174e-05,
+      "loss": 1.5988,
+      "step": 10164
+    },
+    {
+      "epoch": 2.6711334450027917,
+      "grad_norm": 0.5797778367996216,
+      "learning_rate": 1.0968985456456984e-05,
+      "loss": 1.6057,
+      "step": 10166
+    },
+    {
+      "epoch": 2.6716589483364537,
+      "grad_norm": 0.6136969327926636,
+      "learning_rate": 1.0951463115472228e-05,
+      "loss": 1.5615,
+      "step": 10168
+    },
+    {
+      "epoch": 2.672184451670115,
+      "grad_norm": 0.5537448525428772,
+      "learning_rate": 1.0933940774487472e-05,
+      "loss": 1.5739,
+      "step": 10170
+    },
+    {
+      "epoch": 2.672709955003777,
+      "grad_norm": 0.6372264623641968,
+      "learning_rate": 1.0916418433502716e-05,
+      "loss": 1.5985,
+      "step": 10172
+    },
+    {
+      "epoch": 2.6732354583374387,
+      "grad_norm": 0.5945001840591431,
+      "learning_rate": 1.0898896092517962e-05,
+      "loss": 1.6051,
+      "step": 10174
+    },
+    {
+      "epoch": 2.6737609616711007,
+      "grad_norm": 0.5689924359321594,
+      "learning_rate": 1.0881373751533205e-05,
+      "loss": 1.6083,
+      "step": 10176
+    },
+    {
+      "epoch": 2.674286465004762,
+      "grad_norm": 0.596948504447937,
+      "learning_rate": 1.0863851410548449e-05,
+      "loss": 1.6086,
+      "step": 10178
+    },
+    {
+      "epoch": 2.674811968338424,
+      "grad_norm": 0.581723153591156,
+      "learning_rate": 1.0846329069563695e-05,
+      "loss": 1.6127,
+      "step": 10180
+    },
+    {
+      "epoch": 2.675337471672086,
+      "grad_norm": 0.5663895010948181,
+      "learning_rate": 1.0828806728578939e-05,
+      "loss": 1.571,
+      "step": 10182
+    },
+    {
+      "epoch": 2.6758629750057477,
+      "grad_norm": 0.5802646279335022,
+      "learning_rate": 1.0811284387594183e-05,
+      "loss": 1.6068,
+      "step": 10184
+    },
+    {
+      "epoch": 2.6763884783394096,
+      "grad_norm": 0.615695595741272,
+      "learning_rate": 1.0793762046609427e-05,
+      "loss": 1.6147,
+      "step": 10186
+    },
+    {
+      "epoch": 2.676913981673071,
+      "grad_norm": 0.6105237007141113,
+      "learning_rate": 1.0776239705624671e-05,
+      "loss": 1.607,
+      "step": 10188
+    },
+    {
+      "epoch": 2.677439485006733,
+      "grad_norm": 0.5385169982910156,
+      "learning_rate": 1.0758717364639917e-05,
+      "loss": 1.6079,
+      "step": 10190
+    },
+    {
+      "epoch": 2.6779649883403946,
+      "grad_norm": 0.7373268008232117,
+      "learning_rate": 1.0741195023655161e-05,
+      "loss": 1.6043,
+      "step": 10192
+    },
+    {
+      "epoch": 2.6784904916740566,
+      "grad_norm": 0.5916065573692322,
+      "learning_rate": 1.0723672682670406e-05,
+      "loss": 1.6349,
+      "step": 10194
+    },
+    {
+      "epoch": 2.6790159950077186,
+      "grad_norm": 0.7171288132667542,
+      "learning_rate": 1.070615034168565e-05,
+      "loss": 1.6027,
+      "step": 10196
+    },
+    {
+      "epoch": 2.67954149834138,
+      "grad_norm": 0.5862658619880676,
+      "learning_rate": 1.0688628000700894e-05,
+      "loss": 1.5782,
+      "step": 10198
+    },
+    {
+      "epoch": 2.6800670016750416,
+      "grad_norm": 0.5845574736595154,
+      "learning_rate": 1.0671105659716138e-05,
+      "loss": 1.6043,
+      "step": 10200
+    },
+    {
+      "epoch": 2.6805925050087036,
+      "grad_norm": 0.6461187601089478,
+      "learning_rate": 1.0653583318731384e-05,
+      "loss": 1.6035,
+      "step": 10202
+    },
+    {
+      "epoch": 2.6811180083423656,
+      "grad_norm": 0.5671385526657104,
+      "learning_rate": 1.0636060977746626e-05,
+      "loss": 1.5873,
+      "step": 10204
+    },
+    {
+      "epoch": 2.681643511676027,
+      "grad_norm": 0.614528238773346,
+      "learning_rate": 1.0618538636761872e-05,
+      "loss": 1.5885,
+      "step": 10206
+    },
+    {
+      "epoch": 2.682169015009689,
+      "grad_norm": 0.6290176510810852,
+      "learning_rate": 1.0601016295777116e-05,
+      "loss": 1.6293,
+      "step": 10208
+    },
+    {
+      "epoch": 2.6826945183433506,
+      "grad_norm": 0.7489297986030579,
+      "learning_rate": 1.058349395479236e-05,
+      "loss": 1.6248,
+      "step": 10210
+    },
+    {
+      "epoch": 2.6832200216770126,
+      "grad_norm": 0.5606198310852051,
+      "learning_rate": 1.0565971613807605e-05,
+      "loss": 1.5935,
+      "step": 10212
+    },
+    {
+      "epoch": 2.683745525010674,
+      "grad_norm": 0.5877952575683594,
+      "learning_rate": 1.0548449272822849e-05,
+      "loss": 1.5945,
+      "step": 10214
+    },
+    {
+      "epoch": 2.684271028344336,
+      "grad_norm": 0.6569766998291016,
+      "learning_rate": 1.0530926931838093e-05,
+      "loss": 1.5946,
+      "step": 10216
+    },
+    {
+      "epoch": 2.684796531677998,
+      "grad_norm": 0.5706822276115417,
+      "learning_rate": 1.0513404590853339e-05,
+      "loss": 1.6302,
+      "step": 10218
+    },
+    {
+      "epoch": 2.6853220350116596,
+      "grad_norm": 0.6443889141082764,
+      "learning_rate": 1.0495882249868583e-05,
+      "loss": 1.6055,
+      "step": 10220
+    },
+    {
+      "epoch": 2.6858475383453215,
+      "grad_norm": 0.6561224460601807,
+      "learning_rate": 1.0478359908883827e-05,
+      "loss": 1.5979,
+      "step": 10222
+    },
+    {
+      "epoch": 2.686373041678983,
+      "grad_norm": 0.6297589540481567,
+      "learning_rate": 1.0460837567899072e-05,
+      "loss": 1.621,
+      "step": 10224
+    },
+    {
+      "epoch": 2.686898545012645,
+      "grad_norm": 0.5593584775924683,
+      "learning_rate": 1.0443315226914316e-05,
+      "loss": 1.5798,
+      "step": 10226
+    },
+    {
+      "epoch": 2.6874240483463065,
+      "grad_norm": 0.5252684354782104,
+      "learning_rate": 1.0425792885929562e-05,
+      "loss": 1.5672,
+      "step": 10228
+    },
+    {
+      "epoch": 2.6879495516799685,
+      "grad_norm": 0.7257969379425049,
+      "learning_rate": 1.0408270544944806e-05,
+      "loss": 1.6178,
+      "step": 10230
+    },
+    {
+      "epoch": 2.6884750550136305,
+      "grad_norm": 0.5459820628166199,
+      "learning_rate": 1.0390748203960048e-05,
+      "loss": 1.589,
+      "step": 10232
+    },
+    {
+      "epoch": 2.689000558347292,
+      "grad_norm": 0.6055821180343628,
+      "learning_rate": 1.0373225862975294e-05,
+      "loss": 1.5899,
+      "step": 10234
+    },
+    {
+      "epoch": 2.6895260616809535,
+      "grad_norm": 0.7138866782188416,
+      "learning_rate": 1.0355703521990538e-05,
+      "loss": 1.6004,
+      "step": 10236
+    },
+    {
+      "epoch": 2.6900515650146155,
+      "grad_norm": 0.5606485605239868,
+      "learning_rate": 1.0338181181005784e-05,
+      "loss": 1.5703,
+      "step": 10238
+    },
+    {
+      "epoch": 2.6905770683482775,
+      "grad_norm": 0.6380048990249634,
+      "learning_rate": 1.0320658840021027e-05,
+      "loss": 1.6236,
+      "step": 10240
+    },
+    {
+      "epoch": 2.691102571681939,
+      "grad_norm": 0.59669029712677,
+      "learning_rate": 1.030313649903627e-05,
+      "loss": 1.5963,
+      "step": 10242
+    },
+    {
+      "epoch": 2.691628075015601,
+      "grad_norm": 0.6684574484825134,
+      "learning_rate": 1.0285614158051517e-05,
+      "loss": 1.6199,
+      "step": 10244
+    },
+    {
+      "epoch": 2.6921535783492625,
+      "grad_norm": 0.5515660047531128,
+      "learning_rate": 1.0268091817066761e-05,
+      "loss": 1.6143,
+      "step": 10246
+    },
+    {
+      "epoch": 2.6926790816829245,
+      "grad_norm": 0.6562058329582214,
+      "learning_rate": 1.0250569476082005e-05,
+      "loss": 1.6112,
+      "step": 10248
+    },
+    {
+      "epoch": 2.693204585016586,
+      "grad_norm": 0.5547401309013367,
+      "learning_rate": 1.023304713509725e-05,
+      "loss": 1.6124,
+      "step": 10250
+    },
+    {
+      "epoch": 2.693730088350248,
+      "grad_norm": 0.6307016611099243,
+      "learning_rate": 1.0215524794112493e-05,
+      "loss": 1.6256,
+      "step": 10252
+    },
+    {
+      "epoch": 2.69425559168391,
+      "grad_norm": 0.5722636580467224,
+      "learning_rate": 1.0198002453127738e-05,
+      "loss": 1.5713,
+      "step": 10254
+    },
+    {
+      "epoch": 2.6947810950175715,
+      "grad_norm": 0.587989330291748,
+      "learning_rate": 1.0180480112142983e-05,
+      "loss": 1.6191,
+      "step": 10256
+    },
+    {
+      "epoch": 2.6953065983512334,
+      "grad_norm": 0.7785957455635071,
+      "learning_rate": 1.0162957771158228e-05,
+      "loss": 1.6129,
+      "step": 10258
+    },
+    {
+      "epoch": 2.695832101684895,
+      "grad_norm": 0.5290049910545349,
+      "learning_rate": 1.0145435430173472e-05,
+      "loss": 1.5561,
+      "step": 10260
+    },
+    {
+      "epoch": 2.696357605018557,
+      "grad_norm": 0.6306317448616028,
+      "learning_rate": 1.0127913089188716e-05,
+      "loss": 1.5872,
+      "step": 10262
+    },
+    {
+      "epoch": 2.6968831083522184,
+      "grad_norm": 0.5847423672676086,
+      "learning_rate": 1.011039074820396e-05,
+      "loss": 1.5783,
+      "step": 10264
+    },
+    {
+      "epoch": 2.6974086116858804,
+      "grad_norm": 0.5815863609313965,
+      "learning_rate": 1.0092868407219206e-05,
+      "loss": 1.6138,
+      "step": 10266
+    },
+    {
+      "epoch": 2.6979341150195424,
+      "grad_norm": 0.6759339570999146,
+      "learning_rate": 1.0075346066234448e-05,
+      "loss": 1.6146,
+      "step": 10268
+    },
+    {
+      "epoch": 2.698459618353204,
+      "grad_norm": 0.7321836948394775,
+      "learning_rate": 1.0057823725249693e-05,
+      "loss": 1.6333,
+      "step": 10270
+    },
+    {
+      "epoch": 2.6989851216868654,
+      "grad_norm": 0.595303475856781,
+      "learning_rate": 1.0040301384264939e-05,
+      "loss": 1.5903,
+      "step": 10272
+    },
+    {
+      "epoch": 2.6995106250205274,
+      "grad_norm": 0.6422771215438843,
+      "learning_rate": 1.0022779043280183e-05,
+      "loss": 1.6063,
+      "step": 10274
+    },
+    {
+      "epoch": 2.7000361283541894,
+      "grad_norm": 0.6339772939682007,
+      "learning_rate": 1.0005256702295427e-05,
+      "loss": 1.6402,
+      "step": 10276
+    },
+    {
+      "epoch": 2.700561631687851,
+      "grad_norm": 0.6175129413604736,
+      "learning_rate": 9.987734361310671e-06,
+      "loss": 1.5768,
+      "step": 10278
+    },
+    {
+      "epoch": 2.701087135021513,
+      "grad_norm": 0.6692789793014526,
+      "learning_rate": 9.970212020325915e-06,
+      "loss": 1.6293,
+      "step": 10280
+    },
+    {
+      "epoch": 2.7016126383551744,
+      "grad_norm": 0.6754617094993591,
+      "learning_rate": 9.952689679341161e-06,
+      "loss": 1.6127,
+      "step": 10282
+    },
+    {
+      "epoch": 2.7021381416888364,
+      "grad_norm": 0.5878320336341858,
+      "learning_rate": 9.935167338356405e-06,
+      "loss": 1.6195,
+      "step": 10284
+    },
+    {
+      "epoch": 2.702663645022498,
+      "grad_norm": 0.9269865155220032,
+      "learning_rate": 9.91764499737165e-06,
+      "loss": 1.5903,
+      "step": 10286
+    },
+    {
+      "epoch": 2.70318914835616,
+      "grad_norm": 0.610824465751648,
+      "learning_rate": 9.900122656386894e-06,
+      "loss": 1.608,
+      "step": 10288
+    },
+    {
+      "epoch": 2.703714651689822,
+      "grad_norm": 0.587494432926178,
+      "learning_rate": 9.882600315402138e-06,
+      "loss": 1.547,
+      "step": 10290
+    },
+    {
+      "epoch": 2.7042401550234834,
+      "grad_norm": 0.6791782975196838,
+      "learning_rate": 9.865077974417384e-06,
+      "loss": 1.6039,
+      "step": 10292
+    },
+    {
+      "epoch": 2.7047656583571453,
+      "grad_norm": 0.5835421681404114,
+      "learning_rate": 9.847555633432628e-06,
+      "loss": 1.6378,
+      "step": 10294
+    },
+    {
+      "epoch": 2.705291161690807,
+      "grad_norm": 0.6221166253089905,
+      "learning_rate": 9.83003329244787e-06,
+      "loss": 1.5753,
+      "step": 10296
+    },
+    {
+      "epoch": 2.705816665024469,
+      "grad_norm": 0.5716819763183594,
+      "learning_rate": 9.812510951463116e-06,
+      "loss": 1.5876,
+      "step": 10298
+    },
+    {
+      "epoch": 2.7063421683581304,
+      "grad_norm": 0.725444495677948,
+      "learning_rate": 9.79498861047836e-06,
+      "loss": 1.6308,
+      "step": 10300
+    },
+    {
+      "epoch": 2.7068676716917923,
+      "grad_norm": 0.6209605932235718,
+      "learning_rate": 9.777466269493605e-06,
+      "loss": 1.5797,
+      "step": 10302
+    },
+    {
+      "epoch": 2.7073931750254543,
+      "grad_norm": 0.6639958620071411,
+      "learning_rate": 9.759943928508849e-06,
+      "loss": 1.6051,
+      "step": 10304
+    },
+    {
+      "epoch": 2.707918678359116,
+      "grad_norm": 0.6042879819869995,
+      "learning_rate": 9.742421587524093e-06,
+      "loss": 1.5985,
+      "step": 10306
+    },
+    {
+      "epoch": 2.7084441816927773,
+      "grad_norm": 0.5975037813186646,
+      "learning_rate": 9.724899246539339e-06,
+      "loss": 1.5945,
+      "step": 10308
+    },
+    {
+      "epoch": 2.7089696850264393,
+      "grad_norm": 0.5482746362686157,
+      "learning_rate": 9.707376905554583e-06,
+      "loss": 1.5645,
+      "step": 10310
+    },
+    {
+      "epoch": 2.7094951883601013,
+      "grad_norm": 0.6123608946800232,
+      "learning_rate": 9.689854564569827e-06,
+      "loss": 1.5862,
+      "step": 10312
+    },
+    {
+      "epoch": 2.710020691693763,
+      "grad_norm": 0.6144323348999023,
+      "learning_rate": 9.672332223585071e-06,
+      "loss": 1.6422,
+      "step": 10314
+    },
+    {
+      "epoch": 2.7105461950274248,
+      "grad_norm": 0.5697504878044128,
+      "learning_rate": 9.654809882600315e-06,
+      "loss": 1.6299,
+      "step": 10316
+    },
+    {
+      "epoch": 2.7110716983610867,
+      "grad_norm": 0.5491029024124146,
+      "learning_rate": 9.63728754161556e-06,
+      "loss": 1.5843,
+      "step": 10318
+    },
+    {
+      "epoch": 2.7115972016947483,
+      "grad_norm": 0.8691264390945435,
+      "learning_rate": 9.619765200630806e-06,
+      "loss": 1.6086,
+      "step": 10320
+    },
+    {
+      "epoch": 2.71212270502841,
+      "grad_norm": 0.5798594355583191,
+      "learning_rate": 9.60224285964605e-06,
+      "loss": 1.6205,
+      "step": 10322
+    },
+    {
+      "epoch": 2.7126482083620718,
+      "grad_norm": 0.5886126756668091,
+      "learning_rate": 9.584720518661294e-06,
+      "loss": 1.5923,
+      "step": 10324
+    },
+    {
+      "epoch": 2.7131737116957337,
+      "grad_norm": 0.7271785140037537,
+      "learning_rate": 9.567198177676538e-06,
+      "loss": 1.623,
+      "step": 10326
+    },
+    {
+      "epoch": 2.7136992150293953,
+      "grad_norm": 0.5842270255088806,
+      "learning_rate": 9.549675836691782e-06,
+      "loss": 1.6141,
+      "step": 10328
+    },
+    {
+      "epoch": 2.7142247183630572,
+      "grad_norm": 0.5803746581077576,
+      "learning_rate": 9.532153495707028e-06,
+      "loss": 1.611,
+      "step": 10330
+    },
+    {
+      "epoch": 2.7147502216967188,
+      "grad_norm": 0.6185279488563538,
+      "learning_rate": 9.51463115472227e-06,
+      "loss": 1.6061,
+      "step": 10332
+    },
+    {
+      "epoch": 2.7152757250303807,
+      "grad_norm": 0.5305478572845459,
+      "learning_rate": 9.497108813737515e-06,
+      "loss": 1.6341,
+      "step": 10334
+    },
+    {
+      "epoch": 2.7158012283640423,
+      "grad_norm": 0.5952558517456055,
+      "learning_rate": 9.47958647275276e-06,
+      "loss": 1.6062,
+      "step": 10336
+    },
+    {
+      "epoch": 2.7163267316977042,
+      "grad_norm": 0.5312665700912476,
+      "learning_rate": 9.462064131768005e-06,
+      "loss": 1.6027,
+      "step": 10338
+    },
+    {
+      "epoch": 2.716852235031366,
+      "grad_norm": 0.6712905168533325,
+      "learning_rate": 9.444541790783249e-06,
+      "loss": 1.5846,
+      "step": 10340
+    },
+    {
+      "epoch": 2.7173777383650277,
+      "grad_norm": 0.6250156164169312,
+      "learning_rate": 9.427019449798493e-06,
+      "loss": 1.6207,
+      "step": 10342
+    },
+    {
+      "epoch": 2.7179032416986897,
+      "grad_norm": 0.6365962624549866,
+      "learning_rate": 9.409497108813737e-06,
+      "loss": 1.614,
+      "step": 10344
+    },
+    {
+      "epoch": 2.718428745032351,
+      "grad_norm": 0.5564280152320862,
+      "learning_rate": 9.391974767828983e-06,
+      "loss": 1.6054,
+      "step": 10346
+    },
+    {
+      "epoch": 2.718954248366013,
+      "grad_norm": 0.599860668182373,
+      "learning_rate": 9.374452426844227e-06,
+      "loss": 1.5914,
+      "step": 10348
+    },
+    {
+      "epoch": 2.7194797516996747,
+      "grad_norm": 0.6647321581840515,
+      "learning_rate": 9.356930085859472e-06,
+      "loss": 1.6001,
+      "step": 10350
+    },
+    {
+      "epoch": 2.7200052550333367,
+      "grad_norm": 0.5299103260040283,
+      "learning_rate": 9.339407744874716e-06,
+      "loss": 1.6181,
+      "step": 10352
+    },
+    {
+      "epoch": 2.7205307583669986,
+      "grad_norm": 0.6260030269622803,
+      "learning_rate": 9.32188540388996e-06,
+      "loss": 1.5963,
+      "step": 10354
+    },
+    {
+      "epoch": 2.72105626170066,
+      "grad_norm": 0.6055660247802734,
+      "learning_rate": 9.304363062905204e-06,
+      "loss": 1.564,
+      "step": 10356
+    },
+    {
+      "epoch": 2.7215817650343217,
+      "grad_norm": 0.6588335633277893,
+      "learning_rate": 9.28684072192045e-06,
+      "loss": 1.5692,
+      "step": 10358
+    },
+    {
+      "epoch": 2.7221072683679837,
+      "grad_norm": 0.5844119787216187,
+      "learning_rate": 9.269318380935692e-06,
+      "loss": 1.5931,
+      "step": 10360
+    },
+    {
+      "epoch": 2.7226327717016456,
+      "grad_norm": 0.5909096598625183,
+      "learning_rate": 9.251796039950938e-06,
+      "loss": 1.6011,
+      "step": 10362
+    },
+    {
+      "epoch": 2.723158275035307,
+      "grad_norm": 0.5176669955253601,
+      "learning_rate": 9.234273698966182e-06,
+      "loss": 1.5484,
+      "step": 10364
+    },
+    {
+      "epoch": 2.723683778368969,
+      "grad_norm": 0.5780521035194397,
+      "learning_rate": 9.216751357981427e-06,
+      "loss": 1.5989,
+      "step": 10366
+    },
+    {
+      "epoch": 2.7242092817026307,
+      "grad_norm": 0.6192241311073303,
+      "learning_rate": 9.19922901699667e-06,
+      "loss": 1.5719,
+      "step": 10368
+    },
+    {
+      "epoch": 2.7247347850362926,
+      "grad_norm": 0.5565381646156311,
+      "learning_rate": 9.181706676011915e-06,
+      "loss": 1.6109,
+      "step": 10370
+    },
+    {
+      "epoch": 2.725260288369954,
+      "grad_norm": 0.6125465631484985,
+      "learning_rate": 9.16418433502716e-06,
+      "loss": 1.61,
+      "step": 10372
+    },
+    {
+      "epoch": 2.725785791703616,
+      "grad_norm": 0.597527265548706,
+      "learning_rate": 9.146661994042405e-06,
+      "loss": 1.6093,
+      "step": 10374
+    },
+    {
+      "epoch": 2.726311295037278,
+      "grad_norm": 0.633154034614563,
+      "learning_rate": 9.12913965305765e-06,
+      "loss": 1.6101,
+      "step": 10376
+    },
+    {
+      "epoch": 2.7268367983709396,
+      "grad_norm": 0.5533692240715027,
+      "learning_rate": 9.111617312072893e-06,
+      "loss": 1.5807,
+      "step": 10378
+    },
+    {
+      "epoch": 2.7273623017046016,
+      "grad_norm": 0.6450961232185364,
+      "learning_rate": 9.094094971088138e-06,
+      "loss": 1.5835,
+      "step": 10380
+    },
+    {
+      "epoch": 2.727887805038263,
+      "grad_norm": 0.629943311214447,
+      "learning_rate": 9.076572630103382e-06,
+      "loss": 1.5977,
+      "step": 10382
+    },
+    {
+      "epoch": 2.728413308371925,
+      "grad_norm": 0.6521432995796204,
+      "learning_rate": 9.059050289118628e-06,
+      "loss": 1.6267,
+      "step": 10384
+    },
+    {
+      "epoch": 2.7289388117055866,
+      "grad_norm": 0.5852161645889282,
+      "learning_rate": 9.041527948133872e-06,
+      "loss": 1.6112,
+      "step": 10386
+    },
+    {
+      "epoch": 2.7294643150392486,
+      "grad_norm": 0.6588388085365295,
+      "learning_rate": 9.024005607149114e-06,
+      "loss": 1.6181,
+      "step": 10388
+    },
+    {
+      "epoch": 2.7299898183729105,
+      "grad_norm": 0.6498621106147766,
+      "learning_rate": 9.00648326616436e-06,
+      "loss": 1.596,
+      "step": 10390
+    },
+    {
+      "epoch": 2.730515321706572,
+      "grad_norm": 0.5801547169685364,
+      "learning_rate": 8.988960925179604e-06,
+      "loss": 1.5948,
+      "step": 10392
+    },
+    {
+      "epoch": 2.7310408250402336,
+      "grad_norm": 0.6173077821731567,
+      "learning_rate": 8.97143858419485e-06,
+      "loss": 1.6164,
+      "step": 10394
+    },
+    {
+      "epoch": 2.7315663283738956,
+      "grad_norm": 0.5936638116836548,
+      "learning_rate": 8.953916243210093e-06,
+      "loss": 1.5946,
+      "step": 10396
+    },
+    {
+      "epoch": 2.7320918317075575,
+      "grad_norm": 0.6268817186355591,
+      "learning_rate": 8.936393902225337e-06,
+      "loss": 1.6422,
+      "step": 10398
+    },
+    {
+      "epoch": 2.732617335041219,
+      "grad_norm": 0.7213032245635986,
+      "learning_rate": 8.918871561240583e-06,
+      "loss": 1.6387,
+      "step": 10400
+    },
+    {
+      "epoch": 2.732617335041219,
+      "eval_loss": 1.649667501449585,
+      "eval_runtime": 487.15,
+      "eval_samples_per_second": 250.003,
+      "eval_steps_per_second": 31.251,
+      "step": 10400
+    },
+    {
+      "epoch": 2.733142838374881,
+      "grad_norm": 0.5801553130149841,
+      "learning_rate": 8.901349220255827e-06,
+      "loss": 1.5527,
+      "step": 10402
+    },
+    {
+      "epoch": 2.7336683417085426,
+      "grad_norm": 0.6011367440223694,
+      "learning_rate": 8.883826879271071e-06,
+      "loss": 1.5901,
+      "step": 10404
+    },
+    {
+      "epoch": 2.7341938450422045,
+      "grad_norm": 0.5433336496353149,
+      "learning_rate": 8.866304538286315e-06,
+      "loss": 1.605,
+      "step": 10406
+    },
+    {
+      "epoch": 2.734719348375866,
+      "grad_norm": 0.5924966335296631,
+      "learning_rate": 8.84878219730156e-06,
+      "loss": 1.594,
+      "step": 10408
+    },
+    {
+      "epoch": 2.735244851709528,
+      "grad_norm": 0.6247406601905823,
+      "learning_rate": 8.831259856316805e-06,
+      "loss": 1.5501,
+      "step": 10410
+    },
+    {
+      "epoch": 2.73577035504319,
+      "grad_norm": 0.8343258500099182,
+      "learning_rate": 8.81373751533205e-06,
+      "loss": 1.6041,
+      "step": 10412
+    },
+    {
+      "epoch": 2.7362958583768515,
+      "grad_norm": 0.5661430358886719,
+      "learning_rate": 8.796215174347294e-06,
+      "loss": 1.6153,
+      "step": 10414
+    },
+    {
+      "epoch": 2.7368213617105135,
+      "grad_norm": 0.6813891530036926,
+      "learning_rate": 8.778692833362538e-06,
+      "loss": 1.6211,
+      "step": 10416
+    },
+    {
+      "epoch": 2.737346865044175,
+      "grad_norm": 0.6394498348236084,
+      "learning_rate": 8.761170492377782e-06,
+      "loss": 1.5989,
+      "step": 10418
+    },
+    {
+      "epoch": 2.737872368377837,
+      "grad_norm": 0.6066767573356628,
+      "learning_rate": 8.743648151393026e-06,
+      "loss": 1.6183,
+      "step": 10420
+    },
+    {
+      "epoch": 2.7383978717114985,
+      "grad_norm": 0.5361884236335754,
+      "learning_rate": 8.726125810408272e-06,
+      "loss": 1.619,
+      "step": 10422
+    },
+    {
+      "epoch": 2.7389233750451605,
+      "grad_norm": 0.5502820014953613,
+      "learning_rate": 8.708603469423514e-06,
+      "loss": 1.5936,
+      "step": 10424
+    },
+    {
+      "epoch": 2.7394488783788224,
+      "grad_norm": 0.6126530170440674,
+      "learning_rate": 8.691081128438759e-06,
+      "loss": 1.5771,
+      "step": 10426
+    },
+    {
+      "epoch": 2.739974381712484,
+      "grad_norm": 0.5665082335472107,
+      "learning_rate": 8.673558787454005e-06,
+      "loss": 1.608,
+      "step": 10428
+    },
+    {
+      "epoch": 2.7404998850461455,
+      "grad_norm": 0.6264111399650574,
+      "learning_rate": 8.656036446469249e-06,
+      "loss": 1.6095,
+      "step": 10430
+    },
+    {
+      "epoch": 2.7410253883798075,
+      "grad_norm": 0.6684525609016418,
+      "learning_rate": 8.638514105484493e-06,
+      "loss": 1.5786,
+      "step": 10432
+    },
+    {
+      "epoch": 2.7415508917134694,
+      "grad_norm": 0.5436965823173523,
+      "learning_rate": 8.620991764499737e-06,
+      "loss": 1.5764,
+      "step": 10434
+    },
+    {
+      "epoch": 2.742076395047131,
+      "grad_norm": 0.5792893171310425,
+      "learning_rate": 8.603469423514981e-06,
+      "loss": 1.6078,
+      "step": 10436
+    },
+    {
+      "epoch": 2.742601898380793,
+      "grad_norm": 0.5539526343345642,
+      "learning_rate": 8.585947082530227e-06,
+      "loss": 1.5919,
+      "step": 10438
+    },
+    {
+      "epoch": 2.7431274017144545,
+      "grad_norm": 0.5818884968757629,
+      "learning_rate": 8.568424741545471e-06,
+      "loss": 1.5926,
+      "step": 10440
+    },
+    {
+      "epoch": 2.7436529050481164,
+      "grad_norm": 0.5815169811248779,
+      "learning_rate": 8.550902400560715e-06,
+      "loss": 1.5986,
+      "step": 10442
+    },
+    {
+      "epoch": 2.744178408381778,
+      "grad_norm": 0.6487326622009277,
+      "learning_rate": 8.53338005957596e-06,
+      "loss": 1.6054,
+      "step": 10444
+    },
+    {
+      "epoch": 2.74470391171544,
+      "grad_norm": 0.6841205358505249,
+      "learning_rate": 8.515857718591204e-06,
+      "loss": 1.6031,
+      "step": 10446
+    },
+    {
+      "epoch": 2.745229415049102,
+      "grad_norm": 0.6351028680801392,
+      "learning_rate": 8.49833537760645e-06,
+      "loss": 1.6272,
+      "step": 10448
+    },
+    {
+      "epoch": 2.7457549183827634,
+      "grad_norm": 0.5550417900085449,
+      "learning_rate": 8.480813036621694e-06,
+      "loss": 1.6228,
+      "step": 10450
+    },
+    {
+      "epoch": 2.7462804217164254,
+      "grad_norm": 0.6113694906234741,
+      "learning_rate": 8.463290695636936e-06,
+      "loss": 1.5799,
+      "step": 10452
+    },
+    {
+      "epoch": 2.746805925050087,
+      "grad_norm": 0.5771463513374329,
+      "learning_rate": 8.445768354652182e-06,
+      "loss": 1.577,
+      "step": 10454
+    },
+    {
+      "epoch": 2.747331428383749,
+      "grad_norm": 0.5693290829658508,
+      "learning_rate": 8.428246013667426e-06,
+      "loss": 1.5841,
+      "step": 10456
+    },
+    {
+      "epoch": 2.7478569317174104,
+      "grad_norm": 0.5757125020027161,
+      "learning_rate": 8.41072367268267e-06,
+      "loss": 1.5785,
+      "step": 10458
+    },
+    {
+      "epoch": 2.7483824350510724,
+      "grad_norm": 0.643251359462738,
+      "learning_rate": 8.393201331697915e-06,
+      "loss": 1.611,
+      "step": 10460
+    },
+    {
+      "epoch": 2.7489079383847344,
+      "grad_norm": 0.6383783221244812,
+      "learning_rate": 8.375678990713159e-06,
+      "loss": 1.6065,
+      "step": 10462
+    },
+    {
+      "epoch": 2.749433441718396,
+      "grad_norm": 0.5447094440460205,
+      "learning_rate": 8.358156649728405e-06,
+      "loss": 1.5887,
+      "step": 10464
+    },
+    {
+      "epoch": 2.7499589450520574,
+      "grad_norm": 0.5493614077568054,
+      "learning_rate": 8.340634308743649e-06,
+      "loss": 1.5946,
+      "step": 10466
+    },
+    {
+      "epoch": 2.7504844483857194,
+      "grad_norm": 0.7292418479919434,
+      "learning_rate": 8.323111967758893e-06,
+      "loss": 1.5853,
+      "step": 10468
+    },
+    {
+      "epoch": 2.7510099517193813,
+      "grad_norm": 0.5726646780967712,
+      "learning_rate": 8.305589626774137e-06,
+      "loss": 1.6051,
+      "step": 10470
+    },
+    {
+      "epoch": 2.751535455053043,
+      "grad_norm": 0.6066914200782776,
+      "learning_rate": 8.288067285789381e-06,
+      "loss": 1.5952,
+      "step": 10472
+    },
+    {
+      "epoch": 2.752060958386705,
+      "grad_norm": 0.6483511328697205,
+      "learning_rate": 8.270544944804626e-06,
+      "loss": 1.6212,
+      "step": 10474
+    },
+    {
+      "epoch": 2.752586461720367,
+      "grad_norm": 0.5741682648658752,
+      "learning_rate": 8.253022603819872e-06,
+      "loss": 1.6178,
+      "step": 10476
+    },
+    {
+      "epoch": 2.7531119650540283,
+      "grad_norm": 0.5253785848617554,
+      "learning_rate": 8.235500262835116e-06,
+      "loss": 1.5889,
+      "step": 10478
+    },
+    {
+      "epoch": 2.75363746838769,
+      "grad_norm": 0.557144820690155,
+      "learning_rate": 8.21797792185036e-06,
+      "loss": 1.6081,
+      "step": 10480
+    },
+    {
+      "epoch": 2.754162971721352,
+      "grad_norm": 0.5865209698677063,
+      "learning_rate": 8.200455580865604e-06,
+      "loss": 1.603,
+      "step": 10482
+    },
+    {
+      "epoch": 2.754688475055014,
+      "grad_norm": 0.5472773909568787,
+      "learning_rate": 8.182933239880848e-06,
+      "loss": 1.6022,
+      "step": 10484
+    },
+    {
+      "epoch": 2.7552139783886753,
+      "grad_norm": 0.6681110262870789,
+      "learning_rate": 8.165410898896094e-06,
+      "loss": 1.5927,
+      "step": 10486
+    },
+    {
+      "epoch": 2.7557394817223373,
+      "grad_norm": 0.6070604920387268,
+      "learning_rate": 8.147888557911337e-06,
+      "loss": 1.6051,
+      "step": 10488
+    },
+    {
+      "epoch": 2.756264985055999,
+      "grad_norm": 0.5712462663650513,
+      "learning_rate": 8.13036621692658e-06,
+      "loss": 1.6124,
+      "step": 10490
+    },
+    {
+      "epoch": 2.756790488389661,
+      "grad_norm": 0.5560539364814758,
+      "learning_rate": 8.112843875941827e-06,
+      "loss": 1.6116,
+      "step": 10492
+    },
+    {
+      "epoch": 2.7573159917233223,
+      "grad_norm": 0.6540628671646118,
+      "learning_rate": 8.09532153495707e-06,
+      "loss": 1.5916,
+      "step": 10494
+    },
+    {
+      "epoch": 2.7578414950569843,
+      "grad_norm": 0.5967172384262085,
+      "learning_rate": 8.077799193972315e-06,
+      "loss": 1.5789,
+      "step": 10496
+    },
+    {
+      "epoch": 2.7583669983906463,
+      "grad_norm": 0.6844902634620667,
+      "learning_rate": 8.060276852987559e-06,
+      "loss": 1.6098,
+      "step": 10498
+    },
+    {
+      "epoch": 2.7588925017243078,
+      "grad_norm": 0.6153303980827332,
+      "learning_rate": 8.042754512002803e-06,
+      "loss": 1.6149,
+      "step": 10500
+    },
+    {
+      "epoch": 2.7594180050579697,
+      "grad_norm": 0.5895024538040161,
+      "learning_rate": 8.02523217101805e-06,
+      "loss": 1.6097,
+      "step": 10502
+    },
+    {
+      "epoch": 2.7599435083916313,
+      "grad_norm": 0.6543148159980774,
+      "learning_rate": 8.007709830033293e-06,
+      "loss": 1.5865,
+      "step": 10504
+    },
+    {
+      "epoch": 2.7604690117252932,
+      "grad_norm": 0.6354900002479553,
+      "learning_rate": 7.990187489048538e-06,
+      "loss": 1.587,
+      "step": 10506
+    },
+    {
+      "epoch": 2.7609945150589548,
+      "grad_norm": 0.6232167482376099,
+      "learning_rate": 7.972665148063782e-06,
+      "loss": 1.622,
+      "step": 10508
+    },
+    {
+      "epoch": 2.7615200183926167,
+      "grad_norm": 0.5841950178146362,
+      "learning_rate": 7.955142807079026e-06,
+      "loss": 1.6159,
+      "step": 10510
+    },
+    {
+      "epoch": 2.7620455217262787,
+      "grad_norm": 0.5530897974967957,
+      "learning_rate": 7.93762046609427e-06,
+      "loss": 1.5696,
+      "step": 10512
+    },
+    {
+      "epoch": 2.7625710250599402,
+      "grad_norm": 0.5793708562850952,
+      "learning_rate": 7.920098125109516e-06,
+      "loss": 1.5672,
+      "step": 10514
+    },
+    {
+      "epoch": 2.7630965283936018,
+      "grad_norm": 0.595795750617981,
+      "learning_rate": 7.902575784124758e-06,
+      "loss": 1.6003,
+      "step": 10516
+    },
+    {
+      "epoch": 2.7636220317272637,
+      "grad_norm": 0.5570265054702759,
+      "learning_rate": 7.885053443140004e-06,
+      "loss": 1.5881,
+      "step": 10518
+    },
+    {
+      "epoch": 2.7641475350609257,
+      "grad_norm": 0.6062209606170654,
+      "learning_rate": 7.867531102155248e-06,
+      "loss": 1.5865,
+      "step": 10520
+    },
+    {
+      "epoch": 2.7646730383945872,
+      "grad_norm": 0.6273199319839478,
+      "learning_rate": 7.850008761170493e-06,
+      "loss": 1.6095,
+      "step": 10522
+    },
+    {
+      "epoch": 2.765198541728249,
+      "grad_norm": 0.6534038782119751,
+      "learning_rate": 7.832486420185737e-06,
+      "loss": 1.5834,
+      "step": 10524
+    },
+    {
+      "epoch": 2.7657240450619107,
+      "grad_norm": 0.7082603573799133,
+      "learning_rate": 7.814964079200981e-06,
+      "loss": 1.5977,
+      "step": 10526
+    },
+    {
+      "epoch": 2.7662495483955727,
+      "grad_norm": 0.8246460556983948,
+      "learning_rate": 7.797441738216225e-06,
+      "loss": 1.5952,
+      "step": 10528
+    },
+    {
+      "epoch": 2.766775051729234,
+      "grad_norm": 0.6671035289764404,
+      "learning_rate": 7.779919397231471e-06,
+      "loss": 1.6045,
+      "step": 10530
+    },
+    {
+      "epoch": 2.767300555062896,
+      "grad_norm": 0.6524311900138855,
+      "learning_rate": 7.762397056246715e-06,
+      "loss": 1.6305,
+      "step": 10532
+    },
+    {
+      "epoch": 2.767826058396558,
+      "grad_norm": 0.5577046275138855,
+      "learning_rate": 7.74487471526196e-06,
+      "loss": 1.5955,
+      "step": 10534
+    },
+    {
+      "epoch": 2.7683515617302197,
+      "grad_norm": 0.5540436506271362,
+      "learning_rate": 7.727352374277204e-06,
+      "loss": 1.6009,
+      "step": 10536
+    },
+    {
+      "epoch": 2.7688770650638816,
+      "grad_norm": 0.5906193256378174,
+      "learning_rate": 7.709830033292448e-06,
+      "loss": 1.5597,
+      "step": 10538
+    },
+    {
+      "epoch": 2.769402568397543,
+      "grad_norm": 0.7687748074531555,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.6085,
+      "step": 10540
+    },
+    {
+      "epoch": 2.769928071731205,
+      "grad_norm": 0.5762370824813843,
+      "learning_rate": 7.674785351322938e-06,
+      "loss": 1.6209,
+      "step": 10542
+    },
+    {
+      "epoch": 2.7704535750648667,
+      "grad_norm": 0.6532847285270691,
+      "learning_rate": 7.65726301033818e-06,
+      "loss": 1.5809,
+      "step": 10544
+    },
+    {
+      "epoch": 2.7709790783985286,
+      "grad_norm": 0.6120353937149048,
+      "learning_rate": 7.639740669353426e-06,
+      "loss": 1.5994,
+      "step": 10546
+    },
+    {
+      "epoch": 2.7715045817321906,
+      "grad_norm": 0.5683396458625793,
+      "learning_rate": 7.62221832836867e-06,
+      "loss": 1.6049,
+      "step": 10548
+    },
+    {
+      "epoch": 2.772030085065852,
+      "grad_norm": 0.6380746364593506,
+      "learning_rate": 7.604695987383915e-06,
+      "loss": 1.5922,
+      "step": 10550
+    },
+    {
+      "epoch": 2.7725555883995137,
+      "grad_norm": 0.619238018989563,
+      "learning_rate": 7.5871736463991595e-06,
+      "loss": 1.6036,
+      "step": 10552
+    },
+    {
+      "epoch": 2.7730810917331756,
+      "grad_norm": 0.6409256458282471,
+      "learning_rate": 7.569651305414404e-06,
+      "loss": 1.6094,
+      "step": 10554
+    },
+    {
+      "epoch": 2.7736065950668376,
+      "grad_norm": 0.6005852818489075,
+      "learning_rate": 7.552128964429649e-06,
+      "loss": 1.5932,
+      "step": 10556
+    },
+    {
+      "epoch": 2.774132098400499,
+      "grad_norm": 0.5933837890625,
+      "learning_rate": 7.534606623444893e-06,
+      "loss": 1.5956,
+      "step": 10558
+    },
+    {
+      "epoch": 2.774657601734161,
+      "grad_norm": 0.6530826687812805,
+      "learning_rate": 7.517084282460136e-06,
+      "loss": 1.6305,
+      "step": 10560
+    },
+    {
+      "epoch": 2.7751831050678226,
+      "grad_norm": 0.7053954601287842,
+      "learning_rate": 7.499561941475382e-06,
+      "loss": 1.6037,
+      "step": 10562
+    },
+    {
+      "epoch": 2.7757086084014846,
+      "grad_norm": 0.5595335364341736,
+      "learning_rate": 7.482039600490625e-06,
+      "loss": 1.6028,
+      "step": 10564
+    },
+    {
+      "epoch": 2.776234111735146,
+      "grad_norm": 0.6155191659927368,
+      "learning_rate": 7.464517259505871e-06,
+      "loss": 1.591,
+      "step": 10566
+    },
+    {
+      "epoch": 2.776759615068808,
+      "grad_norm": 0.5995579361915588,
+      "learning_rate": 7.446994918521115e-06,
+      "loss": 1.6206,
+      "step": 10568
+    },
+    {
+      "epoch": 2.77728511840247,
+      "grad_norm": 0.5798934698104858,
+      "learning_rate": 7.429472577536359e-06,
+      "loss": 1.5958,
+      "step": 10570
+    },
+    {
+      "epoch": 2.7778106217361316,
+      "grad_norm": 0.6855894923210144,
+      "learning_rate": 7.411950236551604e-06,
+      "loss": 1.6102,
+      "step": 10572
+    },
+    {
+      "epoch": 2.7783361250697935,
+      "grad_norm": 0.6109877228736877,
+      "learning_rate": 7.394427895566848e-06,
+      "loss": 1.5838,
+      "step": 10574
+    },
+    {
+      "epoch": 2.778861628403455,
+      "grad_norm": 0.5831050276756287,
+      "learning_rate": 7.376905554582092e-06,
+      "loss": 1.5936,
+      "step": 10576
+    },
+    {
+      "epoch": 2.779387131737117,
+      "grad_norm": 0.6219046711921692,
+      "learning_rate": 7.359383213597337e-06,
+      "loss": 1.6013,
+      "step": 10578
+    },
+    {
+      "epoch": 2.7799126350707786,
+      "grad_norm": 0.6402764916419983,
+      "learning_rate": 7.341860872612581e-06,
+      "loss": 1.6106,
+      "step": 10580
+    },
+    {
+      "epoch": 2.7804381384044405,
+      "grad_norm": 0.6024214625358582,
+      "learning_rate": 7.324338531627826e-06,
+      "loss": 1.6195,
+      "step": 10582
+    },
+    {
+      "epoch": 2.7809636417381025,
+      "grad_norm": 0.5932384729385376,
+      "learning_rate": 7.3068161906430705e-06,
+      "loss": 1.5976,
+      "step": 10584
+    },
+    {
+      "epoch": 2.781489145071764,
+      "grad_norm": 0.631115198135376,
+      "learning_rate": 7.289293849658315e-06,
+      "loss": 1.5967,
+      "step": 10586
+    },
+    {
+      "epoch": 2.7820146484054256,
+      "grad_norm": 0.6470726132392883,
+      "learning_rate": 7.27177150867356e-06,
+      "loss": 1.5992,
+      "step": 10588
+    },
+    {
+      "epoch": 2.7825401517390875,
+      "grad_norm": 0.6247832775115967,
+      "learning_rate": 7.254249167688804e-06,
+      "loss": 1.6331,
+      "step": 10590
+    },
+    {
+      "epoch": 2.7830656550727495,
+      "grad_norm": 0.5455314517021179,
+      "learning_rate": 7.236726826704047e-06,
+      "loss": 1.6089,
+      "step": 10592
+    },
+    {
+      "epoch": 2.783591158406411,
+      "grad_norm": 0.7327371835708618,
+      "learning_rate": 7.219204485719293e-06,
+      "loss": 1.5804,
+      "step": 10594
+    },
+    {
+      "epoch": 2.784116661740073,
+      "grad_norm": 0.6657344102859497,
+      "learning_rate": 7.201682144734536e-06,
+      "loss": 1.5925,
+      "step": 10596
+    },
+    {
+      "epoch": 2.7846421650737345,
+      "grad_norm": 0.6697341203689575,
+      "learning_rate": 7.184159803749781e-06,
+      "loss": 1.5906,
+      "step": 10598
+    },
+    {
+      "epoch": 2.7851676684073965,
+      "grad_norm": 0.6150529384613037,
+      "learning_rate": 7.166637462765026e-06,
+      "loss": 1.6017,
+      "step": 10600
+    },
+    {
+      "epoch": 2.785693171741058,
+      "grad_norm": 0.5665842890739441,
+      "learning_rate": 7.14911512178027e-06,
+      "loss": 1.581,
+      "step": 10602
+    },
+    {
+      "epoch": 2.78621867507472,
+      "grad_norm": 0.5935694575309753,
+      "learning_rate": 7.131592780795515e-06,
+      "loss": 1.5845,
+      "step": 10604
+    },
+    {
+      "epoch": 2.786744178408382,
+      "grad_norm": 0.6057782769203186,
+      "learning_rate": 7.114070439810759e-06,
+      "loss": 1.5985,
+      "step": 10606
+    },
+    {
+      "epoch": 2.7872696817420435,
+      "grad_norm": 0.5478567481040955,
+      "learning_rate": 7.096548098826003e-06,
+      "loss": 1.6067,
+      "step": 10608
+    },
+    {
+      "epoch": 2.7877951850757055,
+      "grad_norm": 0.5651966333389282,
+      "learning_rate": 7.079025757841248e-06,
+      "loss": 1.6157,
+      "step": 10610
+    },
+    {
+      "epoch": 2.788320688409367,
+      "grad_norm": 0.6189813017845154,
+      "learning_rate": 7.061503416856492e-06,
+      "loss": 1.5912,
+      "step": 10612
+    },
+    {
+      "epoch": 2.788846191743029,
+      "grad_norm": 0.6159524917602539,
+      "learning_rate": 7.0439810758717365e-06,
+      "loss": 1.622,
+      "step": 10614
+    },
+    {
+      "epoch": 2.7893716950766905,
+      "grad_norm": 0.5482856631278992,
+      "learning_rate": 7.0264587348869816e-06,
+      "loss": 1.6197,
+      "step": 10616
+    },
+    {
+      "epoch": 2.7898971984103524,
+      "grad_norm": 0.6579813957214355,
+      "learning_rate": 7.008936393902226e-06,
+      "loss": 1.6182,
+      "step": 10618
+    },
+    {
+      "epoch": 2.7904227017440144,
+      "grad_norm": 0.607364296913147,
+      "learning_rate": 6.991414052917471e-06,
+      "loss": 1.6183,
+      "step": 10620
+    },
+    {
+      "epoch": 2.790948205077676,
+      "grad_norm": 0.649757444858551,
+      "learning_rate": 6.973891711932715e-06,
+      "loss": 1.6179,
+      "step": 10622
+    },
+    {
+      "epoch": 2.7914737084113375,
+      "grad_norm": 0.5615342259407043,
+      "learning_rate": 6.956369370947958e-06,
+      "loss": 1.632,
+      "step": 10624
+    },
+    {
+      "epoch": 2.7919992117449994,
+      "grad_norm": 0.6041872501373291,
+      "learning_rate": 6.938847029963204e-06,
+      "loss": 1.5943,
+      "step": 10626
+    },
+    {
+      "epoch": 2.7925247150786614,
+      "grad_norm": 0.6032666563987732,
+      "learning_rate": 6.9213246889784475e-06,
+      "loss": 1.5552,
+      "step": 10628
+    },
+    {
+      "epoch": 2.793050218412323,
+      "grad_norm": 0.5664829611778259,
+      "learning_rate": 6.903802347993692e-06,
+      "loss": 1.5918,
+      "step": 10630
+    },
+    {
+      "epoch": 2.793575721745985,
+      "grad_norm": 0.633831262588501,
+      "learning_rate": 6.886280007008937e-06,
+      "loss": 1.5763,
+      "step": 10632
+    },
+    {
+      "epoch": 2.794101225079647,
+      "grad_norm": 0.5904597640037537,
+      "learning_rate": 6.868757666024181e-06,
+      "loss": 1.5659,
+      "step": 10634
+    },
+    {
+      "epoch": 2.7946267284133084,
+      "grad_norm": 0.556845486164093,
+      "learning_rate": 6.851235325039426e-06,
+      "loss": 1.5793,
+      "step": 10636
+    },
+    {
+      "epoch": 2.79515223174697,
+      "grad_norm": 0.6646136045455933,
+      "learning_rate": 6.83371298405467e-06,
+      "loss": 1.6201,
+      "step": 10638
+    },
+    {
+      "epoch": 2.795677735080632,
+      "grad_norm": 0.5443913340568542,
+      "learning_rate": 6.816190643069914e-06,
+      "loss": 1.6126,
+      "step": 10640
+    },
+    {
+      "epoch": 2.796203238414294,
+      "grad_norm": 0.5882983803749084,
+      "learning_rate": 6.798668302085159e-06,
+      "loss": 1.572,
+      "step": 10642
+    },
+    {
+      "epoch": 2.7967287417479554,
+      "grad_norm": 0.5482323169708252,
+      "learning_rate": 6.781145961100403e-06,
+      "loss": 1.6127,
+      "step": 10644
+    },
+    {
+      "epoch": 2.7972542450816174,
+      "grad_norm": 0.5871306657791138,
+      "learning_rate": 6.763623620115648e-06,
+      "loss": 1.6031,
+      "step": 10646
+    },
+    {
+      "epoch": 2.797779748415279,
+      "grad_norm": 0.6310222744941711,
+      "learning_rate": 6.746101279130893e-06,
+      "loss": 1.6017,
+      "step": 10648
+    },
+    {
+      "epoch": 2.798305251748941,
+      "grad_norm": 0.8194197416305542,
+      "learning_rate": 6.728578938146137e-06,
+      "loss": 1.5373,
+      "step": 10650
+    },
+    {
+      "epoch": 2.7988307550826024,
+      "grad_norm": 0.5392462015151978,
+      "learning_rate": 6.711056597161382e-06,
+      "loss": 1.6042,
+      "step": 10652
+    },
+    {
+      "epoch": 2.7993562584162643,
+      "grad_norm": 0.5910424590110779,
+      "learning_rate": 6.693534256176626e-06,
+      "loss": 1.6332,
+      "step": 10654
+    },
+    {
+      "epoch": 2.7998817617499263,
+      "grad_norm": 0.6581763029098511,
+      "learning_rate": 6.676011915191869e-06,
+      "loss": 1.617,
+      "step": 10656
+    },
+    {
+      "epoch": 2.800407265083588,
+      "grad_norm": 0.6362857222557068,
+      "learning_rate": 6.658489574207115e-06,
+      "loss": 1.5952,
+      "step": 10658
+    },
+    {
+      "epoch": 2.80093276841725,
+      "grad_norm": 0.6781132817268372,
+      "learning_rate": 6.6409672332223585e-06,
+      "loss": 1.571,
+      "step": 10660
+    },
+    {
+      "epoch": 2.8014582717509113,
+      "grad_norm": 0.609855055809021,
+      "learning_rate": 6.623444892237603e-06,
+      "loss": 1.6155,
+      "step": 10662
+    },
+    {
+      "epoch": 2.8019837750845733,
+      "grad_norm": 0.5758035182952881,
+      "learning_rate": 6.605922551252848e-06,
+      "loss": 1.5991,
+      "step": 10664
+    },
+    {
+      "epoch": 2.802509278418235,
+      "grad_norm": 0.6021389961242676,
+      "learning_rate": 6.588400210268092e-06,
+      "loss": 1.6027,
+      "step": 10666
+    },
+    {
+      "epoch": 2.803034781751897,
+      "grad_norm": 0.6531217098236084,
+      "learning_rate": 6.570877869283337e-06,
+      "loss": 1.6246,
+      "step": 10668
+    },
+    {
+      "epoch": 2.8035602850855588,
+      "grad_norm": 0.7034488916397095,
+      "learning_rate": 6.553355528298581e-06,
+      "loss": 1.599,
+      "step": 10670
+    },
+    {
+      "epoch": 2.8040857884192203,
+      "grad_norm": 0.5464747548103333,
+      "learning_rate": 6.535833187313825e-06,
+      "loss": 1.6301,
+      "step": 10672
+    },
+    {
+      "epoch": 2.804611291752882,
+      "grad_norm": 0.6323386430740356,
+      "learning_rate": 6.51831084632907e-06,
+      "loss": 1.634,
+      "step": 10674
+    },
+    {
+      "epoch": 2.805136795086544,
+      "grad_norm": 0.5574644207954407,
+      "learning_rate": 6.5007885053443144e-06,
+      "loss": 1.5634,
+      "step": 10676
+    },
+    {
+      "epoch": 2.8056622984202058,
+      "grad_norm": 0.641542911529541,
+      "learning_rate": 6.483266164359559e-06,
+      "loss": 1.5822,
+      "step": 10678
+    },
+    {
+      "epoch": 2.8061878017538673,
+      "grad_norm": 0.5666943788528442,
+      "learning_rate": 6.465743823374804e-06,
+      "loss": 1.5812,
+      "step": 10680
+    },
+    {
+      "epoch": 2.8067133050875293,
+      "grad_norm": 0.5939244627952576,
+      "learning_rate": 6.448221482390048e-06,
+      "loss": 1.5907,
+      "step": 10682
+    },
+    {
+      "epoch": 2.807238808421191,
+      "grad_norm": 0.5618119835853577,
+      "learning_rate": 6.430699141405291e-06,
+      "loss": 1.5859,
+      "step": 10684
+    },
+    {
+      "epoch": 2.8077643117548527,
+      "grad_norm": 0.5734422206878662,
+      "learning_rate": 6.413176800420537e-06,
+      "loss": 1.5904,
+      "step": 10686
+    },
+    {
+      "epoch": 2.8082898150885143,
+      "grad_norm": 0.5530012845993042,
+      "learning_rate": 6.39565445943578e-06,
+      "loss": 1.6041,
+      "step": 10688
+    },
+    {
+      "epoch": 2.8088153184221762,
+      "grad_norm": 0.6470913290977478,
+      "learning_rate": 6.378132118451026e-06,
+      "loss": 1.594,
+      "step": 10690
+    },
+    {
+      "epoch": 2.809340821755838,
+      "grad_norm": 0.6118698120117188,
+      "learning_rate": 6.3606097774662695e-06,
+      "loss": 1.625,
+      "step": 10692
+    },
+    {
+      "epoch": 2.8098663250894997,
+      "grad_norm": 0.6750172972679138,
+      "learning_rate": 6.343087436481514e-06,
+      "loss": 1.5389,
+      "step": 10694
+    },
+    {
+      "epoch": 2.8103918284231617,
+      "grad_norm": 0.6908929347991943,
+      "learning_rate": 6.325565095496759e-06,
+      "loss": 1.6215,
+      "step": 10696
+    },
+    {
+      "epoch": 2.8109173317568232,
+      "grad_norm": 0.6111420392990112,
+      "learning_rate": 6.308042754512003e-06,
+      "loss": 1.5701,
+      "step": 10698
+    },
+    {
+      "epoch": 2.811442835090485,
+      "grad_norm": 0.6819601655006409,
+      "learning_rate": 6.290520413527247e-06,
+      "loss": 1.5895,
+      "step": 10700
+    },
+    {
+      "epoch": 2.8119683384241467,
+      "grad_norm": 0.6061453819274902,
+      "learning_rate": 6.272998072542492e-06,
+      "loss": 1.6076,
+      "step": 10702
+    },
+    {
+      "epoch": 2.8124938417578087,
+      "grad_norm": 0.6342339515686035,
+      "learning_rate": 6.255475731557736e-06,
+      "loss": 1.6128,
+      "step": 10704
+    },
+    {
+      "epoch": 2.8130193450914707,
+      "grad_norm": 0.6219684481620789,
+      "learning_rate": 6.2379533905729805e-06,
+      "loss": 1.6264,
+      "step": 10706
+    },
+    {
+      "epoch": 2.813544848425132,
+      "grad_norm": 0.779617190361023,
+      "learning_rate": 6.2204310495882255e-06,
+      "loss": 1.6218,
+      "step": 10708
+    },
+    {
+      "epoch": 2.8140703517587937,
+      "grad_norm": 0.5725176334381104,
+      "learning_rate": 6.20290870860347e-06,
+      "loss": 1.5772,
+      "step": 10710
+    },
+    {
+      "epoch": 2.8145958550924557,
+      "grad_norm": 0.6861016750335693,
+      "learning_rate": 6.185386367618714e-06,
+      "loss": 1.604,
+      "step": 10712
+    },
+    {
+      "epoch": 2.8151213584261177,
+      "grad_norm": 0.6791368126869202,
+      "learning_rate": 6.167864026633959e-06,
+      "loss": 1.6244,
+      "step": 10714
+    },
+    {
+      "epoch": 2.815646861759779,
+      "grad_norm": 0.5471606850624084,
+      "learning_rate": 6.150341685649203e-06,
+      "loss": 1.589,
+      "step": 10716
+    },
+    {
+      "epoch": 2.816172365093441,
+      "grad_norm": 0.5449687242507935,
+      "learning_rate": 6.132819344664448e-06,
+      "loss": 1.6234,
+      "step": 10718
+    },
+    {
+      "epoch": 2.8166978684271027,
+      "grad_norm": 0.6119715571403503,
+      "learning_rate": 6.115297003679691e-06,
+      "loss": 1.5809,
+      "step": 10720
+    },
+    {
+      "epoch": 2.8172233717607646,
+      "grad_norm": 0.5771490931510925,
+      "learning_rate": 6.097774662694936e-06,
+      "loss": 1.5833,
+      "step": 10722
+    },
+    {
+      "epoch": 2.817748875094426,
+      "grad_norm": 0.5519348978996277,
+      "learning_rate": 6.0802523217101806e-06,
+      "loss": 1.6195,
+      "step": 10724
+    },
+    {
+      "epoch": 2.818274378428088,
+      "grad_norm": 0.7355902791023254,
+      "learning_rate": 6.062729980725426e-06,
+      "loss": 1.54,
+      "step": 10726
+    },
+    {
+      "epoch": 2.81879988176175,
+      "grad_norm": 0.5478994250297546,
+      "learning_rate": 6.04520763974067e-06,
+      "loss": 1.6225,
+      "step": 10728
+    },
+    {
+      "epoch": 2.8193253850954116,
+      "grad_norm": 0.5732230544090271,
+      "learning_rate": 6.027685298755914e-06,
+      "loss": 1.5899,
+      "step": 10730
+    },
+    {
+      "epoch": 2.8198508884290736,
+      "grad_norm": 0.6284743547439575,
+      "learning_rate": 6.010162957771159e-06,
+      "loss": 1.6116,
+      "step": 10732
+    },
+    {
+      "epoch": 2.820376391762735,
+      "grad_norm": 0.6271458864212036,
+      "learning_rate": 5.992640616786403e-06,
+      "loss": 1.5938,
+      "step": 10734
+    },
+    {
+      "epoch": 2.820901895096397,
+      "grad_norm": 0.5997607111930847,
+      "learning_rate": 5.975118275801647e-06,
+      "loss": 1.6035,
+      "step": 10736
+    },
+    {
+      "epoch": 2.8214273984300586,
+      "grad_norm": 0.6365160942077637,
+      "learning_rate": 5.9575959348168915e-06,
+      "loss": 1.6172,
+      "step": 10738
+    },
+    {
+      "epoch": 2.8219529017637206,
+      "grad_norm": 0.7437217235565186,
+      "learning_rate": 5.9400735938321365e-06,
+      "loss": 1.605,
+      "step": 10740
+    },
+    {
+      "epoch": 2.8224784050973826,
+      "grad_norm": 0.6684809923171997,
+      "learning_rate": 5.922551252847381e-06,
+      "loss": 1.6066,
+      "step": 10742
+    },
+    {
+      "epoch": 2.823003908431044,
+      "grad_norm": 0.5696150064468384,
+      "learning_rate": 5.905028911862625e-06,
+      "loss": 1.6137,
+      "step": 10744
+    },
+    {
+      "epoch": 2.8235294117647056,
+      "grad_norm": 0.6045554876327515,
+      "learning_rate": 5.88750657087787e-06,
+      "loss": 1.6169,
+      "step": 10746
+    },
+    {
+      "epoch": 2.8240549150983676,
+      "grad_norm": 0.5769687294960022,
+      "learning_rate": 5.869984229893114e-06,
+      "loss": 1.612,
+      "step": 10748
+    },
+    {
+      "epoch": 2.8245804184320296,
+      "grad_norm": 0.6565601825714111,
+      "learning_rate": 5.852461888908359e-06,
+      "loss": 1.6078,
+      "step": 10750
+    },
+    {
+      "epoch": 2.825105921765691,
+      "grad_norm": 0.614622175693512,
+      "learning_rate": 5.834939547923602e-06,
+      "loss": 1.6229,
+      "step": 10752
+    },
+    {
+      "epoch": 2.825631425099353,
+      "grad_norm": 0.5989246964454651,
+      "learning_rate": 5.8174172069388474e-06,
+      "loss": 1.6167,
+      "step": 10754
+    },
+    {
+      "epoch": 2.8261569284330146,
+      "grad_norm": 0.6011819243431091,
+      "learning_rate": 5.799894865954092e-06,
+      "loss": 1.5975,
+      "step": 10756
+    },
+    {
+      "epoch": 2.8266824317666766,
+      "grad_norm": 0.5927401185035706,
+      "learning_rate": 5.782372524969336e-06,
+      "loss": 1.6034,
+      "step": 10758
+    },
+    {
+      "epoch": 2.827207935100338,
+      "grad_norm": 0.6807177066802979,
+      "learning_rate": 5.764850183984581e-06,
+      "loss": 1.6005,
+      "step": 10760
+    },
+    {
+      "epoch": 2.827733438434,
+      "grad_norm": 0.637870728969574,
+      "learning_rate": 5.747327842999825e-06,
+      "loss": 1.6184,
+      "step": 10762
+    },
+    {
+      "epoch": 2.828258941767662,
+      "grad_norm": 0.5749346017837524,
+      "learning_rate": 5.72980550201507e-06,
+      "loss": 1.6022,
+      "step": 10764
+    },
+    {
+      "epoch": 2.8287844451013235,
+      "grad_norm": 0.5985031127929688,
+      "learning_rate": 5.712283161030313e-06,
+      "loss": 1.5792,
+      "step": 10766
+    },
+    {
+      "epoch": 2.8293099484349855,
+      "grad_norm": 0.5938543677330017,
+      "learning_rate": 5.694760820045558e-06,
+      "loss": 1.621,
+      "step": 10768
+    },
+    {
+      "epoch": 2.829835451768647,
+      "grad_norm": 0.6295526027679443,
+      "learning_rate": 5.6772384790608025e-06,
+      "loss": 1.6134,
+      "step": 10770
+    },
+    {
+      "epoch": 2.830360955102309,
+      "grad_norm": 0.5686202049255371,
+      "learning_rate": 5.6597161380760476e-06,
+      "loss": 1.5748,
+      "step": 10772
+    },
+    {
+      "epoch": 2.8308864584359705,
+      "grad_norm": 0.5600228309631348,
+      "learning_rate": 5.642193797091292e-06,
+      "loss": 1.5901,
+      "step": 10774
+    },
+    {
+      "epoch": 2.8314119617696325,
+      "grad_norm": 0.5651440024375916,
+      "learning_rate": 5.624671456106536e-06,
+      "loss": 1.6054,
+      "step": 10776
+    },
+    {
+      "epoch": 2.8319374651032945,
+      "grad_norm": 0.5856150984764099,
+      "learning_rate": 5.607149115121781e-06,
+      "loss": 1.5641,
+      "step": 10778
+    },
+    {
+      "epoch": 2.832462968436956,
+      "grad_norm": 0.6541823148727417,
+      "learning_rate": 5.589626774137025e-06,
+      "loss": 1.5969,
+      "step": 10780
+    },
+    {
+      "epoch": 2.8329884717706175,
+      "grad_norm": 0.577987551689148,
+      "learning_rate": 5.572104433152269e-06,
+      "loss": 1.5768,
+      "step": 10782
+    },
+    {
+      "epoch": 2.8335139751042795,
+      "grad_norm": 0.6325380206108093,
+      "learning_rate": 5.5545820921675135e-06,
+      "loss": 1.5626,
+      "step": 10784
+    },
+    {
+      "epoch": 2.8340394784379415,
+      "grad_norm": 0.5774263143539429,
+      "learning_rate": 5.5370597511827585e-06,
+      "loss": 1.6056,
+      "step": 10786
+    },
+    {
+      "epoch": 2.834564981771603,
+      "grad_norm": 0.5753619074821472,
+      "learning_rate": 5.519537410198003e-06,
+      "loss": 1.5734,
+      "step": 10788
+    },
+    {
+      "epoch": 2.835090485105265,
+      "grad_norm": 0.5787561535835266,
+      "learning_rate": 5.502015069213247e-06,
+      "loss": 1.6027,
+      "step": 10790
+    },
+    {
+      "epoch": 2.835615988438927,
+      "grad_norm": 0.616862952709198,
+      "learning_rate": 5.484492728228492e-06,
+      "loss": 1.6296,
+      "step": 10792
+    },
+    {
+      "epoch": 2.8361414917725885,
+      "grad_norm": 0.5734919905662537,
+      "learning_rate": 5.466970387243736e-06,
+      "loss": 1.5966,
+      "step": 10794
+    },
+    {
+      "epoch": 2.83666699510625,
+      "grad_norm": 0.5936645865440369,
+      "learning_rate": 5.449448046258981e-06,
+      "loss": 1.6018,
+      "step": 10796
+    },
+    {
+      "epoch": 2.837192498439912,
+      "grad_norm": 0.6087591052055359,
+      "learning_rate": 5.431925705274224e-06,
+      "loss": 1.6201,
+      "step": 10798
+    },
+    {
+      "epoch": 2.837718001773574,
+      "grad_norm": 0.609794557094574,
+      "learning_rate": 5.414403364289469e-06,
+      "loss": 1.6401,
+      "step": 10800
+    },
+    {
+      "epoch": 2.837718001773574,
+      "eval_loss": 1.6476643085479736,
+      "eval_runtime": 487.2582,
+      "eval_samples_per_second": 249.948,
+      "eval_steps_per_second": 31.244,
+      "step": 10800
+    },
+    {
+      "epoch": 2.8382435051072354,
+      "grad_norm": 0.8440867066383362,
+      "learning_rate": 5.3968810233047136e-06,
+      "loss": 1.613,
+      "step": 10802
+    },
+    {
+      "epoch": 2.8387690084408974,
+      "grad_norm": 0.6034703254699707,
+      "learning_rate": 5.379358682319959e-06,
+      "loss": 1.5959,
+      "step": 10804
+    },
+    {
+      "epoch": 2.839294511774559,
+      "grad_norm": 0.5443819165229797,
+      "learning_rate": 5.361836341335203e-06,
+      "loss": 1.5844,
+      "step": 10806
+    },
+    {
+      "epoch": 2.839820015108221,
+      "grad_norm": 0.7041229605674744,
+      "learning_rate": 5.344314000350447e-06,
+      "loss": 1.611,
+      "step": 10808
+    },
+    {
+      "epoch": 2.8403455184418824,
+      "grad_norm": 0.5578668713569641,
+      "learning_rate": 5.326791659365692e-06,
+      "loss": 1.6177,
+      "step": 10810
+    },
+    {
+      "epoch": 2.8408710217755444,
+      "grad_norm": 0.5947692394256592,
+      "learning_rate": 5.309269318380936e-06,
+      "loss": 1.6223,
+      "step": 10812
+    },
+    {
+      "epoch": 2.8413965251092064,
+      "grad_norm": 0.6893056631088257,
+      "learning_rate": 5.29174697739618e-06,
+      "loss": 1.6284,
+      "step": 10814
+    },
+    {
+      "epoch": 2.841922028442868,
+      "grad_norm": 0.6605462431907654,
+      "learning_rate": 5.2742246364114245e-06,
+      "loss": 1.6525,
+      "step": 10816
+    },
+    {
+      "epoch": 2.84244753177653,
+      "grad_norm": 0.5912427306175232,
+      "learning_rate": 5.2567022954266695e-06,
+      "loss": 1.5903,
+      "step": 10818
+    },
+    {
+      "epoch": 2.8429730351101914,
+      "grad_norm": 0.6220155954360962,
+      "learning_rate": 5.239179954441914e-06,
+      "loss": 1.6177,
+      "step": 10820
+    },
+    {
+      "epoch": 2.8434985384438534,
+      "grad_norm": 0.5568714141845703,
+      "learning_rate": 5.221657613457158e-06,
+      "loss": 1.5975,
+      "step": 10822
+    },
+    {
+      "epoch": 2.844024041777515,
+      "grad_norm": 0.6185790300369263,
+      "learning_rate": 5.204135272472403e-06,
+      "loss": 1.6108,
+      "step": 10824
+    },
+    {
+      "epoch": 2.844549545111177,
+      "grad_norm": 0.5641743540763855,
+      "learning_rate": 5.186612931487647e-06,
+      "loss": 1.5923,
+      "step": 10826
+    },
+    {
+      "epoch": 2.845075048444839,
+      "grad_norm": 0.5651496052742004,
+      "learning_rate": 5.169090590502892e-06,
+      "loss": 1.6242,
+      "step": 10828
+    },
+    {
+      "epoch": 2.8456005517785004,
+      "grad_norm": 0.5548931360244751,
+      "learning_rate": 5.151568249518135e-06,
+      "loss": 1.5939,
+      "step": 10830
+    },
+    {
+      "epoch": 2.846126055112162,
+      "grad_norm": 0.6729891896247864,
+      "learning_rate": 5.1340459085333804e-06,
+      "loss": 1.6279,
+      "step": 10832
+    },
+    {
+      "epoch": 2.846651558445824,
+      "grad_norm": 0.603522002696991,
+      "learning_rate": 5.116523567548625e-06,
+      "loss": 1.5939,
+      "step": 10834
+    },
+    {
+      "epoch": 2.847177061779486,
+      "grad_norm": 0.5417622923851013,
+      "learning_rate": 5.099001226563869e-06,
+      "loss": 1.5809,
+      "step": 10836
+    },
+    {
+      "epoch": 2.8477025651131473,
+      "grad_norm": 0.6103736162185669,
+      "learning_rate": 5.081478885579114e-06,
+      "loss": 1.5864,
+      "step": 10838
+    },
+    {
+      "epoch": 2.8482280684468093,
+      "grad_norm": 0.6213535666465759,
+      "learning_rate": 5.063956544594358e-06,
+      "loss": 1.6082,
+      "step": 10840
+    },
+    {
+      "epoch": 2.848753571780471,
+      "grad_norm": 0.5709004998207092,
+      "learning_rate": 5.046434203609603e-06,
+      "loss": 1.5648,
+      "step": 10842
+    },
+    {
+      "epoch": 2.849279075114133,
+      "grad_norm": 0.6318028569221497,
+      "learning_rate": 5.028911862624846e-06,
+      "loss": 1.6319,
+      "step": 10844
+    },
+    {
+      "epoch": 2.8498045784477943,
+      "grad_norm": 0.5908393263816833,
+      "learning_rate": 5.011389521640091e-06,
+      "loss": 1.63,
+      "step": 10846
+    },
+    {
+      "epoch": 2.8503300817814563,
+      "grad_norm": 0.5498826503753662,
+      "learning_rate": 4.9938671806553355e-06,
+      "loss": 1.6219,
+      "step": 10848
+    },
+    {
+      "epoch": 2.8508555851151183,
+      "grad_norm": 0.5806742906570435,
+      "learning_rate": 4.9763448396705806e-06,
+      "loss": 1.574,
+      "step": 10850
+    },
+    {
+      "epoch": 2.85138108844878,
+      "grad_norm": 0.5563404560089111,
+      "learning_rate": 4.958822498685825e-06,
+      "loss": 1.6111,
+      "step": 10852
+    },
+    {
+      "epoch": 2.8519065917824418,
+      "grad_norm": 0.5694655179977417,
+      "learning_rate": 4.941300157701069e-06,
+      "loss": 1.604,
+      "step": 10854
+    },
+    {
+      "epoch": 2.8524320951161033,
+      "grad_norm": 0.5293794274330139,
+      "learning_rate": 4.923777816716314e-06,
+      "loss": 1.624,
+      "step": 10856
+    },
+    {
+      "epoch": 2.8529575984497653,
+      "grad_norm": 0.5831487774848938,
+      "learning_rate": 4.906255475731558e-06,
+      "loss": 1.6151,
+      "step": 10858
+    },
+    {
+      "epoch": 2.853483101783427,
+      "grad_norm": 0.6377184987068176,
+      "learning_rate": 4.888733134746802e-06,
+      "loss": 1.5698,
+      "step": 10860
+    },
+    {
+      "epoch": 2.8540086051170888,
+      "grad_norm": 0.6772559881210327,
+      "learning_rate": 4.8712107937620465e-06,
+      "loss": 1.5918,
+      "step": 10862
+    },
+    {
+      "epoch": 2.8545341084507507,
+      "grad_norm": 0.6782848238945007,
+      "learning_rate": 4.8536884527772915e-06,
+      "loss": 1.6303,
+      "step": 10864
+    },
+    {
+      "epoch": 2.8550596117844123,
+      "grad_norm": 0.6130311489105225,
+      "learning_rate": 4.836166111792536e-06,
+      "loss": 1.6051,
+      "step": 10866
+    },
+    {
+      "epoch": 2.855585115118074,
+      "grad_norm": 0.6167743802070618,
+      "learning_rate": 4.81864377080778e-06,
+      "loss": 1.6214,
+      "step": 10868
+    },
+    {
+      "epoch": 2.8561106184517357,
+      "grad_norm": 0.5957754850387573,
+      "learning_rate": 4.801121429823025e-06,
+      "loss": 1.6118,
+      "step": 10870
+    },
+    {
+      "epoch": 2.8566361217853977,
+      "grad_norm": 0.5984094738960266,
+      "learning_rate": 4.783599088838269e-06,
+      "loss": 1.5894,
+      "step": 10872
+    },
+    {
+      "epoch": 2.8571616251190592,
+      "grad_norm": 0.6340751051902771,
+      "learning_rate": 4.766076747853514e-06,
+      "loss": 1.6018,
+      "step": 10874
+    },
+    {
+      "epoch": 2.857687128452721,
+      "grad_norm": 0.6450504660606384,
+      "learning_rate": 4.748554406868757e-06,
+      "loss": 1.6009,
+      "step": 10876
+    },
+    {
+      "epoch": 2.8582126317863827,
+      "grad_norm": 0.5570113062858582,
+      "learning_rate": 4.731032065884002e-06,
+      "loss": 1.5877,
+      "step": 10878
+    },
+    {
+      "epoch": 2.8587381351200447,
+      "grad_norm": 0.6482218503952026,
+      "learning_rate": 4.7135097248992466e-06,
+      "loss": 1.6192,
+      "step": 10880
+    },
+    {
+      "epoch": 2.8592636384537062,
+      "grad_norm": 0.5896407961845398,
+      "learning_rate": 4.695987383914492e-06,
+      "loss": 1.5885,
+      "step": 10882
+    },
+    {
+      "epoch": 2.859789141787368,
+      "grad_norm": 0.6039836406707764,
+      "learning_rate": 4.678465042929736e-06,
+      "loss": 1.6141,
+      "step": 10884
+    },
+    {
+      "epoch": 2.86031464512103,
+      "grad_norm": 0.6409982442855835,
+      "learning_rate": 4.66094270194498e-06,
+      "loss": 1.6277,
+      "step": 10886
+    },
+    {
+      "epoch": 2.8608401484546917,
+      "grad_norm": 0.5397644639015198,
+      "learning_rate": 4.643420360960225e-06,
+      "loss": 1.6035,
+      "step": 10888
+    },
+    {
+      "epoch": 2.8613656517883537,
+      "grad_norm": 0.587805986404419,
+      "learning_rate": 4.625898019975469e-06,
+      "loss": 1.578,
+      "step": 10890
+    },
+    {
+      "epoch": 2.861891155122015,
+      "grad_norm": 0.6628287434577942,
+      "learning_rate": 4.608375678990713e-06,
+      "loss": 1.6003,
+      "step": 10892
+    },
+    {
+      "epoch": 2.862416658455677,
+      "grad_norm": 0.5740228891372681,
+      "learning_rate": 4.5908533380059575e-06,
+      "loss": 1.6358,
+      "step": 10894
+    },
+    {
+      "epoch": 2.8629421617893387,
+      "grad_norm": 0.6140313148498535,
+      "learning_rate": 4.5733309970212025e-06,
+      "loss": 1.5789,
+      "step": 10896
+    },
+    {
+      "epoch": 2.8634676651230007,
+      "grad_norm": 0.6435587406158447,
+      "learning_rate": 4.555808656036447e-06,
+      "loss": 1.6386,
+      "step": 10898
+    },
+    {
+      "epoch": 2.8639931684566626,
+      "grad_norm": 0.6358049511909485,
+      "learning_rate": 4.538286315051691e-06,
+      "loss": 1.6128,
+      "step": 10900
+    },
+    {
+      "epoch": 2.864518671790324,
+      "grad_norm": 0.5897725224494934,
+      "learning_rate": 4.520763974066936e-06,
+      "loss": 1.5958,
+      "step": 10902
+    },
+    {
+      "epoch": 2.8650441751239857,
+      "grad_norm": 0.6232768893241882,
+      "learning_rate": 4.50324163308218e-06,
+      "loss": 1.6026,
+      "step": 10904
+    },
+    {
+      "epoch": 2.8655696784576477,
+      "grad_norm": 0.5649431347846985,
+      "learning_rate": 4.485719292097425e-06,
+      "loss": 1.6001,
+      "step": 10906
+    },
+    {
+      "epoch": 2.8660951817913096,
+      "grad_norm": 0.5962793827056885,
+      "learning_rate": 4.468196951112668e-06,
+      "loss": 1.5408,
+      "step": 10908
+    },
+    {
+      "epoch": 2.866620685124971,
+      "grad_norm": 0.5510377883911133,
+      "learning_rate": 4.4506746101279134e-06,
+      "loss": 1.6295,
+      "step": 10910
+    },
+    {
+      "epoch": 2.867146188458633,
+      "grad_norm": 0.5945908427238464,
+      "learning_rate": 4.433152269143158e-06,
+      "loss": 1.6192,
+      "step": 10912
+    },
+    {
+      "epoch": 2.8676716917922946,
+      "grad_norm": 0.5374131798744202,
+      "learning_rate": 4.415629928158403e-06,
+      "loss": 1.6022,
+      "step": 10914
+    },
+    {
+      "epoch": 2.8681971951259566,
+      "grad_norm": 0.5591030716896057,
+      "learning_rate": 4.398107587173647e-06,
+      "loss": 1.6104,
+      "step": 10916
+    },
+    {
+      "epoch": 2.868722698459618,
+      "grad_norm": 0.6745641231536865,
+      "learning_rate": 4.380585246188891e-06,
+      "loss": 1.5931,
+      "step": 10918
+    },
+    {
+      "epoch": 2.86924820179328,
+      "grad_norm": 0.7577717900276184,
+      "learning_rate": 4.363062905204136e-06,
+      "loss": 1.6324,
+      "step": 10920
+    },
+    {
+      "epoch": 2.869773705126942,
+      "grad_norm": 0.7548211812973022,
+      "learning_rate": 4.345540564219379e-06,
+      "loss": 1.62,
+      "step": 10922
+    },
+    {
+      "epoch": 2.8702992084606036,
+      "grad_norm": 0.6899915337562561,
+      "learning_rate": 4.328018223234624e-06,
+      "loss": 1.5757,
+      "step": 10924
+    },
+    {
+      "epoch": 2.8708247117942656,
+      "grad_norm": 0.580837070941925,
+      "learning_rate": 4.3104958822498685e-06,
+      "loss": 1.5889,
+      "step": 10926
+    },
+    {
+      "epoch": 2.871350215127927,
+      "grad_norm": 0.6023684740066528,
+      "learning_rate": 4.2929735412651136e-06,
+      "loss": 1.5814,
+      "step": 10928
+    },
+    {
+      "epoch": 2.871875718461589,
+      "grad_norm": 0.63701331615448,
+      "learning_rate": 4.275451200280358e-06,
+      "loss": 1.6205,
+      "step": 10930
+    },
+    {
+      "epoch": 2.8724012217952506,
+      "grad_norm": 0.5388381481170654,
+      "learning_rate": 4.257928859295602e-06,
+      "loss": 1.6093,
+      "step": 10932
+    },
+    {
+      "epoch": 2.8729267251289126,
+      "grad_norm": 0.5374415516853333,
+      "learning_rate": 4.240406518310847e-06,
+      "loss": 1.6366,
+      "step": 10934
+    },
+    {
+      "epoch": 2.8734522284625745,
+      "grad_norm": 0.7570417523384094,
+      "learning_rate": 4.222884177326091e-06,
+      "loss": 1.6263,
+      "step": 10936
+    },
+    {
+      "epoch": 2.873977731796236,
+      "grad_norm": 0.6314011812210083,
+      "learning_rate": 4.205361836341335e-06,
+      "loss": 1.5977,
+      "step": 10938
+    },
+    {
+      "epoch": 2.8745032351298976,
+      "grad_norm": 0.649953305721283,
+      "learning_rate": 4.1878394953565794e-06,
+      "loss": 1.6034,
+      "step": 10940
+    },
+    {
+      "epoch": 2.8750287384635596,
+      "grad_norm": 0.6486716866493225,
+      "learning_rate": 4.1703171543718245e-06,
+      "loss": 1.5938,
+      "step": 10942
+    },
+    {
+      "epoch": 2.8755542417972215,
+      "grad_norm": 0.6346645951271057,
+      "learning_rate": 4.152794813387069e-06,
+      "loss": 1.5878,
+      "step": 10944
+    },
+    {
+      "epoch": 2.876079745130883,
+      "grad_norm": 0.624394953250885,
+      "learning_rate": 4.135272472402313e-06,
+      "loss": 1.6165,
+      "step": 10946
+    },
+    {
+      "epoch": 2.876605248464545,
+      "grad_norm": 0.5654635429382324,
+      "learning_rate": 4.117750131417558e-06,
+      "loss": 1.5565,
+      "step": 10948
+    },
+    {
+      "epoch": 2.877130751798207,
+      "grad_norm": 0.6420629024505615,
+      "learning_rate": 4.100227790432802e-06,
+      "loss": 1.6144,
+      "step": 10950
+    },
+    {
+      "epoch": 2.8776562551318685,
+      "grad_norm": 0.6588164567947388,
+      "learning_rate": 4.082705449448047e-06,
+      "loss": 1.6324,
+      "step": 10952
+    },
+    {
+      "epoch": 2.87818175846553,
+      "grad_norm": 0.5764179229736328,
+      "learning_rate": 4.06518310846329e-06,
+      "loss": 1.608,
+      "step": 10954
+    },
+    {
+      "epoch": 2.878707261799192,
+      "grad_norm": 0.5966598391532898,
+      "learning_rate": 4.047660767478535e-06,
+      "loss": 1.5683,
+      "step": 10956
+    },
+    {
+      "epoch": 2.879232765132854,
+      "grad_norm": 0.6094986796379089,
+      "learning_rate": 4.0301384264937796e-06,
+      "loss": 1.6101,
+      "step": 10958
+    },
+    {
+      "epoch": 2.8797582684665155,
+      "grad_norm": 0.6777652502059937,
+      "learning_rate": 4.012616085509025e-06,
+      "loss": 1.6454,
+      "step": 10960
+    },
+    {
+      "epoch": 2.8802837718001775,
+      "grad_norm": 0.54978346824646,
+      "learning_rate": 3.995093744524269e-06,
+      "loss": 1.5811,
+      "step": 10962
+    },
+    {
+      "epoch": 2.880809275133839,
+      "grad_norm": 0.7338440418243408,
+      "learning_rate": 3.977571403539513e-06,
+      "loss": 1.6004,
+      "step": 10964
+    },
+    {
+      "epoch": 2.881334778467501,
+      "grad_norm": 0.5676441192626953,
+      "learning_rate": 3.960049062554758e-06,
+      "loss": 1.6194,
+      "step": 10966
+    },
+    {
+      "epoch": 2.8818602818011625,
+      "grad_norm": 0.6784756183624268,
+      "learning_rate": 3.942526721570002e-06,
+      "loss": 1.5941,
+      "step": 10968
+    },
+    {
+      "epoch": 2.8823857851348245,
+      "grad_norm": 0.6024194955825806,
+      "learning_rate": 3.925004380585246e-06,
+      "loss": 1.6035,
+      "step": 10970
+    },
+    {
+      "epoch": 2.8829112884684864,
+      "grad_norm": 0.5432921051979065,
+      "learning_rate": 3.9074820396004905e-06,
+      "loss": 1.6307,
+      "step": 10972
+    },
+    {
+      "epoch": 2.883436791802148,
+      "grad_norm": 0.5449672937393188,
+      "learning_rate": 3.8899596986157355e-06,
+      "loss": 1.6078,
+      "step": 10974
+    },
+    {
+      "epoch": 2.88396229513581,
+      "grad_norm": 0.582300066947937,
+      "learning_rate": 3.87243735763098e-06,
+      "loss": 1.5869,
+      "step": 10976
+    },
+    {
+      "epoch": 2.8844877984694715,
+      "grad_norm": 0.5711895823478699,
+      "learning_rate": 3.854915016646224e-06,
+      "loss": 1.578,
+      "step": 10978
+    },
+    {
+      "epoch": 2.8850133018031334,
+      "grad_norm": 0.6558634638786316,
+      "learning_rate": 3.837392675661469e-06,
+      "loss": 1.6074,
+      "step": 10980
+    },
+    {
+      "epoch": 2.885538805136795,
+      "grad_norm": 0.6054571866989136,
+      "learning_rate": 3.819870334676713e-06,
+      "loss": 1.6015,
+      "step": 10982
+    },
+    {
+      "epoch": 2.886064308470457,
+      "grad_norm": 0.5859559178352356,
+      "learning_rate": 3.8023479936919577e-06,
+      "loss": 1.575,
+      "step": 10984
+    },
+    {
+      "epoch": 2.886589811804119,
+      "grad_norm": 0.7558565139770508,
+      "learning_rate": 3.784825652707202e-06,
+      "loss": 1.6048,
+      "step": 10986
+    },
+    {
+      "epoch": 2.8871153151377804,
+      "grad_norm": 0.5738922357559204,
+      "learning_rate": 3.7673033117224464e-06,
+      "loss": 1.5895,
+      "step": 10988
+    },
+    {
+      "epoch": 2.887640818471442,
+      "grad_norm": 0.6217196583747864,
+      "learning_rate": 3.749780970737691e-06,
+      "loss": 1.6017,
+      "step": 10990
+    },
+    {
+      "epoch": 2.888166321805104,
+      "grad_norm": 0.590238094329834,
+      "learning_rate": 3.7322586297529356e-06,
+      "loss": 1.5937,
+      "step": 10992
+    },
+    {
+      "epoch": 2.888691825138766,
+      "grad_norm": 0.5736632943153381,
+      "learning_rate": 3.7147362887681794e-06,
+      "loss": 1.6011,
+      "step": 10994
+    },
+    {
+      "epoch": 2.8892173284724274,
+      "grad_norm": 0.6404598951339722,
+      "learning_rate": 3.697213947783424e-06,
+      "loss": 1.6158,
+      "step": 10996
+    },
+    {
+      "epoch": 2.8897428318060894,
+      "grad_norm": 0.5618501305580139,
+      "learning_rate": 3.6796916067986686e-06,
+      "loss": 1.5836,
+      "step": 10998
+    },
+    {
+      "epoch": 2.890268335139751,
+      "grad_norm": 0.5837007761001587,
+      "learning_rate": 3.662169265813913e-06,
+      "loss": 1.6126,
+      "step": 11000
+    },
+    {
+      "epoch": 2.890793838473413,
+      "grad_norm": 0.6110183596611023,
+      "learning_rate": 3.6446469248291574e-06,
+      "loss": 1.5715,
+      "step": 11002
+    },
+    {
+      "epoch": 2.8913193418070744,
+      "grad_norm": 0.5909767746925354,
+      "learning_rate": 3.627124583844402e-06,
+      "loss": 1.5992,
+      "step": 11004
+    },
+    {
+      "epoch": 2.8918448451407364,
+      "grad_norm": 0.5684041976928711,
+      "learning_rate": 3.6096022428596465e-06,
+      "loss": 1.6002,
+      "step": 11006
+    },
+    {
+      "epoch": 2.8923703484743983,
+      "grad_norm": 0.5662478804588318,
+      "learning_rate": 3.5920799018748903e-06,
+      "loss": 1.6093,
+      "step": 11008
+    },
+    {
+      "epoch": 2.89289585180806,
+      "grad_norm": 0.5797233581542969,
+      "learning_rate": 3.574557560890135e-06,
+      "loss": 1.5847,
+      "step": 11010
+    },
+    {
+      "epoch": 2.893421355141722,
+      "grad_norm": 0.60663902759552,
+      "learning_rate": 3.5570352199053795e-06,
+      "loss": 1.5927,
+      "step": 11012
+    },
+    {
+      "epoch": 2.8939468584753834,
+      "grad_norm": 0.5697915554046631,
+      "learning_rate": 3.539512878920624e-06,
+      "loss": 1.5718,
+      "step": 11014
+    },
+    {
+      "epoch": 2.8944723618090453,
+      "grad_norm": 0.6556651592254639,
+      "learning_rate": 3.5219905379358683e-06,
+      "loss": 1.6116,
+      "step": 11016
+    },
+    {
+      "epoch": 2.894997865142707,
+      "grad_norm": 0.6983522176742554,
+      "learning_rate": 3.504468196951113e-06,
+      "loss": 1.6157,
+      "step": 11018
+    },
+    {
+      "epoch": 2.895523368476369,
+      "grad_norm": 0.5559704303741455,
+      "learning_rate": 3.4869458559663575e-06,
+      "loss": 1.5646,
+      "step": 11020
+    },
+    {
+      "epoch": 2.896048871810031,
+      "grad_norm": 0.5538905262947083,
+      "learning_rate": 3.469423514981602e-06,
+      "loss": 1.6231,
+      "step": 11022
+    },
+    {
+      "epoch": 2.8965743751436923,
+      "grad_norm": 0.5700134038925171,
+      "learning_rate": 3.451901173996846e-06,
+      "loss": 1.5936,
+      "step": 11024
+    },
+    {
+      "epoch": 2.897099878477354,
+      "grad_norm": 0.5542116761207581,
+      "learning_rate": 3.4343788330120904e-06,
+      "loss": 1.6293,
+      "step": 11026
+    },
+    {
+      "epoch": 2.897625381811016,
+      "grad_norm": 0.5474702715873718,
+      "learning_rate": 3.416856492027335e-06,
+      "loss": 1.6112,
+      "step": 11028
+    },
+    {
+      "epoch": 2.898150885144678,
+      "grad_norm": 0.6167372465133667,
+      "learning_rate": 3.3993341510425796e-06,
+      "loss": 1.5763,
+      "step": 11030
+    },
+    {
+      "epoch": 2.8986763884783393,
+      "grad_norm": 0.6581651568412781,
+      "learning_rate": 3.381811810057824e-06,
+      "loss": 1.6004,
+      "step": 11032
+    },
+    {
+      "epoch": 2.8992018918120013,
+      "grad_norm": 0.62739497423172,
+      "learning_rate": 3.3642894690730684e-06,
+      "loss": 1.6136,
+      "step": 11034
+    },
+    {
+      "epoch": 2.899727395145663,
+      "grad_norm": 0.5928983688354492,
+      "learning_rate": 3.346767128088313e-06,
+      "loss": 1.622,
+      "step": 11036
+    },
+    {
+      "epoch": 2.9002528984793248,
+      "grad_norm": 0.6407076716423035,
+      "learning_rate": 3.3292447871035576e-06,
+      "loss": 1.5815,
+      "step": 11038
+    },
+    {
+      "epoch": 2.9007784018129863,
+      "grad_norm": 0.6269816756248474,
+      "learning_rate": 3.3117224461188013e-06,
+      "loss": 1.6119,
+      "step": 11040
+    },
+    {
+      "epoch": 2.9013039051466483,
+      "grad_norm": 0.5729198455810547,
+      "learning_rate": 3.294200105134046e-06,
+      "loss": 1.6173,
+      "step": 11042
+    },
+    {
+      "epoch": 2.9018294084803102,
+      "grad_norm": 0.5712758302688599,
+      "learning_rate": 3.2766777641492905e-06,
+      "loss": 1.5912,
+      "step": 11044
+    },
+    {
+      "epoch": 2.9023549118139718,
+      "grad_norm": 0.6735134720802307,
+      "learning_rate": 3.259155423164535e-06,
+      "loss": 1.603,
+      "step": 11046
+    },
+    {
+      "epoch": 2.9028804151476337,
+      "grad_norm": 0.6073388457298279,
+      "learning_rate": 3.2416330821797793e-06,
+      "loss": 1.6013,
+      "step": 11048
+    },
+    {
+      "epoch": 2.9034059184812953,
+      "grad_norm": 0.611774742603302,
+      "learning_rate": 3.224110741195024e-06,
+      "loss": 1.6021,
+      "step": 11050
+    },
+    {
+      "epoch": 2.9039314218149572,
+      "grad_norm": 0.6439489722251892,
+      "learning_rate": 3.2065884002102685e-06,
+      "loss": 1.6037,
+      "step": 11052
+    },
+    {
+      "epoch": 2.9044569251486188,
+      "grad_norm": 0.572325587272644,
+      "learning_rate": 3.189066059225513e-06,
+      "loss": 1.5629,
+      "step": 11054
+    },
+    {
+      "epoch": 2.9049824284822807,
+      "grad_norm": 0.5802271962165833,
+      "learning_rate": 3.171543718240757e-06,
+      "loss": 1.5982,
+      "step": 11056
+    },
+    {
+      "epoch": 2.9055079318159427,
+      "grad_norm": 0.6686844825744629,
+      "learning_rate": 3.1540213772560015e-06,
+      "loss": 1.626,
+      "step": 11058
+    },
+    {
+      "epoch": 2.906033435149604,
+      "grad_norm": 0.5929732918739319,
+      "learning_rate": 3.136499036271246e-06,
+      "loss": 1.5946,
+      "step": 11060
+    },
+    {
+      "epoch": 2.9065589384832657,
+      "grad_norm": 0.5740990042686462,
+      "learning_rate": 3.1189766952864902e-06,
+      "loss": 1.6365,
+      "step": 11062
+    },
+    {
+      "epoch": 2.9070844418169277,
+      "grad_norm": 0.6125746369361877,
+      "learning_rate": 3.101454354301735e-06,
+      "loss": 1.5778,
+      "step": 11064
+    },
+    {
+      "epoch": 2.9076099451505897,
+      "grad_norm": 0.5508874654769897,
+      "learning_rate": 3.0839320133169794e-06,
+      "loss": 1.6019,
+      "step": 11066
+    },
+    {
+      "epoch": 2.908135448484251,
+      "grad_norm": 0.5743160247802734,
+      "learning_rate": 3.066409672332224e-06,
+      "loss": 1.6103,
+      "step": 11068
+    },
+    {
+      "epoch": 2.908660951817913,
+      "grad_norm": 0.6552396416664124,
+      "learning_rate": 3.048887331347468e-06,
+      "loss": 1.6245,
+      "step": 11070
+    },
+    {
+      "epoch": 2.9091864551515747,
+      "grad_norm": 0.5878915190696716,
+      "learning_rate": 3.031364990362713e-06,
+      "loss": 1.6111,
+      "step": 11072
+    },
+    {
+      "epoch": 2.9097119584852367,
+      "grad_norm": 0.5548945665359497,
+      "learning_rate": 3.013842649377957e-06,
+      "loss": 1.6289,
+      "step": 11074
+    },
+    {
+      "epoch": 2.910237461818898,
+      "grad_norm": 0.5816505551338196,
+      "learning_rate": 2.9963203083932016e-06,
+      "loss": 1.6,
+      "step": 11076
+    },
+    {
+      "epoch": 2.91076296515256,
+      "grad_norm": 0.5567723512649536,
+      "learning_rate": 2.9787979674084457e-06,
+      "loss": 1.6144,
+      "step": 11078
+    },
+    {
+      "epoch": 2.911288468486222,
+      "grad_norm": 0.542295515537262,
+      "learning_rate": 2.9612756264236903e-06,
+      "loss": 1.6051,
+      "step": 11080
+    },
+    {
+      "epoch": 2.9118139718198837,
+      "grad_norm": 0.6353909969329834,
+      "learning_rate": 2.943753285438935e-06,
+      "loss": 1.6135,
+      "step": 11082
+    },
+    {
+      "epoch": 2.9123394751535456,
+      "grad_norm": 0.5646031498908997,
+      "learning_rate": 2.9262309444541795e-06,
+      "loss": 1.5934,
+      "step": 11084
+    },
+    {
+      "epoch": 2.912864978487207,
+      "grad_norm": 0.5628175139427185,
+      "learning_rate": 2.9087086034694237e-06,
+      "loss": 1.5589,
+      "step": 11086
+    },
+    {
+      "epoch": 2.913390481820869,
+      "grad_norm": 0.6169978380203247,
+      "learning_rate": 2.891186262484668e-06,
+      "loss": 1.5969,
+      "step": 11088
+    },
+    {
+      "epoch": 2.9139159851545307,
+      "grad_norm": 0.5552890300750732,
+      "learning_rate": 2.8736639214999125e-06,
+      "loss": 1.6083,
+      "step": 11090
+    },
+    {
+      "epoch": 2.9144414884881926,
+      "grad_norm": 0.5658928751945496,
+      "learning_rate": 2.8561415805151567e-06,
+      "loss": 1.5807,
+      "step": 11092
+    },
+    {
+      "epoch": 2.9149669918218546,
+      "grad_norm": 0.5916785001754761,
+      "learning_rate": 2.8386192395304013e-06,
+      "loss": 1.6128,
+      "step": 11094
+    },
+    {
+      "epoch": 2.915492495155516,
+      "grad_norm": 0.5797626376152039,
+      "learning_rate": 2.821096898545646e-06,
+      "loss": 1.6042,
+      "step": 11096
+    },
+    {
+      "epoch": 2.9160179984891776,
+      "grad_norm": 0.661906361579895,
+      "learning_rate": 2.8035745575608905e-06,
+      "loss": 1.6357,
+      "step": 11098
+    },
+    {
+      "epoch": 2.9165435018228396,
+      "grad_norm": 0.5823834538459778,
+      "learning_rate": 2.7860522165761346e-06,
+      "loss": 1.5987,
+      "step": 11100
+    },
+    {
+      "epoch": 2.9170690051565016,
+      "grad_norm": 0.5549127459526062,
+      "learning_rate": 2.7685298755913792e-06,
+      "loss": 1.6271,
+      "step": 11102
+    },
+    {
+      "epoch": 2.917594508490163,
+      "grad_norm": 0.633703351020813,
+      "learning_rate": 2.7510075346066234e-06,
+      "loss": 1.6639,
+      "step": 11104
+    },
+    {
+      "epoch": 2.918120011823825,
+      "grad_norm": 0.6295391321182251,
+      "learning_rate": 2.733485193621868e-06,
+      "loss": 1.5803,
+      "step": 11106
+    },
+    {
+      "epoch": 2.918645515157487,
+      "grad_norm": 0.7048977017402649,
+      "learning_rate": 2.715962852637112e-06,
+      "loss": 1.6112,
+      "step": 11108
+    },
+    {
+      "epoch": 2.9191710184911486,
+      "grad_norm": 0.7996454238891602,
+      "learning_rate": 2.6984405116523568e-06,
+      "loss": 1.5931,
+      "step": 11110
+    },
+    {
+      "epoch": 2.91969652182481,
+      "grad_norm": 0.6191883683204651,
+      "learning_rate": 2.6809181706676014e-06,
+      "loss": 1.6265,
+      "step": 11112
+    },
+    {
+      "epoch": 2.920222025158472,
+      "grad_norm": 0.6032472252845764,
+      "learning_rate": 2.663395829682846e-06,
+      "loss": 1.6078,
+      "step": 11114
+    },
+    {
+      "epoch": 2.920747528492134,
+      "grad_norm": 0.5909383296966553,
+      "learning_rate": 2.64587348869809e-06,
+      "loss": 1.6047,
+      "step": 11116
+    },
+    {
+      "epoch": 2.9212730318257956,
+      "grad_norm": 0.741261899471283,
+      "learning_rate": 2.6283511477133348e-06,
+      "loss": 1.5909,
+      "step": 11118
+    },
+    {
+      "epoch": 2.9217985351594575,
+      "grad_norm": 0.5791400074958801,
+      "learning_rate": 2.610828806728579e-06,
+      "loss": 1.6037,
+      "step": 11120
+    },
+    {
+      "epoch": 2.922324038493119,
+      "grad_norm": 0.5476728677749634,
+      "learning_rate": 2.5933064657438235e-06,
+      "loss": 1.5841,
+      "step": 11122
+    },
+    {
+      "epoch": 2.922849541826781,
+      "grad_norm": 0.6799639463424683,
+      "learning_rate": 2.5757841247590677e-06,
+      "loss": 1.6055,
+      "step": 11124
+    },
+    {
+      "epoch": 2.9233750451604426,
+      "grad_norm": 0.6471995115280151,
+      "learning_rate": 2.5582617837743123e-06,
+      "loss": 1.614,
+      "step": 11126
+    },
+    {
+      "epoch": 2.9239005484941045,
+      "grad_norm": 0.5741569399833679,
+      "learning_rate": 2.540739442789557e-06,
+      "loss": 1.5667,
+      "step": 11128
+    },
+    {
+      "epoch": 2.9244260518277665,
+      "grad_norm": 0.5387163162231445,
+      "learning_rate": 2.5232171018048015e-06,
+      "loss": 1.617,
+      "step": 11130
+    },
+    {
+      "epoch": 2.924951555161428,
+      "grad_norm": 0.5384705066680908,
+      "learning_rate": 2.5056947608200457e-06,
+      "loss": 1.5939,
+      "step": 11132
+    },
+    {
+      "epoch": 2.92547705849509,
+      "grad_norm": 0.5429417490959167,
+      "learning_rate": 2.4881724198352903e-06,
+      "loss": 1.5842,
+      "step": 11134
+    },
+    {
+      "epoch": 2.9260025618287515,
+      "grad_norm": 0.6191485524177551,
+      "learning_rate": 2.4706500788505345e-06,
+      "loss": 1.6338,
+      "step": 11136
+    },
+    {
+      "epoch": 2.9265280651624135,
+      "grad_norm": 0.7055456042289734,
+      "learning_rate": 2.453127737865779e-06,
+      "loss": 1.6349,
+      "step": 11138
+    },
+    {
+      "epoch": 2.927053568496075,
+      "grad_norm": 0.5733761787414551,
+      "learning_rate": 2.4356053968810232e-06,
+      "loss": 1.5677,
+      "step": 11140
+    },
+    {
+      "epoch": 2.927579071829737,
+      "grad_norm": 0.6309702396392822,
+      "learning_rate": 2.418083055896268e-06,
+      "loss": 1.577,
+      "step": 11142
+    },
+    {
+      "epoch": 2.928104575163399,
+      "grad_norm": 0.56844162940979,
+      "learning_rate": 2.4005607149115124e-06,
+      "loss": 1.6017,
+      "step": 11144
+    },
+    {
+      "epoch": 2.9286300784970605,
+      "grad_norm": 0.5193071365356445,
+      "learning_rate": 2.383038373926757e-06,
+      "loss": 1.5855,
+      "step": 11146
+    },
+    {
+      "epoch": 2.929155581830722,
+      "grad_norm": 0.6680015325546265,
+      "learning_rate": 2.365516032942001e-06,
+      "loss": 1.6215,
+      "step": 11148
+    },
+    {
+      "epoch": 2.929681085164384,
+      "grad_norm": 0.6055448055267334,
+      "learning_rate": 2.347993691957246e-06,
+      "loss": 1.5748,
+      "step": 11150
+    },
+    {
+      "epoch": 2.930206588498046,
+      "grad_norm": 0.763596773147583,
+      "learning_rate": 2.33047135097249e-06,
+      "loss": 1.567,
+      "step": 11152
+    },
+    {
+      "epoch": 2.9307320918317075,
+      "grad_norm": 0.5230932235717773,
+      "learning_rate": 2.3129490099877346e-06,
+      "loss": 1.6069,
+      "step": 11154
+    },
+    {
+      "epoch": 2.9312575951653694,
+      "grad_norm": 0.5715333819389343,
+      "learning_rate": 2.2954266690029787e-06,
+      "loss": 1.5798,
+      "step": 11156
+    },
+    {
+      "epoch": 2.931783098499031,
+      "grad_norm": 0.5592380166053772,
+      "learning_rate": 2.2779043280182233e-06,
+      "loss": 1.616,
+      "step": 11158
+    },
+    {
+      "epoch": 2.932308601832693,
+      "grad_norm": 0.5946754813194275,
+      "learning_rate": 2.260381987033468e-06,
+      "loss": 1.5689,
+      "step": 11160
+    },
+    {
+      "epoch": 2.9328341051663545,
+      "grad_norm": 0.6536034345626831,
+      "learning_rate": 2.2428596460487125e-06,
+      "loss": 1.5949,
+      "step": 11162
+    },
+    {
+      "epoch": 2.9333596085000164,
+      "grad_norm": 0.5874665975570679,
+      "learning_rate": 2.2253373050639567e-06,
+      "loss": 1.6125,
+      "step": 11164
+    },
+    {
+      "epoch": 2.9338851118336784,
+      "grad_norm": 0.5944022536277771,
+      "learning_rate": 2.2078149640792013e-06,
+      "loss": 1.5724,
+      "step": 11166
+    },
+    {
+      "epoch": 2.93441061516734,
+      "grad_norm": 0.5689641237258911,
+      "learning_rate": 2.1902926230944455e-06,
+      "loss": 1.587,
+      "step": 11168
+    },
+    {
+      "epoch": 2.934936118501002,
+      "grad_norm": 0.6851060390472412,
+      "learning_rate": 2.1727702821096897e-06,
+      "loss": 1.5942,
+      "step": 11170
+    },
+    {
+      "epoch": 2.9354616218346634,
+      "grad_norm": 0.6660142540931702,
+      "learning_rate": 2.1552479411249343e-06,
+      "loss": 1.5892,
+      "step": 11172
+    },
+    {
+      "epoch": 2.9359871251683254,
+      "grad_norm": 0.6187747716903687,
+      "learning_rate": 2.137725600140179e-06,
+      "loss": 1.5622,
+      "step": 11174
+    },
+    {
+      "epoch": 2.936512628501987,
+      "grad_norm": 0.5741645097732544,
+      "learning_rate": 2.1202032591554235e-06,
+      "loss": 1.5856,
+      "step": 11176
+    },
+    {
+      "epoch": 2.937038131835649,
+      "grad_norm": 0.5658064484596252,
+      "learning_rate": 2.1026809181706676e-06,
+      "loss": 1.5984,
+      "step": 11178
+    },
+    {
+      "epoch": 2.937563635169311,
+      "grad_norm": 0.6636450886726379,
+      "learning_rate": 2.0851585771859122e-06,
+      "loss": 1.6533,
+      "step": 11180
+    },
+    {
+      "epoch": 2.9380891385029724,
+      "grad_norm": 0.5776405334472656,
+      "learning_rate": 2.0676362362011564e-06,
+      "loss": 1.567,
+      "step": 11182
+    },
+    {
+      "epoch": 2.938614641836634,
+      "grad_norm": 0.6802220940589905,
+      "learning_rate": 2.050113895216401e-06,
+      "loss": 1.6211,
+      "step": 11184
+    },
+    {
+      "epoch": 2.939140145170296,
+      "grad_norm": 0.6535943746566772,
+      "learning_rate": 2.032591554231645e-06,
+      "loss": 1.596,
+      "step": 11186
+    },
+    {
+      "epoch": 2.939665648503958,
+      "grad_norm": 0.5814671516418457,
+      "learning_rate": 2.0150692132468898e-06,
+      "loss": 1.5689,
+      "step": 11188
+    },
+    {
+      "epoch": 2.9401911518376194,
+      "grad_norm": 0.6042841076850891,
+      "learning_rate": 1.9975468722621344e-06,
+      "loss": 1.6142,
+      "step": 11190
+    },
+    {
+      "epoch": 2.9407166551712813,
+      "grad_norm": 0.6148901581764221,
+      "learning_rate": 1.980024531277379e-06,
+      "loss": 1.5813,
+      "step": 11192
+    },
+    {
+      "epoch": 2.941242158504943,
+      "grad_norm": 0.6761669516563416,
+      "learning_rate": 1.962502190292623e-06,
+      "loss": 1.5915,
+      "step": 11194
+    },
+    {
+      "epoch": 2.941767661838605,
+      "grad_norm": 0.6184157133102417,
+      "learning_rate": 1.9449798493078678e-06,
+      "loss": 1.6291,
+      "step": 11196
+    },
+    {
+      "epoch": 2.9422931651722664,
+      "grad_norm": 0.567001223564148,
+      "learning_rate": 1.927457508323112e-06,
+      "loss": 1.5796,
+      "step": 11198
+    },
+    {
+      "epoch": 2.9428186685059283,
+      "grad_norm": 0.5846920609474182,
+      "learning_rate": 1.9099351673383565e-06,
+      "loss": 1.6186,
+      "step": 11200
+    },
+    {
+      "epoch": 2.9428186685059283,
+      "eval_loss": 1.6465901136398315,
+      "eval_runtime": 487.2703,
+      "eval_samples_per_second": 249.941,
+      "eval_steps_per_second": 31.243,
+      "step": 11200
+    },
+    {
+      "epoch": 2.9433441718395903,
+      "grad_norm": 0.5758314728736877,
+      "learning_rate": 1.892412826353601e-06,
+      "loss": 1.609,
+      "step": 11202
+    },
+    {
+      "epoch": 2.943869675173252,
+      "grad_norm": 0.6517518758773804,
+      "learning_rate": 1.8748904853688455e-06,
+      "loss": 1.614,
+      "step": 11204
+    },
+    {
+      "epoch": 2.944395178506914,
+      "grad_norm": 0.5758547186851501,
+      "learning_rate": 1.8573681443840897e-06,
+      "loss": 1.5874,
+      "step": 11206
+    },
+    {
+      "epoch": 2.9449206818405753,
+      "grad_norm": 0.5640631914138794,
+      "learning_rate": 1.8398458033993343e-06,
+      "loss": 1.5723,
+      "step": 11208
+    },
+    {
+      "epoch": 2.9454461851742373,
+      "grad_norm": 0.6348847150802612,
+      "learning_rate": 1.8223234624145787e-06,
+      "loss": 1.5789,
+      "step": 11210
+    },
+    {
+      "epoch": 2.945971688507899,
+      "grad_norm": 0.6184114217758179,
+      "learning_rate": 1.8048011214298233e-06,
+      "loss": 1.5986,
+      "step": 11212
+    },
+    {
+      "epoch": 2.946497191841561,
+      "grad_norm": 0.8769078850746155,
+      "learning_rate": 1.7872787804450674e-06,
+      "loss": 1.6315,
+      "step": 11214
+    },
+    {
+      "epoch": 2.9470226951752228,
+      "grad_norm": 0.6675733923912048,
+      "learning_rate": 1.769756439460312e-06,
+      "loss": 1.6477,
+      "step": 11216
+    },
+    {
+      "epoch": 2.9475481985088843,
+      "grad_norm": 0.5792946219444275,
+      "learning_rate": 1.7522340984755564e-06,
+      "loss": 1.6266,
+      "step": 11218
+    },
+    {
+      "epoch": 2.948073701842546,
+      "grad_norm": 0.6874922513961792,
+      "learning_rate": 1.734711757490801e-06,
+      "loss": 1.6014,
+      "step": 11220
+    },
+    {
+      "epoch": 2.9485992051762078,
+      "grad_norm": 0.6224451065063477,
+      "learning_rate": 1.7171894165060452e-06,
+      "loss": 1.5985,
+      "step": 11222
+    },
+    {
+      "epoch": 2.9491247085098697,
+      "grad_norm": 0.5994266271591187,
+      "learning_rate": 1.6996670755212898e-06,
+      "loss": 1.6229,
+      "step": 11224
+    },
+    {
+      "epoch": 2.9496502118435313,
+      "grad_norm": 0.6327415704727173,
+      "learning_rate": 1.6821447345365342e-06,
+      "loss": 1.5961,
+      "step": 11226
+    },
+    {
+      "epoch": 2.9501757151771932,
+      "grad_norm": 0.5327082276344299,
+      "learning_rate": 1.6646223935517788e-06,
+      "loss": 1.5832,
+      "step": 11228
+    },
+    {
+      "epoch": 2.9507012185108548,
+      "grad_norm": 0.6451945900917053,
+      "learning_rate": 1.647100052567023e-06,
+      "loss": 1.6072,
+      "step": 11230
+    },
+    {
+      "epoch": 2.9512267218445167,
+      "grad_norm": 0.6457750201225281,
+      "learning_rate": 1.6295777115822676e-06,
+      "loss": 1.59,
+      "step": 11232
+    },
+    {
+      "epoch": 2.9517522251781783,
+      "grad_norm": 0.7860383987426758,
+      "learning_rate": 1.612055370597512e-06,
+      "loss": 1.6288,
+      "step": 11234
+    },
+    {
+      "epoch": 2.9522777285118402,
+      "grad_norm": 0.5900551080703735,
+      "learning_rate": 1.5945330296127566e-06,
+      "loss": 1.6113,
+      "step": 11236
+    },
+    {
+      "epoch": 2.952803231845502,
+      "grad_norm": 0.606599748134613,
+      "learning_rate": 1.5770106886280007e-06,
+      "loss": 1.6201,
+      "step": 11238
+    },
+    {
+      "epoch": 2.9533287351791637,
+      "grad_norm": 0.6085255742073059,
+      "learning_rate": 1.5594883476432451e-06,
+      "loss": 1.5788,
+      "step": 11240
+    },
+    {
+      "epoch": 2.9538542385128257,
+      "grad_norm": 0.689579427242279,
+      "learning_rate": 1.5419660066584897e-06,
+      "loss": 1.5928,
+      "step": 11242
+    },
+    {
+      "epoch": 2.954379741846487,
+      "grad_norm": 0.5633424520492554,
+      "learning_rate": 1.524443665673734e-06,
+      "loss": 1.5989,
+      "step": 11244
+    },
+    {
+      "epoch": 2.954905245180149,
+      "grad_norm": 0.6903582811355591,
+      "learning_rate": 1.5069213246889785e-06,
+      "loss": 1.6527,
+      "step": 11246
+    },
+    {
+      "epoch": 2.9554307485138107,
+      "grad_norm": 0.6674286723136902,
+      "learning_rate": 1.4893989837042229e-06,
+      "loss": 1.6041,
+      "step": 11248
+    },
+    {
+      "epoch": 2.9559562518474727,
+      "grad_norm": 0.5618950724601746,
+      "learning_rate": 1.4718766427194675e-06,
+      "loss": 1.6209,
+      "step": 11250
+    },
+    {
+      "epoch": 2.9564817551811347,
+      "grad_norm": 0.60519939661026,
+      "learning_rate": 1.4543543017347119e-06,
+      "loss": 1.57,
+      "step": 11252
+    },
+    {
+      "epoch": 2.957007258514796,
+      "grad_norm": 0.6382578611373901,
+      "learning_rate": 1.4368319607499562e-06,
+      "loss": 1.6074,
+      "step": 11254
+    },
+    {
+      "epoch": 2.9575327618484577,
+      "grad_norm": 0.5759736895561218,
+      "learning_rate": 1.4193096197652006e-06,
+      "loss": 1.5892,
+      "step": 11256
+    },
+    {
+      "epoch": 2.9580582651821197,
+      "grad_norm": 0.5800794363021851,
+      "learning_rate": 1.4017872787804452e-06,
+      "loss": 1.643,
+      "step": 11258
+    },
+    {
+      "epoch": 2.9585837685157816,
+      "grad_norm": 0.5824633240699768,
+      "learning_rate": 1.3842649377956896e-06,
+      "loss": 1.5935,
+      "step": 11260
+    },
+    {
+      "epoch": 2.959109271849443,
+      "grad_norm": 0.5774669647216797,
+      "learning_rate": 1.366742596810934e-06,
+      "loss": 1.6421,
+      "step": 11262
+    },
+    {
+      "epoch": 2.959634775183105,
+      "grad_norm": 0.5631816387176514,
+      "learning_rate": 1.3492202558261784e-06,
+      "loss": 1.6194,
+      "step": 11264
+    },
+    {
+      "epoch": 2.960160278516767,
+      "grad_norm": 0.5980478525161743,
+      "learning_rate": 1.331697914841423e-06,
+      "loss": 1.6124,
+      "step": 11266
+    },
+    {
+      "epoch": 2.9606857818504286,
+      "grad_norm": 0.5869792699813843,
+      "learning_rate": 1.3141755738566674e-06,
+      "loss": 1.6033,
+      "step": 11268
+    },
+    {
+      "epoch": 2.96121128518409,
+      "grad_norm": 0.7216488122940063,
+      "learning_rate": 1.2966532328719118e-06,
+      "loss": 1.6396,
+      "step": 11270
+    },
+    {
+      "epoch": 2.961736788517752,
+      "grad_norm": 0.6487390398979187,
+      "learning_rate": 1.2791308918871562e-06,
+      "loss": 1.611,
+      "step": 11272
+    },
+    {
+      "epoch": 2.962262291851414,
+      "grad_norm": 0.6345615983009338,
+      "learning_rate": 1.2616085509024008e-06,
+      "loss": 1.575,
+      "step": 11274
+    },
+    {
+      "epoch": 2.9627877951850756,
+      "grad_norm": 0.692871630191803,
+      "learning_rate": 1.2440862099176451e-06,
+      "loss": 1.6005,
+      "step": 11276
+    },
+    {
+      "epoch": 2.9633132985187376,
+      "grad_norm": 0.6169196367263794,
+      "learning_rate": 1.2265638689328895e-06,
+      "loss": 1.5758,
+      "step": 11278
+    },
+    {
+      "epoch": 2.963838801852399,
+      "grad_norm": 0.5923426151275635,
+      "learning_rate": 1.209041527948134e-06,
+      "loss": 1.5737,
+      "step": 11280
+    },
+    {
+      "epoch": 2.964364305186061,
+      "grad_norm": 0.629149854183197,
+      "learning_rate": 1.1915191869633785e-06,
+      "loss": 1.6043,
+      "step": 11282
+    },
+    {
+      "epoch": 2.9648898085197226,
+      "grad_norm": 0.5396657586097717,
+      "learning_rate": 1.173996845978623e-06,
+      "loss": 1.5922,
+      "step": 11284
+    },
+    {
+      "epoch": 2.9654153118533846,
+      "grad_norm": 0.6270614862442017,
+      "learning_rate": 1.1564745049938673e-06,
+      "loss": 1.6032,
+      "step": 11286
+    },
+    {
+      "epoch": 2.9659408151870466,
+      "grad_norm": 0.6442562937736511,
+      "learning_rate": 1.1389521640091117e-06,
+      "loss": 1.5929,
+      "step": 11288
+    },
+    {
+      "epoch": 2.966466318520708,
+      "grad_norm": 0.5687601566314697,
+      "learning_rate": 1.1214298230243563e-06,
+      "loss": 1.5735,
+      "step": 11290
+    },
+    {
+      "epoch": 2.96699182185437,
+      "grad_norm": 0.6251326203346252,
+      "learning_rate": 1.1039074820396007e-06,
+      "loss": 1.5841,
+      "step": 11292
+    },
+    {
+      "epoch": 2.9675173251880316,
+      "grad_norm": 0.6409140229225159,
+      "learning_rate": 1.0863851410548448e-06,
+      "loss": 1.6202,
+      "step": 11294
+    },
+    {
+      "epoch": 2.9680428285216935,
+      "grad_norm": 0.5543487668037415,
+      "learning_rate": 1.0688628000700894e-06,
+      "loss": 1.6237,
+      "step": 11296
+    },
+    {
+      "epoch": 2.968568331855355,
+      "grad_norm": 0.6813648343086243,
+      "learning_rate": 1.0513404590853338e-06,
+      "loss": 1.6548,
+      "step": 11298
+    },
+    {
+      "epoch": 2.969093835189017,
+      "grad_norm": 0.5688581466674805,
+      "learning_rate": 1.0338181181005782e-06,
+      "loss": 1.597,
+      "step": 11300
+    },
+    {
+      "epoch": 2.969619338522679,
+      "grad_norm": 0.6080614328384399,
+      "learning_rate": 1.0162957771158226e-06,
+      "loss": 1.6439,
+      "step": 11302
+    },
+    {
+      "epoch": 2.9701448418563405,
+      "grad_norm": 0.5709330439567566,
+      "learning_rate": 9.987734361310672e-07,
+      "loss": 1.6394,
+      "step": 11304
+    },
+    {
+      "epoch": 2.970670345190002,
+      "grad_norm": 0.6443371772766113,
+      "learning_rate": 9.812510951463116e-07,
+      "loss": 1.6215,
+      "step": 11306
+    },
+    {
+      "epoch": 2.971195848523664,
+      "grad_norm": 0.6284047365188599,
+      "learning_rate": 9.63728754161556e-07,
+      "loss": 1.6182,
+      "step": 11308
+    },
+    {
+      "epoch": 2.971721351857326,
+      "grad_norm": 0.5524962544441223,
+      "learning_rate": 9.462064131768005e-07,
+      "loss": 1.5953,
+      "step": 11310
+    },
+    {
+      "epoch": 2.9722468551909875,
+      "grad_norm": 0.56756991147995,
+      "learning_rate": 9.286840721920448e-07,
+      "loss": 1.6048,
+      "step": 11312
+    },
+    {
+      "epoch": 2.9727723585246495,
+      "grad_norm": 0.729558527469635,
+      "learning_rate": 9.111617312072893e-07,
+      "loss": 1.6145,
+      "step": 11314
+    },
+    {
+      "epoch": 2.973297861858311,
+      "grad_norm": 0.6247385144233704,
+      "learning_rate": 8.936393902225337e-07,
+      "loss": 1.6176,
+      "step": 11316
+    },
+    {
+      "epoch": 2.973823365191973,
+      "grad_norm": 0.6063811779022217,
+      "learning_rate": 8.761170492377782e-07,
+      "loss": 1.5839,
+      "step": 11318
+    },
+    {
+      "epoch": 2.9743488685256345,
+      "grad_norm": 0.6044145226478577,
+      "learning_rate": 8.585947082530226e-07,
+      "loss": 1.6051,
+      "step": 11320
+    },
+    {
+      "epoch": 2.9748743718592965,
+      "grad_norm": 0.5743279457092285,
+      "learning_rate": 8.410723672682671e-07,
+      "loss": 1.6128,
+      "step": 11322
+    },
+    {
+      "epoch": 2.9753998751929585,
+      "grad_norm": 0.5868866443634033,
+      "learning_rate": 8.235500262835115e-07,
+      "loss": 1.6428,
+      "step": 11324
+    },
+    {
+      "epoch": 2.97592537852662,
+      "grad_norm": 0.5260694026947021,
+      "learning_rate": 8.06027685298756e-07,
+      "loss": 1.584,
+      "step": 11326
+    },
+    {
+      "epoch": 2.976450881860282,
+      "grad_norm": 0.6764076948165894,
+      "learning_rate": 7.885053443140004e-07,
+      "loss": 1.6163,
+      "step": 11328
+    },
+    {
+      "epoch": 2.9769763851939435,
+      "grad_norm": 0.5896710157394409,
+      "learning_rate": 7.709830033292449e-07,
+      "loss": 1.5717,
+      "step": 11330
+    },
+    {
+      "epoch": 2.9775018885276054,
+      "grad_norm": 0.6759716272354126,
+      "learning_rate": 7.534606623444892e-07,
+      "loss": 1.5856,
+      "step": 11332
+    },
+    {
+      "epoch": 2.978027391861267,
+      "grad_norm": 0.6693316102027893,
+      "learning_rate": 7.359383213597337e-07,
+      "loss": 1.6189,
+      "step": 11334
+    },
+    {
+      "epoch": 2.978552895194929,
+      "grad_norm": 0.5483260154724121,
+      "learning_rate": 7.184159803749781e-07,
+      "loss": 1.6106,
+      "step": 11336
+    },
+    {
+      "epoch": 2.979078398528591,
+      "grad_norm": 0.6094440817832947,
+      "learning_rate": 7.008936393902226e-07,
+      "loss": 1.6105,
+      "step": 11338
+    },
+    {
+      "epoch": 2.9796039018622524,
+      "grad_norm": 0.6723849773406982,
+      "learning_rate": 6.83371298405467e-07,
+      "loss": 1.5995,
+      "step": 11340
+    },
+    {
+      "epoch": 2.980129405195914,
+      "grad_norm": 0.6282168030738831,
+      "learning_rate": 6.658489574207115e-07,
+      "loss": 1.5797,
+      "step": 11342
+    },
+    {
+      "epoch": 2.980654908529576,
+      "grad_norm": 0.5661284923553467,
+      "learning_rate": 6.483266164359559e-07,
+      "loss": 1.5908,
+      "step": 11344
+    },
+    {
+      "epoch": 2.981180411863238,
+      "grad_norm": 0.598994255065918,
+      "learning_rate": 6.308042754512004e-07,
+      "loss": 1.592,
+      "step": 11346
+    },
+    {
+      "epoch": 2.9817059151968994,
+      "grad_norm": 0.6033176183700562,
+      "learning_rate": 6.132819344664448e-07,
+      "loss": 1.5935,
+      "step": 11348
+    },
+    {
+      "epoch": 2.9822314185305614,
+      "grad_norm": 0.6345306038856506,
+      "learning_rate": 5.957595934816893e-07,
+      "loss": 1.5789,
+      "step": 11350
+    },
+    {
+      "epoch": 2.982756921864223,
+      "grad_norm": 0.606160044670105,
+      "learning_rate": 5.782372524969336e-07,
+      "loss": 1.5949,
+      "step": 11352
+    },
+    {
+      "epoch": 2.983282425197885,
+      "grad_norm": 0.7126883268356323,
+      "learning_rate": 5.607149115121781e-07,
+      "loss": 1.5786,
+      "step": 11354
+    },
+    {
+      "epoch": 2.9838079285315464,
+      "grad_norm": 0.6453995704650879,
+      "learning_rate": 5.431925705274224e-07,
+      "loss": 1.6359,
+      "step": 11356
+    },
+    {
+      "epoch": 2.9843334318652084,
+      "grad_norm": 0.5591650605201721,
+      "learning_rate": 5.256702295426669e-07,
+      "loss": 1.596,
+      "step": 11358
+    },
+    {
+      "epoch": 2.9848589351988704,
+      "grad_norm": 0.5831902027130127,
+      "learning_rate": 5.081478885579113e-07,
+      "loss": 1.5785,
+      "step": 11360
+    },
+    {
+      "epoch": 2.985384438532532,
+      "grad_norm": 0.5874847173690796,
+      "learning_rate": 4.906255475731558e-07,
+      "loss": 1.6092,
+      "step": 11362
+    },
+    {
+      "epoch": 2.985909941866194,
+      "grad_norm": 0.6131449937820435,
+      "learning_rate": 4.7310320658840023e-07,
+      "loss": 1.6002,
+      "step": 11364
+    },
+    {
+      "epoch": 2.9864354451998554,
+      "grad_norm": 0.5965538620948792,
+      "learning_rate": 4.5558086560364467e-07,
+      "loss": 1.5937,
+      "step": 11366
+    },
+    {
+      "epoch": 2.9869609485335173,
+      "grad_norm": 0.6877683401107788,
+      "learning_rate": 4.380585246188891e-07,
+      "loss": 1.5981,
+      "step": 11368
+    },
+    {
+      "epoch": 2.987486451867179,
+      "grad_norm": 0.6720339059829712,
+      "learning_rate": 4.2053618363413355e-07,
+      "loss": 1.6071,
+      "step": 11370
+    },
+    {
+      "epoch": 2.988011955200841,
+      "grad_norm": 0.7178999185562134,
+      "learning_rate": 4.03013842649378e-07,
+      "loss": 1.6089,
+      "step": 11372
+    },
+    {
+      "epoch": 2.988537458534503,
+      "grad_norm": 0.5988190770149231,
+      "learning_rate": 3.8549150166462243e-07,
+      "loss": 1.6001,
+      "step": 11374
+    },
+    {
+      "epoch": 2.9890629618681643,
+      "grad_norm": 0.6235753297805786,
+      "learning_rate": 3.6796916067986687e-07,
+      "loss": 1.6124,
+      "step": 11376
+    },
+    {
+      "epoch": 2.989588465201826,
+      "grad_norm": 0.6031942367553711,
+      "learning_rate": 3.504468196951113e-07,
+      "loss": 1.6445,
+      "step": 11378
+    },
+    {
+      "epoch": 2.990113968535488,
+      "grad_norm": 0.6021082401275635,
+      "learning_rate": 3.3292447871035575e-07,
+      "loss": 1.6139,
+      "step": 11380
+    },
+    {
+      "epoch": 2.99063947186915,
+      "grad_norm": 0.5596811175346375,
+      "learning_rate": 3.154021377256002e-07,
+      "loss": 1.5923,
+      "step": 11382
+    },
+    {
+      "epoch": 2.9911649752028113,
+      "grad_norm": 0.5226958990097046,
+      "learning_rate": 2.9787979674084463e-07,
+      "loss": 1.6145,
+      "step": 11384
+    },
+    {
+      "epoch": 2.9916904785364733,
+      "grad_norm": 0.6241964101791382,
+      "learning_rate": 2.8035745575608907e-07,
+      "loss": 1.5997,
+      "step": 11386
+    },
+    {
+      "epoch": 2.992215981870135,
+      "grad_norm": 0.5991497039794922,
+      "learning_rate": 2.6283511477133345e-07,
+      "loss": 1.6279,
+      "step": 11388
+    },
+    {
+      "epoch": 2.992741485203797,
+      "grad_norm": 0.5512558817863464,
+      "learning_rate": 2.453127737865779e-07,
+      "loss": 1.5999,
+      "step": 11390
+    },
+    {
+      "epoch": 2.9932669885374583,
+      "grad_norm": 0.6609853506088257,
+      "learning_rate": 2.2779043280182233e-07,
+      "loss": 1.6145,
+      "step": 11392
+    },
+    {
+      "epoch": 2.9937924918711203,
+      "grad_norm": 0.6005377769470215,
+      "learning_rate": 2.1026809181706677e-07,
+      "loss": 1.5893,
+      "step": 11394
+    },
+    {
+      "epoch": 2.9943179952047823,
+      "grad_norm": 0.5884697437286377,
+      "learning_rate": 1.9274575083231121e-07,
+      "loss": 1.5813,
+      "step": 11396
+    },
+    {
+      "epoch": 2.994843498538444,
+      "grad_norm": 0.7103734612464905,
+      "learning_rate": 1.7522340984755565e-07,
+      "loss": 1.592,
+      "step": 11398
+    },
+    {
+      "epoch": 2.9953690018721058,
+      "grad_norm": 0.5749437808990479,
+      "learning_rate": 1.577010688628001e-07,
+      "loss": 1.6045,
+      "step": 11400
+    },
+    {
+      "epoch": 2.9958945052057673,
+      "grad_norm": 0.6688903570175171,
+      "learning_rate": 1.4017872787804453e-07,
+      "loss": 1.6188,
+      "step": 11402
+    },
+    {
+      "epoch": 2.9964200085394292,
+      "grad_norm": 0.6899111866950989,
+      "learning_rate": 1.2265638689328895e-07,
+      "loss": 1.6102,
+      "step": 11404
+    },
+    {
+      "epoch": 2.9969455118730908,
+      "grad_norm": 0.6025496125221252,
+      "learning_rate": 1.0513404590853339e-07,
+      "loss": 1.5927,
+      "step": 11406
+    },
+    {
+      "epoch": 2.9974710152067527,
+      "grad_norm": 0.8233180642127991,
+      "learning_rate": 8.761170492377783e-08,
+      "loss": 1.6267,
+      "step": 11408
+    },
+    {
+      "epoch": 2.9979965185404147,
+      "grad_norm": 0.633036196231842,
+      "learning_rate": 7.008936393902227e-08,
+      "loss": 1.6246,
+      "step": 11410
+    },
+    {
+      "epoch": 2.9985220218740762,
+      "grad_norm": 0.6616869568824768,
+      "learning_rate": 5.2567022954266694e-08,
+      "loss": 1.5989,
+      "step": 11412
+    },
+    {
+      "epoch": 2.9990475252077378,
+      "grad_norm": 0.7166438102722168,
+      "learning_rate": 3.5044681969511133e-08,
+      "loss": 1.6247,
+      "step": 11414
+    },
+    {
+      "epoch": 2.999310276874569,
+      "step": 11415,
+      "total_flos": 8.316227983158804e+17,
+      "train_loss": 1.7114452323838383,
+      "train_runtime": 54638.3985,
+      "train_samples_per_second": 53.496,
+      "train_steps_per_second": 0.209
+    },
+    {
+      "epoch": 2.999310276874569,
+      "eval_loss": 1.6470075845718384,
+      "eval_runtime": 487.9081,
+      "eval_samples_per_second": 249.615,
+      "eval_steps_per_second": 31.203,
+      "step": 11415
+    },
+    {
+      "epoch": 2.999310276874569,
+      "eval_loss": 1.6461076736450195,
+      "eval_runtime": 486.7382,
+      "eval_samples_per_second": 250.215,
+      "eval_steps_per_second": 31.278,
+      "step": 11415
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 11415,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.316227983158804e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}