diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14615 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 200.0,
+  "global_step": 10410,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019212295869356388,
+      "grad_norm": 0.08390819281339645,
+      "learning_rate": 5.758157389635316e-07,
+      "loss": 3.274601936340332,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009606147934678194,
+      "grad_norm": 0.08173709362745285,
+      "learning_rate": 2.879078694817658e-06,
+      "loss": 3.2788190841674805,
+      "step": 5
+    },
+    {
+      "epoch": 0.0019212295869356388,
+      "grad_norm": 0.09356535971164703,
+      "learning_rate": 5.758157389635316e-06,
+      "loss": 3.2803062438964843,
+      "step": 10
+    },
+    {
+      "epoch": 0.002881844380403458,
+      "grad_norm": 0.11097322404384613,
+      "learning_rate": 8.637236084452974e-06,
+      "loss": 3.277487564086914,
+      "step": 15
+    },
+    {
+      "epoch": 0.0038424591738712775,
+      "grad_norm": 0.15006589889526367,
+      "learning_rate": 1.1516314779270632e-05,
+      "loss": 3.281480407714844,
+      "step": 20
+    },
+    {
+      "epoch": 0.004803073967339097,
+      "grad_norm": 0.14452151954174042,
+      "learning_rate": 1.439539347408829e-05,
+      "loss": 3.2803359985351563,
+      "step": 25
+    },
+    {
+      "epoch": 0.005763688760806916,
+      "grad_norm": 0.16946451365947723,
+      "learning_rate": 1.7274472168905948e-05,
+      "loss": 3.287946319580078,
+      "step": 30
+    },
+    {
+      "epoch": 0.0067243035542747355,
+      "grad_norm": 0.14511597156524658,
+      "learning_rate": 2.015355086372361e-05,
+      "loss": 3.278676986694336,
+      "step": 35
+    },
+    {
+      "epoch": 0.007684918347742555,
+      "grad_norm": 0.21068508923053741,
+      "learning_rate": 2.3032629558541264e-05,
+      "loss": 3.2816986083984374,
+      "step": 40
+    },
+    {
+      "epoch": 0.008645533141210375,
+      "grad_norm": 0.2152123898267746,
+      "learning_rate": 2.591170825335892e-05,
+      "loss": 3.2830833435058593,
+      "step": 45
+    },
+    {
+      "epoch": 0.009606147934678195,
+      "grad_norm": 0.2621569037437439,
+      "learning_rate": 2.879078694817658e-05,
+      "loss": 3.279366302490234,
+      "step": 50
+    },
+    {
+      "epoch": 0.010566762728146013,
+      "grad_norm": 0.29736942052841187,
+      "learning_rate": 3.166986564299424e-05,
+      "loss": 3.2836788177490233,
+      "step": 55
+    },
+    {
+      "epoch": 0.011527377521613832,
+      "grad_norm": 0.31358596682548523,
+      "learning_rate": 3.4548944337811895e-05,
+      "loss": 3.2797149658203124,
+      "step": 60
+    },
+    {
+      "epoch": 0.012487992315081652,
+      "grad_norm": 0.42882370948791504,
+      "learning_rate": 3.742802303262955e-05,
+      "loss": 3.2872390747070312,
+      "step": 65
+    },
+    {
+      "epoch": 0.013448607108549471,
+      "grad_norm": 0.3247455060482025,
+      "learning_rate": 4.030710172744722e-05,
+      "loss": 3.2825721740722655,
+      "step": 70
+    },
+    {
+      "epoch": 0.01440922190201729,
+      "grad_norm": 0.40256041288375854,
+      "learning_rate": 4.318618042226487e-05,
+      "loss": 3.279859924316406,
+      "step": 75
+    },
+    {
+      "epoch": 0.01536983669548511,
+      "grad_norm": 0.5199999809265137,
+      "learning_rate": 4.606525911708253e-05,
+      "loss": 3.284808349609375,
+      "step": 80
+    },
+    {
+      "epoch": 0.01633045148895293,
+      "grad_norm": 0.6197025179862976,
+      "learning_rate": 4.894433781190019e-05,
+      "loss": 3.2813262939453125,
+      "step": 85
+    },
+    {
+      "epoch": 0.01729106628242075,
+      "grad_norm": 0.4358411133289337,
+      "learning_rate": 5.182341650671784e-05,
+      "loss": 3.2836116790771483,
+      "step": 90
+    },
+    {
+      "epoch": 0.01825168107588857,
+      "grad_norm": 0.5619639158248901,
+      "learning_rate": 5.4702495201535504e-05,
+      "loss": 3.281964874267578,
+      "step": 95
+    },
+    {
+      "epoch": 0.01921229586935639,
+      "grad_norm": 0.42559099197387695,
+      "learning_rate": 5.758157389635316e-05,
+      "loss": 3.284071350097656,
+      "step": 100
+    },
+    {
+      "epoch": 0.020172910662824207,
+      "grad_norm": 0.5353518128395081,
+      "learning_rate": 6.0460652591170814e-05,
+      "loss": 3.286463165283203,
+      "step": 105
+    },
+    {
+      "epoch": 0.021133525456292025,
+      "grad_norm": 0.8242971897125244,
+      "learning_rate": 6.333973128598848e-05,
+      "loss": 3.28314094543457,
+      "step": 110
+    },
+    {
+      "epoch": 0.022094140249759846,
+      "grad_norm": 0.5128299593925476,
+      "learning_rate": 6.621880998080614e-05,
+      "loss": 3.2824722290039063,
+      "step": 115
+    },
+    {
+      "epoch": 0.023054755043227664,
+      "grad_norm": 0.7521133422851562,
+      "learning_rate": 6.909788867562379e-05,
+      "loss": 3.2825164794921875,
+      "step": 120
+    },
+    {
+      "epoch": 0.024015369836695485,
+      "grad_norm": 0.7099725008010864,
+      "learning_rate": 7.197696737044145e-05,
+      "loss": 3.2856216430664062,
+      "step": 125
+    },
+    {
+      "epoch": 0.024975984630163303,
+      "grad_norm": 0.673387885093689,
+      "learning_rate": 7.48560460652591e-05,
+      "loss": 3.2845367431640624,
+      "step": 130
+    },
+    {
+      "epoch": 0.025936599423631124,
+      "grad_norm": 0.47837865352630615,
+      "learning_rate": 7.773512476007677e-05,
+      "loss": 3.2803184509277346,
+      "step": 135
+    },
+    {
+      "epoch": 0.026897214217098942,
+      "grad_norm": 0.5746335983276367,
+      "learning_rate": 8.061420345489444e-05,
+      "loss": 3.2853370666503907,
+      "step": 140
+    },
+    {
+      "epoch": 0.027857829010566763,
+      "grad_norm": 0.9516947269439697,
+      "learning_rate": 8.349328214971209e-05,
+      "loss": 3.2811763763427733,
+      "step": 145
+    },
+    {
+      "epoch": 0.02881844380403458,
+      "grad_norm": 0.7206563949584961,
+      "learning_rate": 8.637236084452975e-05,
+      "loss": 3.2858177185058595,
+      "step": 150
+    },
+    {
+      "epoch": 0.029779058597502402,
+      "grad_norm": 0.9468950033187866,
+      "learning_rate": 8.92514395393474e-05,
+      "loss": 3.2817359924316407,
+      "step": 155
+    },
+    {
+      "epoch": 0.03073967339097022,
+      "grad_norm": 0.9247293472290039,
+      "learning_rate": 9.213051823416505e-05,
+      "loss": 3.283827209472656,
+      "step": 160
+    },
+    {
+      "epoch": 0.03170028818443804,
+      "grad_norm": 0.7639815807342529,
+      "learning_rate": 9.500959692898272e-05,
+      "loss": 3.2805789947509765,
+      "step": 165
+    },
+    {
+      "epoch": 0.03266090297790586,
+      "grad_norm": 1.0864217281341553,
+      "learning_rate": 9.788867562380038e-05,
+      "loss": 3.282984161376953,
+      "step": 170
+    },
+    {
+      "epoch": 0.03362151777137368,
+      "grad_norm": 1.0888683795928955,
+      "learning_rate": 0.00010076775431861803,
+      "loss": 3.2868572235107423,
+      "step": 175
+    },
+    {
+      "epoch": 0.0345821325648415,
+      "grad_norm": 0.7901690602302551,
+      "learning_rate": 0.00010364683301343569,
+      "loss": 3.2854034423828127,
+      "step": 180
+    },
+    {
+      "epoch": 0.03554274735830932,
+      "grad_norm": 1.132866621017456,
+      "learning_rate": 0.00010652591170825334,
+      "loss": 3.283061218261719,
+      "step": 185
+    },
+    {
+      "epoch": 0.03650336215177714,
+      "grad_norm": 1.1301084756851196,
+      "learning_rate": 0.00010940499040307101,
+      "loss": 3.2865623474121093,
+      "step": 190
+    },
+    {
+      "epoch": 0.037463976945244955,
+      "grad_norm": 0.9956747889518738,
+      "learning_rate": 0.00011228406909788866,
+      "loss": 3.2860641479492188,
+      "step": 195
+    },
+    {
+      "epoch": 0.03842459173871278,
+      "grad_norm": 1.1139202117919922,
+      "learning_rate": 0.00011516314779270632,
+      "loss": 3.2860763549804686,
+      "step": 200
+    },
+    {
+      "epoch": 0.0393852065321806,
+      "grad_norm": 1.1494112014770508,
+      "learning_rate": 0.00011804222648752397,
+      "loss": 3.2874046325683595,
+      "step": 205
+    },
+    {
+      "epoch": 0.040345821325648415,
+      "grad_norm": 1.1791850328445435,
+      "learning_rate": 0.00012092130518234163,
+      "loss": 3.28511962890625,
+      "step": 210
+    },
+    {
+      "epoch": 0.04130643611911623,
+      "grad_norm": 0.8412664532661438,
+      "learning_rate": 0.0001238003838771593,
+      "loss": 3.2826396942138674,
+      "step": 215
+    },
+    {
+      "epoch": 0.04226705091258405,
+      "grad_norm": 1.0537320375442505,
+      "learning_rate": 0.00012667946257197696,
+      "loss": 3.2859474182128907,
+      "step": 220
+    },
+    {
+      "epoch": 0.043227665706051875,
+      "grad_norm": 0.9535348415374756,
+      "learning_rate": 0.00012955854126679462,
+      "loss": 3.2872699737548827,
+      "step": 225
+    },
+    {
+      "epoch": 0.04418828049951969,
+      "grad_norm": 1.2287579774856567,
+      "learning_rate": 0.00013243761996161227,
+      "loss": 3.2854686737060548,
+      "step": 230
+    },
+    {
+      "epoch": 0.04514889529298751,
+      "grad_norm": 0.8720296621322632,
+      "learning_rate": 0.00013531669865642993,
+      "loss": 3.2873931884765626,
+      "step": 235
+    },
+    {
+      "epoch": 0.04610951008645533,
+      "grad_norm": 0.9984288811683655,
+      "learning_rate": 0.00013819577735124758,
+      "loss": 3.28588752746582,
+      "step": 240
+    },
+    {
+      "epoch": 0.04707012487992315,
+      "grad_norm": 1.3232122659683228,
+      "learning_rate": 0.00014107485604606524,
+      "loss": 3.2899200439453127,
+      "step": 245
+    },
+    {
+      "epoch": 0.04803073967339097,
+      "grad_norm": 1.0983723402023315,
+      "learning_rate": 0.0001439539347408829,
+      "loss": 3.287739562988281,
+      "step": 250
+    },
+    {
+      "epoch": 0.04899135446685879,
+      "grad_norm": 0.9880014657974243,
+      "learning_rate": 0.00014683301343570055,
+      "loss": 3.2914413452148437,
+      "step": 255
+    },
+    {
+      "epoch": 0.049951969260326606,
+      "grad_norm": 1.0988938808441162,
+      "learning_rate": 0.0001497120921305182,
+      "loss": 3.2892181396484377,
+      "step": 260
+    },
+    {
+      "epoch": 0.05091258405379443,
+      "grad_norm": 0.9497994780540466,
+      "learning_rate": 0.00015259117082533588,
+      "loss": 3.289907455444336,
+      "step": 265
+    },
+    {
+      "epoch": 0.05187319884726225,
+      "grad_norm": 1.4786131381988525,
+      "learning_rate": 0.00015547024952015354,
+      "loss": 3.2885669708251952,
+      "step": 270
+    },
+    {
+      "epoch": 0.052833813640730067,
+      "grad_norm": 1.1913130283355713,
+      "learning_rate": 0.0001583493282149712,
+      "loss": 3.2930023193359377,
+      "step": 275
+    },
+    {
+      "epoch": 0.053794428434197884,
+      "grad_norm": 1.0576707124710083,
+      "learning_rate": 0.00016122840690978887,
+      "loss": 3.285129165649414,
+      "step": 280
+    },
+    {
+      "epoch": 0.05475504322766571,
+      "grad_norm": 1.3840175867080688,
+      "learning_rate": 0.0001641074856046065,
+      "loss": 3.2881790161132813,
+      "step": 285
+    },
+    {
+      "epoch": 0.05571565802113353,
+      "grad_norm": 1.1912461519241333,
+      "learning_rate": 0.00016698656429942418,
+      "loss": 3.2907310485839845,
+      "step": 290
+    },
+    {
+      "epoch": 0.056676272814601344,
+      "grad_norm": 0.8021464347839355,
+      "learning_rate": 0.0001698656429942418,
+      "loss": 3.2897079467773436,
+      "step": 295
+    },
+    {
+      "epoch": 0.05763688760806916,
+      "grad_norm": 1.2219209671020508,
+      "learning_rate": 0.0001727447216890595,
+      "loss": 3.288962554931641,
+      "step": 300
+    },
+    {
+      "epoch": 0.05859750240153699,
+      "grad_norm": 1.13152015209198,
+      "learning_rate": 0.00017562380038387714,
+      "loss": 3.2893699645996093,
+      "step": 305
+    },
+    {
+      "epoch": 0.059558117195004805,
+      "grad_norm": 1.5439852476119995,
+      "learning_rate": 0.0001785028790786948,
+      "loss": 3.293457794189453,
+      "step": 310
+    },
+    {
+      "epoch": 0.06051873198847262,
+      "grad_norm": 1.4060869216918945,
+      "learning_rate": 0.00018138195777351245,
+      "loss": 3.2903827667236327,
+      "step": 315
+    },
+    {
+      "epoch": 0.06147934678194044,
+      "grad_norm": 1.0909712314605713,
+      "learning_rate": 0.0001842610364683301,
+      "loss": 3.293254852294922,
+      "step": 320
+    },
+    {
+      "epoch": 0.06243996157540826,
+      "grad_norm": 1.4758312702178955,
+      "learning_rate": 0.00018714011516314776,
+      "loss": 3.291452407836914,
+      "step": 325
+    },
+    {
+      "epoch": 0.06340057636887608,
+      "grad_norm": 1.3612134456634521,
+      "learning_rate": 0.00019001919385796544,
+      "loss": 3.2947055816650392,
+      "step": 330
+    },
+    {
+      "epoch": 0.0643611911623439,
+      "grad_norm": 1.017069935798645,
+      "learning_rate": 0.00019289827255278307,
+      "loss": 3.2914665222167967,
+      "step": 335
+    },
+    {
+      "epoch": 0.06532180595581172,
+      "grad_norm": 1.114537239074707,
+      "learning_rate": 0.00019577735124760075,
+      "loss": 3.2918880462646483,
+      "step": 340
+    },
+    {
+      "epoch": 0.06628242074927954,
+      "grad_norm": 1.359885334968567,
+      "learning_rate": 0.00019865642994241838,
+      "loss": 3.291158676147461,
+      "step": 345
+    },
+    {
+      "epoch": 0.06724303554274735,
+      "grad_norm": 1.5216479301452637,
+      "learning_rate": 0.00020153550863723606,
+      "loss": 3.2951950073242187,
+      "step": 350
+    },
+    {
+      "epoch": 0.06820365033621517,
+      "grad_norm": 1.0338709354400635,
+      "learning_rate": 0.00020441458733205374,
+      "loss": 3.2910877227783204,
+      "step": 355
+    },
+    {
+      "epoch": 0.069164265129683,
+      "grad_norm": 1.3676224946975708,
+      "learning_rate": 0.00020729366602687137,
+      "loss": 3.2979248046875,
+      "step": 360
+    },
+    {
+      "epoch": 0.07012487992315082,
+      "grad_norm": 1.508606195449829,
+      "learning_rate": 0.00021017274472168905,
+      "loss": 3.2947929382324217,
+      "step": 365
+    },
+    {
+      "epoch": 0.07108549471661864,
+      "grad_norm": 1.4694204330444336,
+      "learning_rate": 0.00021305182341650668,
+      "loss": 3.290261077880859,
+      "step": 370
+    },
+    {
+      "epoch": 0.07204610951008646,
+      "grad_norm": 1.629552960395813,
+      "learning_rate": 0.00021593090211132436,
+      "loss": 3.292864990234375,
+      "step": 375
+    },
+    {
+      "epoch": 0.07300672430355427,
+      "grad_norm": 1.3059035539627075,
+      "learning_rate": 0.00021880998080614202,
+      "loss": 3.297530746459961,
+      "step": 380
+    },
+    {
+      "epoch": 0.07396733909702209,
+      "grad_norm": 1.6734189987182617,
+      "learning_rate": 0.00022168905950095967,
+      "loss": 3.2937080383300783,
+      "step": 385
+    },
+    {
+      "epoch": 0.07492795389048991,
+      "grad_norm": 1.9466580152511597,
+      "learning_rate": 0.00022456813819577733,
+      "loss": 3.2989707946777345,
+      "step": 390
+    },
+    {
+      "epoch": 0.07588856868395773,
+      "grad_norm": 1.4589389562606812,
+      "learning_rate": 0.00022744721689059498,
+      "loss": 3.2969600677490236,
+      "step": 395
+    },
+    {
+      "epoch": 0.07684918347742556,
+      "grad_norm": 1.273814082145691,
+      "learning_rate": 0.00023032629558541264,
+      "loss": 3.2937450408935547,
+      "step": 400
+    },
+    {
+      "epoch": 0.07780979827089338,
+      "grad_norm": 1.6920170783996582,
+      "learning_rate": 0.00023320537428023032,
+      "loss": 3.2936920166015624,
+      "step": 405
+    },
+    {
+      "epoch": 0.0787704130643612,
+      "grad_norm": 1.7424761056900024,
+      "learning_rate": 0.00023608445297504794,
+      "loss": 3.29752197265625,
+      "step": 410
+    },
+    {
+      "epoch": 0.07973102785782901,
+      "grad_norm": 1.4278359413146973,
+      "learning_rate": 0.00023896353166986563,
+      "loss": 3.2931358337402346,
+      "step": 415
+    },
+    {
+      "epoch": 0.08069164265129683,
+      "grad_norm": 1.3596338033676147,
+      "learning_rate": 0.00024184261036468325,
+      "loss": 3.2951515197753904,
+      "step": 420
+    },
+    {
+      "epoch": 0.08165225744476465,
+      "grad_norm": 1.4509109258651733,
+      "learning_rate": 0.00024472168905950096,
+      "loss": 3.2955284118652344,
+      "step": 425
+    },
+    {
+      "epoch": 0.08261287223823247,
+      "grad_norm": 2.0564208030700684,
+      "learning_rate": 0.0002476007677543186,
+      "loss": 3.2958652496337892,
+      "step": 430
+    },
+    {
+      "epoch": 0.08357348703170028,
+      "grad_norm": 1.80084228515625,
+      "learning_rate": 0.00025047984644913627,
+      "loss": 3.299582672119141,
+      "step": 435
+    },
+    {
+      "epoch": 0.0845341018251681,
+      "grad_norm": 1.2458295822143555,
+      "learning_rate": 0.0002533589251439539,
+      "loss": 3.2957221984863283,
+      "step": 440
+    },
+    {
+      "epoch": 0.08549471661863593,
+      "grad_norm": 1.5943655967712402,
+      "learning_rate": 0.0002562380038387716,
+      "loss": 3.296315002441406,
+      "step": 445
+    },
+    {
+      "epoch": 0.08645533141210375,
+      "grad_norm": 1.4959216117858887,
+      "learning_rate": 0.00025911708253358924,
+      "loss": 3.2936065673828123,
+      "step": 450
+    },
+    {
+      "epoch": 0.08741594620557157,
+      "grad_norm": 0.9593575596809387,
+      "learning_rate": 0.0002619961612284069,
+      "loss": 3.2923255920410157,
+      "step": 455
+    },
+    {
+      "epoch": 0.08837656099903939,
+      "grad_norm": 1.7525019645690918,
+      "learning_rate": 0.00026487523992322454,
+      "loss": 3.297338104248047,
+      "step": 460
+    },
+    {
+      "epoch": 0.0893371757925072,
+      "grad_norm": 1.5253691673278809,
+      "learning_rate": 0.0002677543186180422,
+      "loss": 3.29698486328125,
+      "step": 465
+    },
+    {
+      "epoch": 0.09029779058597502,
+      "grad_norm": 1.7394542694091797,
+      "learning_rate": 0.00027063339731285985,
+      "loss": 3.2963401794433596,
+      "step": 470
+    },
+    {
+      "epoch": 0.09125840537944284,
+      "grad_norm": 1.3029968738555908,
+      "learning_rate": 0.0002735124760076775,
+      "loss": 3.2992328643798827,
+      "step": 475
+    },
+    {
+      "epoch": 0.09221902017291066,
+      "grad_norm": 1.3961071968078613,
+      "learning_rate": 0.00027639155470249516,
+      "loss": 3.3040702819824217,
+      "step": 480
+    },
+    {
+      "epoch": 0.09317963496637849,
+      "grad_norm": 0.9128854274749756,
+      "learning_rate": 0.0002792706333973128,
+      "loss": 3.2994720458984377,
+      "step": 485
+    },
+    {
+      "epoch": 0.0941402497598463,
+      "grad_norm": 1.39821195602417,
+      "learning_rate": 0.00028214971209213047,
+      "loss": 3.3009670257568358,
+      "step": 490
+    },
+    {
+      "epoch": 0.09510086455331412,
+      "grad_norm": 1.3767975568771362,
+      "learning_rate": 0.0002850287907869481,
+      "loss": 3.298174285888672,
+      "step": 495
+    },
+    {
+      "epoch": 0.09606147934678194,
+      "grad_norm": 1.5919218063354492,
+      "learning_rate": 0.0002879078694817658,
+      "loss": 3.2930580139160157,
+      "step": 500
+    },
+    {
+      "epoch": 0.09702209414024976,
+      "grad_norm": 2.060764789581299,
+      "learning_rate": 0.0002907869481765835,
+      "loss": 3.3007606506347655,
+      "step": 505
+    },
+    {
+      "epoch": 0.09798270893371758,
+      "grad_norm": 1.864414095878601,
+      "learning_rate": 0.0002936660268714011,
+      "loss": 3.302916717529297,
+      "step": 510
+    },
+    {
+      "epoch": 0.0989433237271854,
+      "grad_norm": 1.410025954246521,
+      "learning_rate": 0.0002965451055662188,
+      "loss": 3.297088623046875,
+      "step": 515
+    },
+    {
+      "epoch": 0.09990393852065321,
+      "grad_norm": 1.4724565744400024,
+      "learning_rate": 0.0002994241842610364,
+      "loss": 3.2969940185546873,
+      "step": 520
+    },
+    {
+      "epoch": 0.10086455331412104,
+      "grad_norm": 1.1308706998825073,
+      "learning_rate": 0.0002999998788910666,
+      "loss": 3.295259475708008,
+      "step": 525
+    },
+    {
+      "epoch": 0.10182516810758886,
+      "grad_norm": 2.0092132091522217,
+      "learning_rate": 0.00029999938688636,
+      "loss": 3.3064605712890627,
+      "step": 530
+    },
+    {
+      "epoch": 0.10278578290105668,
+      "grad_norm": 1.5756242275238037,
+      "learning_rate": 0.0002999985164178122,
+      "loss": 3.302138900756836,
+      "step": 535
+    },
+    {
+      "epoch": 0.1037463976945245,
+      "grad_norm": 1.6854816675186157,
+      "learning_rate": 0.00029999726748761955,
+      "loss": 3.29990234375,
+      "step": 540
+    },
+    {
+      "epoch": 0.10470701248799232,
+      "grad_norm": 1.5925079584121704,
+      "learning_rate": 0.0002999956400989331,
+      "loss": 3.3017444610595703,
+      "step": 545
+    },
+    {
+      "epoch": 0.10566762728146013,
+      "grad_norm": 0.9805465340614319,
+      "learning_rate": 0.00029999363425585907,
+      "loss": 3.2947647094726564,
+      "step": 550
+    },
+    {
+      "epoch": 0.10662824207492795,
+      "grad_norm": 1.143500804901123,
+      "learning_rate": 0.0002999912499634584,
+      "loss": 3.3032691955566404,
+      "step": 555
+    },
+    {
+      "epoch": 0.10758885686839577,
+      "grad_norm": 1.57884681224823,
+      "learning_rate": 0.00029998848722774676,
+      "loss": 3.3033897399902346,
+      "step": 560
+    },
+    {
+      "epoch": 0.10854947166186359,
+      "grad_norm": 1.9447771310806274,
+      "learning_rate": 0.00029998534605569496,
+      "loss": 3.2998809814453125,
+      "step": 565
+    },
+    {
+      "epoch": 0.10951008645533142,
+      "grad_norm": 1.4967366456985474,
+      "learning_rate": 0.0002999818264552284,
+      "loss": 3.2998199462890625,
+      "step": 570
+    },
+    {
+      "epoch": 0.11047070124879924,
+      "grad_norm": 1.5965176820755005,
+      "learning_rate": 0.0002999779284352275,
+      "loss": 3.2984855651855467,
+      "step": 575
+    },
+    {
+      "epoch": 0.11143131604226705,
+      "grad_norm": 1.3109067678451538,
+      "learning_rate": 0.0002999736520055273,
+      "loss": 3.293305206298828,
+      "step": 580
+    },
+    {
+      "epoch": 0.11239193083573487,
+      "grad_norm": 1.374250054359436,
+      "learning_rate": 0.00029996899717691766,
+      "loss": 3.2990135192871093,
+      "step": 585
+    },
+    {
+      "epoch": 0.11335254562920269,
+      "grad_norm": 1.8241665363311768,
+      "learning_rate": 0.0002999639639611432,
+      "loss": 3.2949737548828124,
+      "step": 590
+    },
+    {
+      "epoch": 0.1143131604226705,
+      "grad_norm": 1.1245280504226685,
+      "learning_rate": 0.00029995855237090327,
+      "loss": 3.3007225036621093,
+      "step": 595
+    },
+    {
+      "epoch": 0.11527377521613832,
+      "grad_norm": 1.290839433670044,
+      "learning_rate": 0.0002999527624198518,
+      "loss": 3.2996570587158205,
+      "step": 600
+    },
+    {
+      "epoch": 0.11623439000960614,
+      "grad_norm": 1.100143551826477,
+      "learning_rate": 0.00029994659412259746,
+      "loss": 3.2947521209716797,
+      "step": 605
+    },
+    {
+      "epoch": 0.11719500480307397,
+      "grad_norm": 1.6502522230148315,
+      "learning_rate": 0.0002999400474947036,
+      "loss": 3.295918273925781,
+      "step": 610
+    },
+    {
+      "epoch": 0.11815561959654179,
+      "grad_norm": 1.4078381061553955,
+      "learning_rate": 0.0002999331225526879,
+      "loss": 3.299440383911133,
+      "step": 615
+    },
+    {
+      "epoch": 0.11911623439000961,
+      "grad_norm": 1.4280340671539307,
+      "learning_rate": 0.0002999258193140227,
+      "loss": 3.296113967895508,
+      "step": 620
+    },
+    {
+      "epoch": 0.12007684918347743,
+      "grad_norm": 1.210864543914795,
+      "learning_rate": 0.00029991813779713485,
+      "loss": 3.292288970947266,
+      "step": 625
+    },
+    {
+      "epoch": 0.12103746397694524,
+      "grad_norm": 1.7756609916687012,
+      "learning_rate": 0.0002999100780214057,
+      "loss": 3.291936492919922,
+      "step": 630
+    },
+    {
+      "epoch": 0.12199807877041306,
+      "grad_norm": 2.3280141353607178,
+      "learning_rate": 0.0002999016400071708,
+      "loss": 3.2992881774902343,
+      "step": 635
+    },
+    {
+      "epoch": 0.12295869356388088,
+      "grad_norm": 1.2652477025985718,
+      "learning_rate": 0.00029989282377572006,
+      "loss": 3.302785110473633,
+      "step": 640
+    },
+    {
+      "epoch": 0.1239193083573487,
+      "grad_norm": 1.1368358135223389,
+      "learning_rate": 0.00029988362934929786,
+      "loss": 3.298853302001953,
+      "step": 645
+    },
+    {
+      "epoch": 0.12487992315081652,
+      "grad_norm": 1.5556303262710571,
+      "learning_rate": 0.0002998740567511026,
+      "loss": 3.2962833404541017,
+      "step": 650
+    },
+    {
+      "epoch": 0.12584053794428435,
+      "grad_norm": 1.2470626831054688,
+      "learning_rate": 0.000299864106005287,
+      "loss": 3.296977996826172,
+      "step": 655
+    },
+    {
+      "epoch": 0.12680115273775217,
+      "grad_norm": 1.8531115055084229,
+      "learning_rate": 0.00029985377713695775,
+      "loss": 3.294302749633789,
+      "step": 660
+    },
+    {
+      "epoch": 0.12776176753121998,
+      "grad_norm": 1.524565577507019,
+      "learning_rate": 0.0002998430701721757,
+      "loss": 3.2964263916015626,
+      "step": 665
+    },
+    {
+      "epoch": 0.1287223823246878,
+      "grad_norm": 1.3688217401504517,
+      "learning_rate": 0.00029983198513795556,
+      "loss": 3.2987823486328125,
+      "step": 670
+    },
+    {
+      "epoch": 0.12968299711815562,
+      "grad_norm": 1.3249455690383911,
+      "learning_rate": 0.000299820522062266,
+      "loss": 3.293718719482422,
+      "step": 675
+    },
+    {
+      "epoch": 0.13064361191162344,
+      "grad_norm": 1.0875604152679443,
+      "learning_rate": 0.0002998086809740296,
+      "loss": 3.294192886352539,
+      "step": 680
+    },
+    {
+      "epoch": 0.13160422670509125,
+      "grad_norm": 1.4566972255706787,
+      "learning_rate": 0.00029979646190312264,
+      "loss": 3.2981857299804687,
+      "step": 685
+    },
+    {
+      "epoch": 0.13256484149855907,
+      "grad_norm": 1.506569743156433,
+      "learning_rate": 0.0002997838648803751,
+      "loss": 3.2970245361328123,
+      "step": 690
+    },
+    {
+      "epoch": 0.1335254562920269,
+      "grad_norm": 1.2062523365020752,
+      "learning_rate": 0.00029977088993757045,
+      "loss": 3.294635772705078,
+      "step": 695
+    },
+    {
+      "epoch": 0.1344860710854947,
+      "grad_norm": 1.4316554069519043,
+      "learning_rate": 0.0002997575371074458,
+      "loss": 3.294428253173828,
+      "step": 700
+    },
+    {
+      "epoch": 0.13544668587896252,
+      "grad_norm": 1.0954806804656982,
+      "learning_rate": 0.0002997438064236919,
+      "loss": 3.2926597595214844,
+      "step": 705
+    },
+    {
+      "epoch": 0.13640730067243034,
+      "grad_norm": 1.2398245334625244,
+      "learning_rate": 0.0002997296979209526,
+      "loss": 3.293048858642578,
+      "step": 710
+    },
+    {
+      "epoch": 0.1373679154658982,
+      "grad_norm": 1.7894244194030762,
+      "learning_rate": 0.00029971521163482494,
+      "loss": 3.2915740966796876,
+      "step": 715
+    },
+    {
+      "epoch": 0.138328530259366,
+      "grad_norm": 1.1798651218414307,
+      "learning_rate": 0.0002997003476018595,
+      "loss": 3.2928890228271483,
+      "step": 720
+    },
+    {
+      "epoch": 0.13928914505283382,
+      "grad_norm": 1.4055883884429932,
+      "learning_rate": 0.0002996851058595596,
+      "loss": 3.295462417602539,
+      "step": 725
+    },
+    {
+      "epoch": 0.14024975984630164,
+      "grad_norm": 1.3189594745635986,
+      "learning_rate": 0.00029966948644638184,
+      "loss": 3.2915382385253906,
+      "step": 730
+    },
+    {
+      "epoch": 0.14121037463976946,
+      "grad_norm": 2.0678322315216064,
+      "learning_rate": 0.0002996534894017356,
+      "loss": 3.2929306030273438,
+      "step": 735
+    },
+    {
+      "epoch": 0.14217098943323728,
+      "grad_norm": 1.0185853242874146,
+      "learning_rate": 0.000299637114765983,
+      "loss": 3.2973453521728517,
+      "step": 740
+    },
+    {
+      "epoch": 0.1431316042267051,
+      "grad_norm": 1.1556357145309448,
+      "learning_rate": 0.00029962036258043893,
+      "loss": 3.2935462951660157,
+      "step": 745
+    },
+    {
+      "epoch": 0.1440922190201729,
+      "grad_norm": 1.392240285873413,
+      "learning_rate": 0.00029960323288737097,
+      "loss": 3.2977447509765625,
+      "step": 750
+    },
+    {
+      "epoch": 0.14505283381364073,
+      "grad_norm": 1.6778427362442017,
+      "learning_rate": 0.000299585725729999,
+      "loss": 3.2953746795654295,
+      "step": 755
+    },
+    {
+      "epoch": 0.14601344860710855,
+      "grad_norm": 0.8692209720611572,
+      "learning_rate": 0.00029956784115249535,
+      "loss": 3.2932708740234373,
+      "step": 760
+    },
+    {
+      "epoch": 0.14697406340057637,
+      "grad_norm": 1.0066865682601929,
+      "learning_rate": 0.0002995495791999847,
+      "loss": 3.291333770751953,
+      "step": 765
+    },
+    {
+      "epoch": 0.14793467819404418,
+      "grad_norm": 1.1920307874679565,
+      "learning_rate": 0.0002995309399185439,
+      "loss": 3.2948654174804686,
+      "step": 770
+    },
+    {
+      "epoch": 0.148895292987512,
+      "grad_norm": 0.9732314348220825,
+      "learning_rate": 0.0002995119233552016,
+      "loss": 3.2924430847167967,
+      "step": 775
+    },
+    {
+      "epoch": 0.14985590778097982,
+      "grad_norm": 1.306164264678955,
+      "learning_rate": 0.0002994925295579386,
+      "loss": 3.2954177856445312,
+      "step": 780
+    },
+    {
+      "epoch": 0.15081652257444764,
+      "grad_norm": 1.2705974578857422,
+      "learning_rate": 0.0002994727585756875,
+      "loss": 3.291736602783203,
+      "step": 785
+    },
+    {
+      "epoch": 0.15177713736791545,
+      "grad_norm": 1.4319415092468262,
+      "learning_rate": 0.00029945261045833256,
+      "loss": 3.2872940063476563,
+      "step": 790
+    },
+    {
+      "epoch": 0.15273775216138327,
+      "grad_norm": 1.5740435123443604,
+      "learning_rate": 0.0002994320852567094,
+      "loss": 3.2921607971191404,
+      "step": 795
+    },
+    {
+      "epoch": 0.15369836695485112,
+      "grad_norm": 1.2682417631149292,
+      "learning_rate": 0.0002994111830226053,
+      "loss": 3.2847610473632813,
+      "step": 800
+    },
+    {
+      "epoch": 0.15465898174831894,
+      "grad_norm": 1.3516907691955566,
+      "learning_rate": 0.0002993899038087587,
+      "loss": 3.2877117156982423,
+      "step": 805
+    },
+    {
+      "epoch": 0.15561959654178675,
+      "grad_norm": 1.1541163921356201,
+      "learning_rate": 0.00029936824766885927,
+      "loss": 3.28739013671875,
+      "step": 810
+    },
+    {
+      "epoch": 0.15658021133525457,
+      "grad_norm": 1.4818735122680664,
+      "learning_rate": 0.0002993462146575477,
+      "loss": 3.2846580505371095,
+      "step": 815
+    },
+    {
+      "epoch": 0.1575408261287224,
+      "grad_norm": 1.0636875629425049,
+      "learning_rate": 0.00029932380483041547,
+      "loss": 3.2877876281738283,
+      "step": 820
+    },
+    {
+      "epoch": 0.1585014409221902,
+      "grad_norm": 1.1833419799804688,
+      "learning_rate": 0.00029930101824400486,
+      "loss": 3.289693069458008,
+      "step": 825
+    },
+    {
+      "epoch": 0.15946205571565802,
+      "grad_norm": 1.19170343875885,
+      "learning_rate": 0.00029927785495580887,
+      "loss": 3.283536911010742,
+      "step": 830
+    },
+    {
+      "epoch": 0.16042267050912584,
+      "grad_norm": 1.074803352355957,
+      "learning_rate": 0.00029925431502427077,
+      "loss": 3.2888442993164064,
+      "step": 835
+    },
+    {
+      "epoch": 0.16138328530259366,
+      "grad_norm": 1.205644130706787,
+      "learning_rate": 0.00029923039850878423,
+      "loss": 3.2893348693847657,
+      "step": 840
+    },
+    {
+      "epoch": 0.16234390009606148,
+      "grad_norm": 1.6655492782592773,
+      "learning_rate": 0.00029920610546969314,
+      "loss": 3.2857383728027343,
+      "step": 845
+    },
+    {
+      "epoch": 0.1633045148895293,
+      "grad_norm": 1.3908863067626953,
+      "learning_rate": 0.0002991814359682912,
+      "loss": 3.2931854248046877,
+      "step": 850
+    },
+    {
+      "epoch": 0.1642651296829971,
+      "grad_norm": 1.3854031562805176,
+      "learning_rate": 0.00029915639006682223,
+      "loss": 3.291329193115234,
+      "step": 855
+    },
+    {
+      "epoch": 0.16522574447646493,
+      "grad_norm": 1.3702648878097534,
+      "learning_rate": 0.0002991309678284795,
+      "loss": 3.283129119873047,
+      "step": 860
+    },
+    {
+      "epoch": 0.16618635926993275,
+      "grad_norm": 1.3857735395431519,
+      "learning_rate": 0.000299105169317406,
+      "loss": 3.285200500488281,
+      "step": 865
+    },
+    {
+      "epoch": 0.16714697406340057,
+      "grad_norm": 1.0670974254608154,
+      "learning_rate": 0.000299078994598694,
+      "loss": 3.287001037597656,
+      "step": 870
+    },
+    {
+      "epoch": 0.16810758885686838,
+      "grad_norm": 1.2670496702194214,
+      "learning_rate": 0.000299052443738385,
+      "loss": 3.28797607421875,
+      "step": 875
+    },
+    {
+      "epoch": 0.1690682036503362,
+      "grad_norm": 1.4439146518707275,
+      "learning_rate": 0.00029902551680346956,
+      "loss": 3.2858448028564453,
+      "step": 880
+    },
+    {
+      "epoch": 0.17002881844380405,
+      "grad_norm": 1.4315037727355957,
+      "learning_rate": 0.000298998213861887,
+      "loss": 3.28741340637207,
+      "step": 885
+    },
+    {
+      "epoch": 0.17098943323727187,
+      "grad_norm": 1.227054238319397,
+      "learning_rate": 0.0002989705349825256,
+      "loss": 3.285395050048828,
+      "step": 890
+    },
+    {
+      "epoch": 0.17195004803073968,
+      "grad_norm": 1.9081028699874878,
+      "learning_rate": 0.0002989424802352218,
+      "loss": 3.2881591796875,
+      "step": 895
+    },
+    {
+      "epoch": 0.1729106628242075,
+      "grad_norm": 1.5766661167144775,
+      "learning_rate": 0.0002989140496907609,
+      "loss": 3.282776641845703,
+      "step": 900
+    },
+    {
+      "epoch": 0.17387127761767532,
+      "grad_norm": 1.3651994466781616,
+      "learning_rate": 0.00029888524342087583,
+      "loss": 3.288460540771484,
+      "step": 905
+    },
+    {
+      "epoch": 0.17483189241114314,
+      "grad_norm": 1.1733394861221313,
+      "learning_rate": 0.00029885606149824786,
+      "loss": 3.2853572845458983,
+      "step": 910
+    },
+    {
+      "epoch": 0.17579250720461095,
+      "grad_norm": 1.798338770866394,
+      "learning_rate": 0.00029882650399650596,
+      "loss": 3.2855796813964844,
+      "step": 915
+    },
+    {
+      "epoch": 0.17675312199807877,
+      "grad_norm": 1.5101557970046997,
+      "learning_rate": 0.00029879657099022677,
+      "loss": 3.2913845062255858,
+      "step": 920
+    },
+    {
+      "epoch": 0.1777137367915466,
+      "grad_norm": 1.1141709089279175,
+      "learning_rate": 0.0002987662625549343,
+      "loss": 3.2880714416503904,
+      "step": 925
+    },
+    {
+      "epoch": 0.1786743515850144,
+      "grad_norm": 1.4672577381134033,
+      "learning_rate": 0.00029873557876709984,
+      "loss": 3.2870162963867187,
+      "step": 930
+    },
+    {
+      "epoch": 0.17963496637848222,
+      "grad_norm": 1.2008938789367676,
+      "learning_rate": 0.00029870451970414176,
+      "loss": 3.281757354736328,
+      "step": 935
+    },
+    {
+      "epoch": 0.18059558117195004,
+      "grad_norm": 1.4659417867660522,
+      "learning_rate": 0.00029867308544442525,
+      "loss": 3.2835357666015623,
+      "step": 940
+    },
+    {
+      "epoch": 0.18155619596541786,
+      "grad_norm": 1.200221300125122,
+      "learning_rate": 0.0002986412760672622,
+      "loss": 3.2824310302734374,
+      "step": 945
+    },
+    {
+      "epoch": 0.18251681075888568,
+      "grad_norm": 1.1019620895385742,
+      "learning_rate": 0.0002986090916529109,
+      "loss": 3.2842033386230467,
+      "step": 950
+    },
+    {
+      "epoch": 0.1834774255523535,
+      "grad_norm": 1.3469750881195068,
+      "learning_rate": 0.00029857653228257587,
+      "loss": 3.285472106933594,
+      "step": 955
+    },
+    {
+      "epoch": 0.1844380403458213,
+      "grad_norm": 1.3108643293380737,
+      "learning_rate": 0.0002985435980384078,
+      "loss": 3.285649871826172,
+      "step": 960
+    },
+    {
+      "epoch": 0.18539865513928913,
+      "grad_norm": 1.452244758605957,
+      "learning_rate": 0.0002985102890035032,
+      "loss": 3.2803878784179688,
+      "step": 965
+    },
+    {
+      "epoch": 0.18635926993275698,
+      "grad_norm": 1.229927897453308,
+      "learning_rate": 0.0002984766052619041,
+      "loss": 3.283854675292969,
+      "step": 970
+    },
+    {
+      "epoch": 0.1873198847262248,
+      "grad_norm": 1.1009323596954346,
+      "learning_rate": 0.0002984425468985982,
+      "loss": 3.2806129455566406,
+      "step": 975
+    },
+    {
+      "epoch": 0.1882804995196926,
+      "grad_norm": 1.5917284488677979,
+      "learning_rate": 0.00029840811399951806,
+      "loss": 3.282550048828125,
+      "step": 980
+    },
+    {
+      "epoch": 0.18924111431316043,
+      "grad_norm": 1.1452780961990356,
+      "learning_rate": 0.00029837330665154156,
+      "loss": 3.2815521240234373,
+      "step": 985
+    },
+    {
+      "epoch": 0.19020172910662825,
+      "grad_norm": 0.9764254093170166,
+      "learning_rate": 0.000298338124942491,
+      "loss": 3.284914016723633,
+      "step": 990
+    },
+    {
+      "epoch": 0.19116234390009607,
+      "grad_norm": 1.045482873916626,
+      "learning_rate": 0.0002983025689611337,
+      "loss": 3.2865413665771483,
+      "step": 995
+    },
+    {
+      "epoch": 0.19212295869356388,
+      "grad_norm": 1.3126083612442017,
+      "learning_rate": 0.0002982666387971809,
+      "loss": 3.281140899658203,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1930835734870317,
+      "grad_norm": 1.3957022428512573,
+      "learning_rate": 0.0002982303345412881,
+      "loss": 3.2785770416259767,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19404418828049952,
+      "grad_norm": 1.2035447359085083,
+      "learning_rate": 0.0002981936562850547,
+      "loss": 3.279229736328125,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19500480307396734,
+      "grad_norm": 0.8585781455039978,
+      "learning_rate": 0.00029815660412102366,
+      "loss": 3.2802391052246094,
+      "step": 1015
+    },
+    {
+      "epoch": 0.19596541786743515,
+      "grad_norm": 1.203058123588562,
+      "learning_rate": 0.00029811917814268144,
+      "loss": 3.2790596008300783,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19692603266090297,
+      "grad_norm": 1.4647698402404785,
+      "learning_rate": 0.0002980813784444576,
+      "loss": 3.2819541931152343,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1978866474543708,
+      "grad_norm": 1.4942032098770142,
+      "learning_rate": 0.0002980432051217246,
+      "loss": 3.2802181243896484,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1988472622478386,
+      "grad_norm": 1.1856545209884644,
+      "learning_rate": 0.0002980046582707978,
+      "loss": 3.290009307861328,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19980787704130643,
+      "grad_norm": 1.166818380355835,
+      "learning_rate": 0.00029796573798893464,
+      "loss": 3.28170166015625,
+      "step": 1040
+    },
+    {
+      "epoch": 0.20076849183477424,
+      "grad_norm": 1.310017466545105,
+      "learning_rate": 0.0002979264443743352,
+      "loss": 3.282215118408203,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2017291066282421,
+      "grad_norm": 1.7260593175888062,
+      "learning_rate": 0.00029788677752614116,
+      "loss": 3.2847213745117188,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2026897214217099,
+      "grad_norm": 1.083449125289917,
+      "learning_rate": 0.0002978467375444361,
+      "loss": 3.278899002075195,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20365033621517772,
+      "grad_norm": 1.2556031942367554,
+      "learning_rate": 0.000297806324530245,
+      "loss": 3.280004119873047,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20461095100864554,
+      "grad_norm": 0.8801208138465881,
+      "learning_rate": 0.00029776553858553405,
+      "loss": 3.2757003784179686,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20557156580211336,
+      "grad_norm": 1.419739007949829,
+      "learning_rate": 0.0002977243798132103,
+      "loss": 3.2807285308837892,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20653218059558118,
+      "grad_norm": 1.3385144472122192,
+      "learning_rate": 0.00029768284831712175,
+      "loss": 3.280821990966797,
+      "step": 1075
+    },
+    {
+      "epoch": 0.207492795389049,
+      "grad_norm": 1.0410194396972656,
+      "learning_rate": 0.0002976409442020564,
+      "loss": 3.279698944091797,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2084534101825168,
+      "grad_norm": 1.2109322547912598,
+      "learning_rate": 0.0002975986675737427,
+      "loss": 3.2776737213134766,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20941402497598463,
+      "grad_norm": 1.598090648651123,
+      "learning_rate": 0.00029755601853884893,
+      "loss": 3.277222442626953,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21037463976945245,
+      "grad_norm": 1.0197858810424805,
+      "learning_rate": 0.000297512997204983,
+      "loss": 3.2803329467773437,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21133525456292027,
+      "grad_norm": 0.9741032123565674,
+      "learning_rate": 0.0002974696036806922,
+      "loss": 3.2853607177734374,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21229586935638808,
+      "grad_norm": 1.3451876640319824,
+      "learning_rate": 0.00029742583807546274,
+      "loss": 3.276333236694336,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2132564841498559,
+      "grad_norm": 1.2260080575942993,
+      "learning_rate": 0.00029738170049971974,
+      "loss": 3.2778556823730467,
+      "step": 1110
+    },
+    {
+      "epoch": 0.21421709894332372,
+      "grad_norm": 1.4238629341125488,
+      "learning_rate": 0.00029733719106482694,
+      "loss": 3.27607421875,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21517771373679154,
+      "grad_norm": 1.0617904663085938,
+      "learning_rate": 0.0002972923098830861,
+      "loss": 3.278319549560547,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21613832853025935,
+      "grad_norm": 1.3632889986038208,
+      "learning_rate": 0.0002972470570677371,
+      "loss": 3.2795860290527346,
+      "step": 1125
+    },
+    {
+      "epoch": 0.21709894332372717,
+      "grad_norm": 1.2691421508789062,
+      "learning_rate": 0.00029720143273295735,
+      "loss": 3.277021026611328,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21805955811719502,
+      "grad_norm": 1.2249642610549927,
+      "learning_rate": 0.0002971554369938619,
+      "loss": 3.274774169921875,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21902017291066284,
+      "grad_norm": 1.6009410619735718,
+      "learning_rate": 0.0002971090699665025,
+      "loss": 3.2758060455322267,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21998078770413065,
+      "grad_norm": 1.1862974166870117,
+      "learning_rate": 0.0002970623317678681,
+      "loss": 3.278969955444336,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22094140249759847,
+      "grad_norm": 1.1896966695785522,
+      "learning_rate": 0.00029701522251588395,
+      "loss": 3.277644729614258,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2219020172910663,
+      "grad_norm": 1.2047041654586792,
+      "learning_rate": 0.00029696774232941135,
+      "loss": 3.276856231689453,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2228626320845341,
+      "grad_norm": 1.4511812925338745,
+      "learning_rate": 0.0002969198913282479,
+      "loss": 3.2762985229492188,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22382324687800192,
+      "grad_norm": 1.0419248342514038,
+      "learning_rate": 0.0002968716696331265,
+      "loss": 3.2781463623046876,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22478386167146974,
+      "grad_norm": 1.0537165403366089,
+      "learning_rate": 0.00029682307736571533,
+      "loss": 3.280584716796875,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22574447646493756,
+      "grad_norm": 0.9234645366668701,
+      "learning_rate": 0.00029677411464861783,
+      "loss": 3.2776565551757812,
+      "step": 1175
+    },
+    {
+      "epoch": 0.22670509125840538,
+      "grad_norm": 1.0644463300704956,
+      "learning_rate": 0.00029672478160537184,
+      "loss": 3.2758762359619142,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2276657060518732,
+      "grad_norm": 0.9237310290336609,
+      "learning_rate": 0.00029667507836044984,
+      "loss": 3.2722846984863283,
+      "step": 1185
+    },
+    {
+      "epoch": 0.228626320845341,
+      "grad_norm": 1.5153244733810425,
+      "learning_rate": 0.00029662500503925803,
+      "loss": 3.2741416931152343,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22958693563880883,
+      "grad_norm": 1.3842270374298096,
+      "learning_rate": 0.00029657456176813664,
+      "loss": 3.2764862060546873,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23054755043227665,
+      "grad_norm": 1.1134364604949951,
+      "learning_rate": 0.0002965237486743592,
+      "loss": 3.275590515136719,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23150816522574447,
+      "grad_norm": 1.4811969995498657,
+      "learning_rate": 0.0002964725658861323,
+      "loss": 3.2718353271484375,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23246878001921228,
+      "grad_norm": 1.2888314723968506,
+      "learning_rate": 0.0002964210135325955,
+      "loss": 3.2731605529785157,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2334293948126801,
+      "grad_norm": 1.530621886253357,
+      "learning_rate": 0.0002963690917438206,
+      "loss": 3.2751121520996094,
+      "step": 1215
+    },
+    {
+      "epoch": 0.23439000960614795,
+      "grad_norm": 1.0951229333877563,
+      "learning_rate": 0.0002963168006508116,
+      "loss": 3.2797927856445312,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23535062439961577,
+      "grad_norm": 1.0309784412384033,
+      "learning_rate": 0.00029626414038550424,
+      "loss": 3.279781723022461,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23631123919308358,
+      "grad_norm": 1.211702823638916,
+      "learning_rate": 0.00029621111108076584,
+      "loss": 3.276560592651367,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2372718539865514,
+      "grad_norm": 1.0704128742218018,
+      "learning_rate": 0.0002961577128703948,
+      "loss": 3.2736587524414062,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23823246878001922,
+      "grad_norm": 1.4519094228744507,
+      "learning_rate": 0.0002961039458891202,
+      "loss": 3.274006652832031,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23919308357348704,
+      "grad_norm": 1.2808977365493774,
+      "learning_rate": 0.00029604981027260167,
+      "loss": 3.27347412109375,
+      "step": 1245
+    },
+    {
+      "epoch": 0.24015369836695485,
+      "grad_norm": 0.8584306836128235,
+      "learning_rate": 0.00029599530615742896,
+      "loss": 3.2714912414550783,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24111431316042267,
+      "grad_norm": 1.1275231838226318,
+      "learning_rate": 0.0002959404336811215,
+      "loss": 3.2730926513671874,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2420749279538905,
+      "grad_norm": 1.422745943069458,
+      "learning_rate": 0.0002958851929821283,
+      "loss": 3.27635498046875,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2430355427473583,
+      "grad_norm": 1.0575655698776245,
+      "learning_rate": 0.00029582958419982713,
+      "loss": 3.272149658203125,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24399615754082613,
+      "grad_norm": 1.2576707601547241,
+      "learning_rate": 0.0002957736074745248,
+      "loss": 3.275560760498047,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24495677233429394,
+      "grad_norm": 1.328376054763794,
+      "learning_rate": 0.0002957172629474562,
+      "loss": 3.272699737548828,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24591738712776176,
+      "grad_norm": 1.0527936220169067,
+      "learning_rate": 0.00029566055076078447,
+      "loss": 3.269004058837891,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24687800192122958,
+      "grad_norm": 1.1557375192642212,
+      "learning_rate": 0.00029560347105760023,
+      "loss": 3.2709724426269533,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2478386167146974,
+      "grad_norm": 1.1958216428756714,
+      "learning_rate": 0.00029554602398192137,
+      "loss": 3.276508331298828,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24879923150816521,
+      "grad_norm": 1.0567305088043213,
+      "learning_rate": 0.00029548820967869287,
+      "loss": 3.270316314697266,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24975984630163303,
+      "grad_norm": 1.1860175132751465,
+      "learning_rate": 0.00029543002829378605,
+      "loss": 3.274994659423828,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2507204610951009,
+      "grad_norm": 1.2902470827102661,
+      "learning_rate": 0.0002953714799739986,
+      "loss": 3.2740650177001953,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2516810758885687,
+      "grad_norm": 1.0249444246292114,
+      "learning_rate": 0.0002953125648670538,
+      "loss": 3.273143768310547,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2526416906820365,
+      "grad_norm": 1.051010251045227,
+      "learning_rate": 0.00029525328312160064,
+      "loss": 3.270255279541016,
+      "step": 1315
+    },
+    {
+      "epoch": 0.25360230547550433,
+      "grad_norm": 1.5545085668563843,
+      "learning_rate": 0.00029519363488721297,
+      "loss": 3.270277404785156,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25456292026897215,
+      "grad_norm": 1.096474289894104,
+      "learning_rate": 0.0002951336203143895,
+      "loss": 3.2695423126220704,
+      "step": 1325
+    },
+    {
+      "epoch": 0.25552353506243997,
+      "grad_norm": 1.3266576528549194,
+      "learning_rate": 0.0002950732395545531,
+      "loss": 3.268444061279297,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2564841498559078,
+      "grad_norm": 1.3308000564575195,
+      "learning_rate": 0.0002950124927600505,
+      "loss": 3.270121765136719,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2574447646493756,
+      "grad_norm": 0.9661091566085815,
+      "learning_rate": 0.0002949513800841523,
+      "loss": 3.2689208984375,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2584053794428434,
+      "grad_norm": 1.0783677101135254,
+      "learning_rate": 0.0002948899016810519,
+      "loss": 3.2726470947265627,
+      "step": 1345
+    },
+    {
+      "epoch": 0.25936599423631124,
+      "grad_norm": 1.44166898727417,
+      "learning_rate": 0.00029482805770586566,
+      "loss": 3.2676158905029298,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26032660902977905,
+      "grad_norm": 0.9048498868942261,
+      "learning_rate": 0.0002947658483146323,
+      "loss": 3.268976593017578,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2612872238232469,
+      "grad_norm": 0.9976827502250671,
+      "learning_rate": 0.00029470327366431257,
+      "loss": 3.2651641845703123,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2622478386167147,
+      "grad_norm": 1.4668549299240112,
+      "learning_rate": 0.00029464033391278857,
+      "loss": 3.267919158935547,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2632084534101825,
+      "grad_norm": 1.3420720100402832,
+      "learning_rate": 0.0002945770292188638,
+      "loss": 3.2697959899902345,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2641690682036503,
+      "grad_norm": 1.3224592208862305,
+      "learning_rate": 0.00029451335974226255,
+      "loss": 3.2723613739013673,
+      "step": 1375
+    },
+    {
+      "epoch": 0.26512968299711814,
+      "grad_norm": 1.1924461126327515,
+      "learning_rate": 0.0002944493256436294,
+      "loss": 3.270499038696289,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26609029779058596,
+      "grad_norm": 1.2803032398223877,
+      "learning_rate": 0.000294384927084529,
+      "loss": 3.269664001464844,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2670509125840538,
+      "grad_norm": 1.0187225341796875,
+      "learning_rate": 0.0002943201642274455,
+      "loss": 3.2647491455078126,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2680115273775216,
+      "grad_norm": 1.3384113311767578,
+      "learning_rate": 0.00029425503723578216,
+      "loss": 3.2721004486083984,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2689721421709894,
+      "grad_norm": 1.2812883853912354,
+      "learning_rate": 0.00029418954627386115,
+      "loss": 3.269911193847656,
+      "step": 1400
+    },
+    {
+      "epoch": 0.26993275696445723,
+      "grad_norm": 1.1734299659729004,
+      "learning_rate": 0.0002941236915069228,
+      "loss": 3.2672080993652344,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27089337175792505,
+      "grad_norm": 1.069024682044983,
+      "learning_rate": 0.00029405747310112557,
+      "loss": 3.2702381134033205,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27185398655139287,
+      "grad_norm": 1.286638617515564,
+      "learning_rate": 0.0002939908912235452,
+      "loss": 3.2697906494140625,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2728146013448607,
+      "grad_norm": 1.1734787225723267,
+      "learning_rate": 0.0002939239460421746,
+      "loss": 3.2664981842041017,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2737752161383285,
+      "grad_norm": 1.150191307067871,
+      "learning_rate": 0.00029385663772592334,
+      "loss": 3.2677589416503907,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2747358309317964,
+      "grad_norm": 1.3920680284500122,
+      "learning_rate": 0.00029378896644461716,
+      "loss": 3.2696159362792967,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2756964457252642,
+      "grad_norm": 1.1501007080078125,
+      "learning_rate": 0.0002937209323689978,
+      "loss": 3.268462371826172,
+      "step": 1435
+    },
+    {
+      "epoch": 0.276657060518732,
+      "grad_norm": 1.2256152629852295,
+      "learning_rate": 0.00029365253567072207,
+      "loss": 3.269298553466797,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27761767531219983,
+      "grad_norm": 1.2467095851898193,
+      "learning_rate": 0.00029358377652236186,
+      "loss": 3.2696155548095702,
+      "step": 1445
+    },
+    {
+      "epoch": 0.27857829010566765,
+      "grad_norm": 0.7786490321159363,
+      "learning_rate": 0.00029351465509740366,
+      "loss": 3.265594482421875,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27953890489913547,
+      "grad_norm": 0.9412602186203003,
+      "learning_rate": 0.00029344517157024773,
+      "loss": 3.2679397583007814,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2804995196926033,
+      "grad_norm": 1.1962471008300781,
+      "learning_rate": 0.0002933753261162084,
+      "loss": 3.276123046875,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2814601344860711,
+      "grad_norm": 1.1592360734939575,
+      "learning_rate": 0.00029330511891151263,
+      "loss": 3.2704566955566405,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2824207492795389,
+      "grad_norm": 0.8931114673614502,
+      "learning_rate": 0.00029323455013330064,
+      "loss": 3.2649574279785156,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28338136407300674,
+      "grad_norm": 0.8980585932731628,
+      "learning_rate": 0.00029316361995962453,
+      "loss": 3.2683528900146483,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28434197886647455,
+      "grad_norm": 1.330986499786377,
+      "learning_rate": 0.0002930923285694485,
+      "loss": 3.2716033935546873,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28530259365994237,
+      "grad_norm": 1.0797970294952393,
+      "learning_rate": 0.000293020676142648,
+      "loss": 3.268834686279297,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2862632084534102,
+      "grad_norm": 1.2156567573547363,
+      "learning_rate": 0.00029294866286000946,
+      "loss": 3.2620380401611326,
+      "step": 1490
+    },
+    {
+      "epoch": 0.287223823246878,
+      "grad_norm": 1.3845075368881226,
+      "learning_rate": 0.00029287628890322986,
+      "loss": 3.2627891540527343,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2881844380403458,
+      "grad_norm": 1.2039449214935303,
+      "learning_rate": 0.00029280355445491596,
+      "loss": 3.2676162719726562,
+      "step": 1500
+    },
+    {
+      "epoch": 0.28914505283381364,
+      "grad_norm": 1.2778565883636475,
+      "learning_rate": 0.00029273045969858437,
+      "loss": 3.267996978759766,
+      "step": 1505
+    },
+    {
+      "epoch": 0.29010566762728146,
+      "grad_norm": 0.9892801642417908,
+      "learning_rate": 0.0002926570048186606,
+      "loss": 3.2660484313964844,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2910662824207493,
+      "grad_norm": 1.0218427181243896,
+      "learning_rate": 0.00029258319000047885,
+      "loss": 3.269457244873047,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2920268972142171,
+      "grad_norm": 1.0072425603866577,
+      "learning_rate": 0.00029250901543028144,
+      "loss": 3.267464828491211,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2929875120076849,
+      "grad_norm": 1.4437980651855469,
+      "learning_rate": 0.00029243448129521847,
+      "loss": 3.2681037902832033,
+      "step": 1525
+    },
+    {
+      "epoch": 0.29394812680115273,
+      "grad_norm": 0.8555430769920349,
+      "learning_rate": 0.0002923595877833472,
+      "loss": 3.2668888092041017,
+      "step": 1530
+    },
+    {
+      "epoch": 0.29490874159462055,
+      "grad_norm": 1.2450693845748901,
+      "learning_rate": 0.0002922843350836317,
+      "loss": 3.2664356231689453,
+      "step": 1535
+    },
+    {
+      "epoch": 0.29586935638808837,
+      "grad_norm": 0.9553253650665283,
+      "learning_rate": 0.00029220872338594215,
+      "loss": 3.263041687011719,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2968299711815562,
+      "grad_norm": 1.2076423168182373,
+      "learning_rate": 0.0002921327528810547,
+      "loss": 3.266950988769531,
+      "step": 1545
+    },
+    {
+      "epoch": 0.297790585975024,
+      "grad_norm": 2.1191205978393555,
+      "learning_rate": 0.00029205642376065066,
+      "loss": 3.2656558990478515,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2987512007684918,
+      "grad_norm": 1.4229607582092285,
+      "learning_rate": 0.0002919797362173163,
+      "loss": 3.2715381622314452,
+      "step": 1555
+    },
+    {
+      "epoch": 0.29971181556195964,
+      "grad_norm": 1.2368606328964233,
+      "learning_rate": 0.00029190269044454206,
+      "loss": 3.2648460388183596,
+      "step": 1560
+    },
+    {
+      "epoch": 0.30067243035542746,
+      "grad_norm": 1.266961693763733,
+      "learning_rate": 0.00029182528663672245,
+      "loss": 3.2667465209960938,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3016330451488953,
+      "grad_norm": 0.9238030314445496,
+      "learning_rate": 0.00029174752498915515,
+      "loss": 3.262348175048828,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3025936599423631,
+      "grad_norm": 1.0845513343811035,
+      "learning_rate": 0.0002916694056980408,
+      "loss": 3.266324996948242,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3035542747358309,
+      "grad_norm": 1.050512433052063,
+      "learning_rate": 0.0002915909289604823,
+      "loss": 3.2667007446289062,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3045148895292987,
+      "grad_norm": 1.25586998462677,
+      "learning_rate": 0.00029151209497448467,
+      "loss": 3.262495422363281,
+      "step": 1585
+    },
+    {
+      "epoch": 0.30547550432276654,
+      "grad_norm": 1.1228755712509155,
+      "learning_rate": 0.000291432903938954,
+      "loss": 3.26649169921875,
+      "step": 1590
+    },
+    {
+      "epoch": 0.30643611911623436,
+      "grad_norm": 1.2967936992645264,
+      "learning_rate": 0.0002913533560536975,
+      "loss": 3.264292526245117,
+      "step": 1595
+    },
+    {
+      "epoch": 0.30739673390970224,
+      "grad_norm": 1.1702841520309448,
+      "learning_rate": 0.00029127345151942253,
+      "loss": 3.2614276885986326,
+      "step": 1600
+    },
+    {
+      "epoch": 0.30835734870317005,
+      "grad_norm": 1.438833236694336,
+      "learning_rate": 0.00029119319053773655,
+      "loss": 3.2649909973144533,
+      "step": 1605
+    },
+    {
+      "epoch": 0.30931796349663787,
+      "grad_norm": 1.1444823741912842,
+      "learning_rate": 0.00029111257331114617,
+      "loss": 3.2618698120117187,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3102785782901057,
+      "grad_norm": 1.2421047687530518,
+      "learning_rate": 0.00029103160004305693,
+      "loss": 3.260860061645508,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3112391930835735,
+      "grad_norm": 1.3885422945022583,
+      "learning_rate": 0.00029095027093777276,
+      "loss": 3.2636131286621093,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3121998078770413,
+      "grad_norm": 1.403899073600769,
+      "learning_rate": 0.00029086858620049527,
+      "loss": 3.261060333251953,
+      "step": 1625
+    },
+    {
+      "epoch": 0.31316042267050914,
+      "grad_norm": 1.4193172454833984,
+      "learning_rate": 0.0002907865460373234,
+      "loss": 3.2587600708007813,
+      "step": 1630
+    },
+    {
+      "epoch": 0.31412103746397696,
+      "grad_norm": 0.9989815354347229,
+      "learning_rate": 0.00029070415065525295,
+      "loss": 3.2618736267089843,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3150816522574448,
+      "grad_norm": 1.0307117700576782,
+      "learning_rate": 0.00029062140026217596,
+      "loss": 3.25679931640625,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3160422670509126,
+      "grad_norm": 1.3937115669250488,
+      "learning_rate": 0.0002905382950668801,
+      "loss": 3.2610740661621094,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3170028818443804,
+      "grad_norm": 1.0371273756027222,
+      "learning_rate": 0.0002904548352790483,
+      "loss": 3.264708709716797,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31796349663784823,
+      "grad_norm": 1.202774167060852,
+      "learning_rate": 0.0002903710211092582,
+      "loss": 3.261525344848633,
+      "step": 1655
+    },
+    {
+      "epoch": 0.31892411143131605,
+      "grad_norm": 1.2215559482574463,
+      "learning_rate": 0.0002902868527689816,
+      "loss": 3.264019012451172,
+      "step": 1660
+    },
+    {
+      "epoch": 0.31988472622478387,
+      "grad_norm": 0.8910490274429321,
+      "learning_rate": 0.0002902023304705837,
+      "loss": 3.262919616699219,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3208453410182517,
+      "grad_norm": 1.35845148563385,
+      "learning_rate": 0.00029011745442732303,
+      "loss": 3.2655372619628906,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3218059558117195,
+      "grad_norm": 1.354686975479126,
+      "learning_rate": 0.0002900322248533506,
+      "loss": 3.26285400390625,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3227665706051873,
+      "grad_norm": 1.0969462394714355,
+      "learning_rate": 0.0002899466419637093,
+      "loss": 3.257940673828125,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32372718539865514,
+      "grad_norm": 1.291715145111084,
+      "learning_rate": 0.00028986070597433354,
+      "loss": 3.2633174896240233,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32468780019212296,
+      "grad_norm": 1.4123157262802124,
+      "learning_rate": 0.00028977441710204867,
+      "loss": 3.2593170166015626,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3256484149855908,
+      "grad_norm": 1.1881605386734009,
+      "learning_rate": 0.0002896877755645703,
+      "loss": 3.2590873718261717,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3266090297790586,
+      "grad_norm": 1.043522596359253,
+      "learning_rate": 0.00028960078158050403,
+      "loss": 3.258655548095703,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3275696445725264,
+      "grad_norm": 1.0990368127822876,
+      "learning_rate": 0.0002895134353693445,
+      "loss": 3.258600616455078,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3285302593659942,
+      "grad_norm": 1.4628812074661255,
+      "learning_rate": 0.0002894257371514752,
+      "loss": 3.262346649169922,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32949087415946204,
+      "grad_norm": 0.8175418376922607,
+      "learning_rate": 0.0002893376871481678,
+      "loss": 3.255454254150391,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33045148895292986,
+      "grad_norm": 1.0393476486206055,
+      "learning_rate": 0.0002892492855815814,
+      "loss": 3.2613197326660157,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3314121037463977,
+      "grad_norm": 1.250258445739746,
+      "learning_rate": 0.0002891605326747622,
+      "loss": 3.260877227783203,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3323727185398655,
+      "grad_norm": 0.8695818781852722,
+      "learning_rate": 0.00028907142865164305,
+      "loss": 3.259082794189453,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.0892332792282104,
+      "learning_rate": 0.00028898197373704234,
+      "loss": 3.261730194091797,
+      "step": 1735
+    },
+    {
+      "epoch": 0.33429394812680113,
+      "grad_norm": 1.2874714136123657,
+      "learning_rate": 0.00028889216815666406,
+      "loss": 3.2563186645507813,
+      "step": 1740
+    },
+    {
+      "epoch": 0.33525456292026895,
+      "grad_norm": 1.2988804578781128,
+      "learning_rate": 0.000288802012137097,
+      "loss": 3.2569385528564454,
+      "step": 1745
+    },
+    {
+      "epoch": 0.33621517771373677,
+      "grad_norm": 1.2552379369735718,
+      "learning_rate": 0.0002887115059058139,
+      "loss": 3.2592254638671876,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3371757925072046,
+      "grad_norm": 0.9785555601119995,
+      "learning_rate": 0.0002886206496911714,
+      "loss": 3.257631301879883,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3381364073006724,
+      "grad_norm": 1.3783308267593384,
+      "learning_rate": 0.00028852944372240896,
+      "loss": 3.263458251953125,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3390970220941403,
+      "grad_norm": 1.2814546823501587,
+      "learning_rate": 0.00028843788822964864,
+      "loss": 3.2578887939453125,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3400576368876081,
+      "grad_norm": 1.2692376375198364,
+      "learning_rate": 0.00028834598344389437,
+      "loss": 3.2599555969238283,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3410182516810759,
+      "grad_norm": 1.3274823427200317,
+      "learning_rate": 0.0002882537295970312,
+      "loss": 3.261073684692383,
+      "step": 1775
+    },
+    {
+      "epoch": 0.34197886647454373,
+      "grad_norm": 1.1028993129730225,
+      "learning_rate": 0.00028816112692182526,
+      "loss": 3.2600685119628907,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34293948126801155,
+      "grad_norm": 1.0940430164337158,
+      "learning_rate": 0.0002880681756519224,
+      "loss": 3.2631156921386717,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34390009606147937,
+      "grad_norm": 1.1741334199905396,
+      "learning_rate": 0.0002879748760218484,
+      "loss": 3.2562606811523436,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3448607108549472,
+      "grad_norm": 1.4189056158065796,
+      "learning_rate": 0.0002878812282670077,
+      "loss": 3.2629486083984376,
+      "step": 1795
+    },
+    {
+      "epoch": 0.345821325648415,
+      "grad_norm": 0.9202475547790527,
+      "learning_rate": 0.00028778723262368313,
+      "loss": 3.258461761474609,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3467819404418828,
+      "grad_norm": 1.0866518020629883,
+      "learning_rate": 0.00028769288932903547,
+      "loss": 3.2561798095703125,
+      "step": 1805
+    },
+    {
+      "epoch": 0.34774255523535064,
+      "grad_norm": 1.1435341835021973,
+      "learning_rate": 0.00028759819862110244,
+      "loss": 3.253205108642578,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34870317002881845,
+      "grad_norm": 0.9997341632843018,
+      "learning_rate": 0.00028750316073879845,
+      "loss": 3.256957244873047,
+      "step": 1815
+    },
+    {
+      "epoch": 0.34966378482228627,
+      "grad_norm": 1.5304245948791504,
+      "learning_rate": 0.0002874077759219138,
+      "loss": 3.253832244873047,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3506243996157541,
+      "grad_norm": 1.3517911434173584,
+      "learning_rate": 0.0002873120444111142,
+      "loss": 3.2599609375,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3515850144092219,
+      "grad_norm": 1.1517263650894165,
+      "learning_rate": 0.0002872159664479401,
+      "loss": 3.253156280517578,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3525456292026897,
+      "grad_norm": 1.2207057476043701,
+      "learning_rate": 0.00028711954227480596,
+      "loss": 3.255885696411133,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35350624399615754,
+      "grad_norm": 1.4356719255447388,
+      "learning_rate": 0.00028702277213499993,
+      "loss": 3.2568824768066404,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35446685878962536,
+      "grad_norm": 0.9663187265396118,
+      "learning_rate": 0.000286925656272683,
+      "loss": 3.2607173919677734,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3554274735830932,
+      "grad_norm": 1.0548568964004517,
+      "learning_rate": 0.0002868281949328884,
+      "loss": 3.254198455810547,
+      "step": 1850
+    },
+    {
+      "epoch": 0.356388088376561,
+      "grad_norm": 1.216261386871338,
+      "learning_rate": 0.0002867303883615212,
+      "loss": 3.25797119140625,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3573487031700288,
+      "grad_norm": 0.9959939122200012,
+      "learning_rate": 0.00028663223680535727,
+      "loss": 3.2549095153808594,
+      "step": 1860
+    },
+    {
+      "epoch": 0.35830931796349663,
+      "grad_norm": 1.0812616348266602,
+      "learning_rate": 0.0002865337405120432,
+      "loss": 3.2560733795166015,
+      "step": 1865
+    },
+    {
+      "epoch": 0.35926993275696445,
+      "grad_norm": 1.2685303688049316,
+      "learning_rate": 0.0002864348997300951,
+      "loss": 3.2555168151855467,
+      "step": 1870
+    },
+    {
+      "epoch": 0.36023054755043227,
+      "grad_norm": 1.0879360437393188,
+      "learning_rate": 0.0002863357147088985,
+      "loss": 3.255245590209961,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3611911623439001,
+      "grad_norm": 1.0502004623413086,
+      "learning_rate": 0.0002862361856987073,
+      "loss": 3.260142517089844,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3621517771373679,
+      "grad_norm": 1.2763992547988892,
+      "learning_rate": 0.0002861363129506435,
+      "loss": 3.2576175689697267,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3631123919308357,
+      "grad_norm": 1.2673193216323853,
+      "learning_rate": 0.0002860360967166963,
+      "loss": 3.259668731689453,
+      "step": 1890
+    },
+    {
+      "epoch": 0.36407300672430354,
+      "grad_norm": 1.0005592107772827,
+      "learning_rate": 0.00028593553724972146,
+      "loss": 3.2603923797607424,
+      "step": 1895
+    },
+    {
+      "epoch": 0.36503362151777136,
+      "grad_norm": 1.2387863397598267,
+      "learning_rate": 0.0002858346348034408,
+      "loss": 3.2571121215820313,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3659942363112392,
+      "grad_norm": 1.0775574445724487,
+      "learning_rate": 0.0002857333896324417,
+      "loss": 3.2519359588623047,
+      "step": 1905
+    },
+    {
+      "epoch": 0.366954851104707,
+      "grad_norm": 1.1767851114273071,
+      "learning_rate": 0.00028563180199217596,
+      "loss": 3.256037139892578,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3679154658981748,
+      "grad_norm": 1.2488412857055664,
+      "learning_rate": 0.00028552987213895965,
+      "loss": 3.2543380737304686,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3688760806916426,
+      "grad_norm": 1.2731399536132812,
+      "learning_rate": 0.0002854276003299723,
+      "loss": 3.251921844482422,
+      "step": 1920
+    },
+    {
+      "epoch": 0.36983669548511044,
+      "grad_norm": 1.2652751207351685,
+      "learning_rate": 0.00028532498682325613,
+      "loss": 3.254206085205078,
+      "step": 1925
+    },
+    {
+      "epoch": 0.37079731027857826,
+      "grad_norm": 1.3380461931228638,
+      "learning_rate": 0.0002852220318777156,
+      "loss": 3.2542865753173826,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37175792507204614,
+      "grad_norm": 1.0234968662261963,
+      "learning_rate": 0.0002851187357531165,
+      "loss": 3.2551605224609377,
+      "step": 1935
+    },
+    {
+      "epoch": 0.37271853986551395,
+      "grad_norm": 0.7750507593154907,
+      "learning_rate": 0.00028501509871008563,
+      "loss": 3.258992004394531,
+      "step": 1940
+    },
+    {
+      "epoch": 0.37367915465898177,
+      "grad_norm": 0.9654103517532349,
+      "learning_rate": 0.00028491112101010986,
+      "loss": 3.2557037353515623,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3746397694524496,
+      "grad_norm": 1.1453462839126587,
+      "learning_rate": 0.0002848068029155356,
+      "loss": 3.249644470214844,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3756003842459174,
+      "grad_norm": 1.0247918367385864,
+      "learning_rate": 0.0002847021446895681,
+      "loss": 3.253766632080078,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3765609990393852,
+      "grad_norm": 1.4280420541763306,
+      "learning_rate": 0.0002845971465962708,
+      "loss": 3.2554515838623046,
+      "step": 1960
+    },
+    {
+      "epoch": 0.37752161383285304,
+      "grad_norm": 1.534204125404358,
+      "learning_rate": 0.00028449180890056464,
+      "loss": 3.2532962799072265,
+      "step": 1965
+    },
+    {
+      "epoch": 0.37848222862632086,
+      "grad_norm": 1.2856396436691284,
+      "learning_rate": 0.00028438613186822746,
+      "loss": 3.251531219482422,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3794428434197887,
+      "grad_norm": 1.0911295413970947,
+      "learning_rate": 0.00028428011576589327,
+      "loss": 3.250580978393555,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3804034582132565,
+      "grad_norm": 1.142147421836853,
+      "learning_rate": 0.00028417376086105155,
+      "loss": 3.254068374633789,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3813640730067243,
+      "grad_norm": 1.335033655166626,
+      "learning_rate": 0.0002840670674220466,
+      "loss": 3.2494667053222654,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38232468780019213,
+      "grad_norm": 1.410131812095642,
+      "learning_rate": 0.00028396003571807697,
+      "loss": 3.2515232086181642,
+      "step": 1990
+    },
+    {
+      "epoch": 0.38328530259365995,
+      "grad_norm": 1.3424606323242188,
+      "learning_rate": 0.0002838526660191946,
+      "loss": 3.250862884521484,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38424591738712777,
+      "grad_norm": 0.8202113509178162,
+      "learning_rate": 0.0002837449585963043,
+      "loss": 3.249342346191406,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3852065321805956,
+      "grad_norm": 0.9713074564933777,
+      "learning_rate": 0.00028363691372116284,
+      "loss": 3.2535247802734375,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3861671469740634,
+      "grad_norm": 0.7178729176521301,
+      "learning_rate": 0.0002835285316663786,
+      "loss": 3.2526882171630858,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3871277617675312,
+      "grad_norm": 0.9730674624443054,
+      "learning_rate": 0.0002834198127054106,
+      "loss": 3.2526809692382814,
+      "step": 2015
+    },
+    {
+      "epoch": 0.38808837656099904,
+      "grad_norm": 1.3725463151931763,
+      "learning_rate": 0.000283310757112568,
+      "loss": 3.2512813568115235,
+      "step": 2020
+    },
+    {
+      "epoch": 0.38904899135446686,
+      "grad_norm": 1.5489643812179565,
+      "learning_rate": 0.0002832013651630091,
+      "loss": 3.2502918243408203,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3900096061479347,
+      "grad_norm": 0.8844899535179138,
+      "learning_rate": 0.0002830916371327412,
+      "loss": 3.250738525390625,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3909702209414025,
+      "grad_norm": 1.0500115156173706,
+      "learning_rate": 0.0002829815732986192,
+      "loss": 3.2540382385253905,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3919308357348703,
+      "grad_norm": 1.0166023969650269,
+      "learning_rate": 0.00028287117393834563,
+      "loss": 3.2548370361328125,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3928914505283381,
+      "grad_norm": 1.216377854347229,
+      "learning_rate": 0.00028276043933046926,
+      "loss": 3.248933792114258,
+      "step": 2045
+    },
+    {
+      "epoch": 0.39385206532180594,
+      "grad_norm": 0.9477090239524841,
+      "learning_rate": 0.0002826493697543849,
+      "loss": 3.2496387481689455,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39481268011527376,
+      "grad_norm": 1.4546550512313843,
+      "learning_rate": 0.00028253796549033245,
+      "loss": 3.257366943359375,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3957732949087416,
+      "grad_norm": 1.3477628231048584,
+      "learning_rate": 0.0002824262268193964,
+      "loss": 3.2546245574951174,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3967339097022094,
+      "grad_norm": 1.1829005479812622,
+      "learning_rate": 0.00028231415402350476,
+      "loss": 3.2527755737304687,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3976945244956772,
+      "grad_norm": 1.00674569606781,
+      "learning_rate": 0.00028220174738542876,
+      "loss": 3.249166488647461,
+      "step": 2070
+    },
+    {
+      "epoch": 0.39865513928914503,
+      "grad_norm": 1.143075942993164,
+      "learning_rate": 0.00028208900718878187,
+      "loss": 3.2508075714111326,
+      "step": 2075
+    },
+    {
+      "epoch": 0.39961575408261285,
+      "grad_norm": 0.9893442392349243,
+      "learning_rate": 0.00028197593371801915,
+      "loss": 3.256524658203125,
+      "step": 2080
+    },
+    {
+      "epoch": 0.40057636887608067,
+      "grad_norm": 0.9596649408340454,
+      "learning_rate": 0.00028186252725843664,
+      "loss": 3.2511486053466796,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4015369836695485,
+      "grad_norm": 1.4209188222885132,
+      "learning_rate": 0.00028174878809617037,
+      "loss": 3.2548614501953126,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4024975984630163,
+      "grad_norm": 1.0415889024734497,
+      "learning_rate": 0.000281634716518196,
+      "loss": 3.2491207122802734,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4034582132564842,
+      "grad_norm": 1.0843944549560547,
+      "learning_rate": 0.00028152031281232775,
+      "loss": 3.2516613006591797,
+      "step": 2100
+    },
+    {
+      "epoch": 0.404418828049952,
+      "grad_norm": 1.5569794178009033,
+      "learning_rate": 0.00028140557726721795,
+      "loss": 3.2495628356933595,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4053794428434198,
+      "grad_norm": 0.8837640881538391,
+      "learning_rate": 0.00028129051017235614,
+      "loss": 3.249878692626953,
+      "step": 2110
+    },
+    {
+      "epoch": 0.40634005763688763,
+      "grad_norm": 1.1749989986419678,
+      "learning_rate": 0.0002811751118180684,
+      "loss": 3.2475624084472656,
+      "step": 2115
+    },
+    {
+      "epoch": 0.40730067243035545,
+      "grad_norm": 1.2945916652679443,
+      "learning_rate": 0.00028105938249551666,
+      "loss": 3.2487873077392577,
+      "step": 2120
+    },
+    {
+      "epoch": 0.40826128722382327,
+      "grad_norm": 0.9443672299385071,
+      "learning_rate": 0.0002809433224966978,
+      "loss": 3.247914123535156,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4092219020172911,
+      "grad_norm": 1.1064391136169434,
+      "learning_rate": 0.0002808269321144431,
+      "loss": 3.250864028930664,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4101825168107589,
+      "grad_norm": 1.2150980234146118,
+      "learning_rate": 0.00028071021164241755,
+      "loss": 3.248881530761719,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4111431316042267,
+      "grad_norm": 0.8171827793121338,
+      "learning_rate": 0.00028059316137511877,
+      "loss": 3.246689224243164,
+      "step": 2140
+    },
+    {
+      "epoch": 0.41210374639769454,
+      "grad_norm": 1.093009114265442,
+      "learning_rate": 0.0002804757816078766,
+      "loss": 3.2493213653564452,
+      "step": 2145
+    },
+    {
+      "epoch": 0.41306436119116235,
+      "grad_norm": 1.021937370300293,
+      "learning_rate": 0.00028035807263685224,
+      "loss": 3.2454933166503905,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4140249759846302,
+      "grad_norm": 1.1384049654006958,
+      "learning_rate": 0.0002802400347590376,
+      "loss": 3.2516769409179687,
+      "step": 2155
+    },
+    {
+      "epoch": 0.414985590778098,
+      "grad_norm": 1.0551131963729858,
+      "learning_rate": 0.00028012166827225425,
+      "loss": 3.2476879119873048,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4159462055715658,
+      "grad_norm": 0.9579499363899231,
+      "learning_rate": 0.0002800029734751529,
+      "loss": 3.247665023803711,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4169068203650336,
+      "grad_norm": 1.290895938873291,
+      "learning_rate": 0.0002798839506672129,
+      "loss": 3.248125457763672,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41786743515850144,
+      "grad_norm": 1.1702419519424438,
+      "learning_rate": 0.00027976460014874087,
+      "loss": 3.2500679016113283,
+      "step": 2175
+    },
+    {
+      "epoch": 0.41882804995196926,
+      "grad_norm": 1.2696362733840942,
+      "learning_rate": 0.0002796449222208704,
+      "loss": 3.2472507476806642,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4197886647454371,
+      "grad_norm": 1.1748182773590088,
+      "learning_rate": 0.0002795249171855613,
+      "loss": 3.245960998535156,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4207492795389049,
+      "grad_norm": 1.2495172023773193,
+      "learning_rate": 0.00027940458534559846,
+      "loss": 3.2419815063476562,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4217098943323727,
+      "grad_norm": 1.2086557149887085,
+      "learning_rate": 0.0002792839270045916,
+      "loss": 3.2496551513671874,
+      "step": 2195
+    },
+    {
+      "epoch": 0.42267050912584053,
+      "grad_norm": 0.9928983449935913,
+      "learning_rate": 0.0002791629424669739,
+      "loss": 3.2451755523681642,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42363112391930835,
+      "grad_norm": 1.264974594116211,
+      "learning_rate": 0.00027904163203800185,
+      "loss": 3.2482452392578125,
+      "step": 2205
+    },
+    {
+      "epoch": 0.42459173871277617,
+      "grad_norm": 1.2751965522766113,
+      "learning_rate": 0.0002789199960237542,
+      "loss": 3.251111602783203,
+      "step": 2210
+    },
+    {
+      "epoch": 0.425552353506244,
+      "grad_norm": 1.115212082862854,
+      "learning_rate": 0.0002787980347311309,
+      "loss": 3.2442550659179688,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4265129682997118,
+      "grad_norm": 1.3036905527114868,
+      "learning_rate": 0.00027867574846785295,
+      "loss": 3.250843048095703,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4274735830931796,
+      "grad_norm": 1.3169866800308228,
+      "learning_rate": 0.0002785531375424611,
+      "loss": 3.2423397064208985,
+      "step": 2225
+    },
+    {
+      "epoch": 0.42843419788664744,
+      "grad_norm": 1.1663439273834229,
+      "learning_rate": 0.00027843020226431535,
+      "loss": 3.2481651306152344,
+      "step": 2230
+    },
+    {
+      "epoch": 0.42939481268011526,
+      "grad_norm": 1.014664649963379,
+      "learning_rate": 0.0002783069429435939,
+      "loss": 3.245722198486328,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4303554274735831,
+      "grad_norm": 1.1078412532806396,
+      "learning_rate": 0.00027818335989129275,
+      "loss": 3.2450366973876954,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4313160422670509,
+      "grad_norm": 1.0631850957870483,
+      "learning_rate": 0.00027805945341922457,
+      "loss": 3.2482345581054686,
+      "step": 2245
+    },
+    {
+      "epoch": 0.4322766570605187,
+      "grad_norm": 1.182057499885559,
+      "learning_rate": 0.0002779352238400181,
+      "loss": 3.2490921020507812,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4332372718539865,
+      "grad_norm": 1.021649718284607,
+      "learning_rate": 0.0002778106714671174,
+      "loss": 3.2460681915283205,
+      "step": 2255
+    },
+    {
+      "epoch": 0.43419788664745435,
+      "grad_norm": 1.0771024227142334,
+      "learning_rate": 0.00027768579661478077,
+      "loss": 3.2438682556152343,
+      "step": 2260
+    },
+    {
+      "epoch": 0.43515850144092216,
+      "grad_norm": 1.1866998672485352,
+      "learning_rate": 0.0002775605995980803,
+      "loss": 3.2406051635742186,
+      "step": 2265
+    },
+    {
+      "epoch": 0.43611911623439004,
+      "grad_norm": 1.2904285192489624,
+      "learning_rate": 0.00027743508073290097,
+      "loss": 3.247574234008789,
+      "step": 2270
+    },
+    {
+      "epoch": 0.43707973102785785,
+      "grad_norm": 0.990243136882782,
+      "learning_rate": 0.0002773092403359397,
+      "loss": 3.2439697265625,
+      "step": 2275
+    },
+    {
+      "epoch": 0.43804034582132567,
+      "grad_norm": 0.976154625415802,
+      "learning_rate": 0.0002771830787247047,
+      "loss": 3.251276397705078,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4390009606147935,
+      "grad_norm": 1.2034424543380737,
+      "learning_rate": 0.0002770565962175148,
+      "loss": 3.2444232940673827,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4399615754082613,
+      "grad_norm": 1.4228894710540771,
+      "learning_rate": 0.00027692979313349815,
+      "loss": 3.2472274780273436,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4409221902017291,
+      "grad_norm": 1.3287744522094727,
+      "learning_rate": 0.000276802669792592,
+      "loss": 3.247750091552734,
+      "step": 2295
+    },
+    {
+      "epoch": 0.44188280499519694,
+      "grad_norm": 1.1008442640304565,
+      "learning_rate": 0.0002766752265155416,
+      "loss": 3.2413604736328123,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44284341978866476,
+      "grad_norm": 1.3397494554519653,
+      "learning_rate": 0.0002765474636238994,
+      "loss": 3.243372344970703,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4438040345821326,
+      "grad_norm": 1.2448689937591553,
+      "learning_rate": 0.00027641938144002434,
+      "loss": 3.24469108581543,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4447646493756004,
+      "grad_norm": 1.2328872680664062,
+      "learning_rate": 0.0002762909802870809,
+      "loss": 3.2434837341308596,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4457252641690682,
+      "grad_norm": 1.3019057512283325,
+      "learning_rate": 0.0002761622604890382,
+      "loss": 3.2452354431152344,
+      "step": 2320
+    },
+    {
+      "epoch": 0.44668587896253603,
+      "grad_norm": 1.0617128610610962,
+      "learning_rate": 0.0002760332223706697,
+      "loss": 3.243347930908203,
+      "step": 2325
+    },
+    {
+      "epoch": 0.44764649375600385,
+      "grad_norm": 1.2202798128128052,
+      "learning_rate": 0.0002759038662575518,
+      "loss": 3.2428775787353517,
+      "step": 2330
+    },
+    {
+      "epoch": 0.44860710854947167,
+      "grad_norm": 0.8098715543746948,
+      "learning_rate": 0.0002757741924760631,
+      "loss": 3.2433307647705076,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4495677233429395,
+      "grad_norm": 1.2258954048156738,
+      "learning_rate": 0.0002756442013533839,
+      "loss": 3.2418006896972655,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4505283381364073,
+      "grad_norm": 1.155310869216919,
+      "learning_rate": 0.0002755138932174952,
+      "loss": 3.2446517944335938,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4514889529298751,
+      "grad_norm": 0.9408679604530334,
+      "learning_rate": 0.0002753832683971778,
+      "loss": 3.2486572265625,
+      "step": 2350
+    },
+    {
+      "epoch": 0.45244956772334294,
+      "grad_norm": 1.1427699327468872,
+      "learning_rate": 0.0002752523272220114,
+      "loss": 3.2479103088378904,
+      "step": 2355
+    },
+    {
+      "epoch": 0.45341018251681076,
+      "grad_norm": 0.9570598602294922,
+      "learning_rate": 0.00027512107002237415,
+      "loss": 3.247629165649414,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4543707973102786,
+      "grad_norm": 0.8978891968727112,
+      "learning_rate": 0.0002749894971294414,
+      "loss": 3.242329406738281,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4553314121037464,
+      "grad_norm": 1.0361249446868896,
+      "learning_rate": 0.000274857608875185,
+      "loss": 3.243325042724609,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4562920268972142,
+      "grad_norm": 0.9221097230911255,
+      "learning_rate": 0.0002747254055923726,
+      "loss": 3.2444053649902345,
+      "step": 2375
+    },
+    {
+      "epoch": 0.457252641690682,
+      "grad_norm": 1.2986423969268799,
+      "learning_rate": 0.0002745928876145667,
+      "loss": 3.2442516326904296,
+      "step": 2380
+    },
+    {
+      "epoch": 0.45821325648414984,
+      "grad_norm": 0.9883876442909241,
+      "learning_rate": 0.0002744600552761236,
+      "loss": 3.2429885864257812,
+      "step": 2385
+    },
+    {
+      "epoch": 0.45917387127761766,
+      "grad_norm": 1.4296432733535767,
+      "learning_rate": 0.000274326908912193,
+      "loss": 3.2417251586914064,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4601344860710855,
+      "grad_norm": 1.216537594795227,
+      "learning_rate": 0.00027419344885871685,
+      "loss": 3.2435916900634765,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4610951008645533,
+      "grad_norm": 1.0442991256713867,
+      "learning_rate": 0.0002740596754524285,
+      "loss": 3.2400821685791015,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4620557156580211,
+      "grad_norm": 1.4031963348388672,
+      "learning_rate": 0.0002739255890308521,
+      "loss": 3.2466384887695314,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46301633045148893,
+      "grad_norm": 1.301132082939148,
+      "learning_rate": 0.0002737911899323013,
+      "loss": 3.241625213623047,
+      "step": 2410
+    },
+    {
+      "epoch": 0.46397694524495675,
+      "grad_norm": 1.3836039304733276,
+      "learning_rate": 0.00027365647849587893,
+      "loss": 3.2370471954345703,
+      "step": 2415
+    },
+    {
+      "epoch": 0.46493756003842457,
+      "grad_norm": 0.9351392984390259,
+      "learning_rate": 0.0002735214550614757,
+      "loss": 3.2448040008544923,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4658981748318924,
+      "grad_norm": 0.9278304576873779,
+      "learning_rate": 0.00027338611996976955,
+      "loss": 3.239888000488281,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4668587896253602,
+      "grad_norm": 0.8525119423866272,
+      "learning_rate": 0.0002732504735622248,
+      "loss": 3.2448410034179687,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4678194044188281,
+      "grad_norm": 1.2899324893951416,
+      "learning_rate": 0.00027311451618109144,
+      "loss": 3.2423675537109373,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4687800192122959,
+      "grad_norm": 0.8825019598007202,
+      "learning_rate": 0.00027297824816940365,
+      "loss": 3.240159606933594,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4697406340057637,
+      "grad_norm": 1.301020860671997,
+      "learning_rate": 0.00027284166987097977,
+      "loss": 3.24395751953125,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47070124879923153,
+      "grad_norm": 0.9350459575653076,
+      "learning_rate": 0.0002727047816304208,
+      "loss": 3.2402999877929686,
+      "step": 2450
+    },
+    {
+      "epoch": 0.47166186359269935,
+      "grad_norm": 1.1744569540023804,
+      "learning_rate": 0.00027256758379310975,
+      "loss": 3.239767837524414,
+      "step": 2455
+    },
+    {
+      "epoch": 0.47262247838616717,
+      "grad_norm": 0.9879583716392517,
+      "learning_rate": 0.00027243007670521086,
+      "loss": 3.242339324951172,
+      "step": 2460
+    },
+    {
+      "epoch": 0.473583093179635,
+      "grad_norm": 0.8559085726737976,
+      "learning_rate": 0.00027229226071366874,
+      "loss": 3.241751861572266,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4745437079731028,
+      "grad_norm": 1.3251603841781616,
+      "learning_rate": 0.00027215413616620714,
+      "loss": 3.2432788848876952,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4755043227665706,
+      "grad_norm": 1.1395015716552734,
+      "learning_rate": 0.00027201570341132855,
+      "loss": 3.239895248413086,
+      "step": 2475
+    },
+    {
+      "epoch": 0.47646493756003844,
+      "grad_norm": 1.1372860670089722,
+      "learning_rate": 0.0002718769627983129,
+      "loss": 3.2404502868652343,
+      "step": 2480
+    },
+    {
+      "epoch": 0.47742555235350626,
+      "grad_norm": 1.014133334159851,
+      "learning_rate": 0.00027173791467721714,
+      "loss": 3.238873291015625,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4783861671469741,
+      "grad_norm": 1.2768558263778687,
+      "learning_rate": 0.00027159855939887386,
+      "loss": 3.2407760620117188,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4793467819404419,
+      "grad_norm": 1.0559134483337402,
+      "learning_rate": 0.0002714588973148907,
+      "loss": 3.2391632080078123,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4803073967339097,
+      "grad_norm": 1.212760090827942,
+      "learning_rate": 0.00027131892877764945,
+      "loss": 3.242381286621094,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4812680115273775,
+      "grad_norm": 1.3705852031707764,
+      "learning_rate": 0.0002711786541403051,
+      "loss": 3.2453689575195312,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48222862632084534,
+      "grad_norm": 1.1016364097595215,
+      "learning_rate": 0.00027103807375678494,
+      "loss": 3.240717315673828,
+      "step": 2510
+    },
+    {
+      "epoch": 0.48318924111431316,
+      "grad_norm": 1.0440564155578613,
+      "learning_rate": 0.0002708971879817877,
+      "loss": 3.2409019470214844,
+      "step": 2515
+    },
+    {
+      "epoch": 0.484149855907781,
+      "grad_norm": 1.175846815109253,
+      "learning_rate": 0.0002707559971707827,
+      "loss": 3.2366703033447264,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4851104707012488,
+      "grad_norm": 1.2363582849502563,
+      "learning_rate": 0.00027061450168000875,
+      "loss": 3.235219192504883,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4860710854947166,
+      "grad_norm": 1.2251267433166504,
+      "learning_rate": 0.00027047270186647353,
+      "loss": 3.2362335205078123,
+      "step": 2530
+    },
+    {
+      "epoch": 0.48703170028818443,
+      "grad_norm": 0.8210157752037048,
+      "learning_rate": 0.00027033059808795253,
+      "loss": 3.2397132873535157,
+      "step": 2535
+    },
+    {
+      "epoch": 0.48799231508165225,
+      "grad_norm": 1.229904055595398,
+      "learning_rate": 0.0002701881907029881,
+      "loss": 3.2385082244873047,
+      "step": 2540
+    },
+    {
+      "epoch": 0.48895292987512007,
+      "grad_norm": 1.1676335334777832,
+      "learning_rate": 0.00027004548007088876,
+      "loss": 3.2385066986083983,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4899135446685879,
+      "grad_norm": 1.271478533744812,
+      "learning_rate": 0.000269902466551728,
+      "loss": 3.2415950775146483,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4908741594620557,
+      "grad_norm": 1.395169973373413,
+      "learning_rate": 0.00026975915050634367,
+      "loss": 3.2399360656738283,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4918347742555235,
+      "grad_norm": 1.3647792339324951,
+      "learning_rate": 0.00026961553229633685,
+      "loss": 3.243121337890625,
+      "step": 2560
+    },
+    {
+      "epoch": 0.49279538904899134,
+      "grad_norm": 1.0769857168197632,
+      "learning_rate": 0.000269471612284071,
+      "loss": 3.2393962860107424,
+      "step": 2565
+    },
+    {
+      "epoch": 0.49375600384245916,
+      "grad_norm": 0.9694387316703796,
+      "learning_rate": 0.00026932739083267105,
+      "loss": 3.2386848449707033,
+      "step": 2570
+    },
+    {
+      "epoch": 0.494716618635927,
+      "grad_norm": 0.9222965240478516,
+      "learning_rate": 0.00026918286830602254,
+      "loss": 3.238786315917969,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4956772334293948,
+      "grad_norm": 1.1549302339553833,
+      "learning_rate": 0.00026903804506877064,
+      "loss": 3.239440155029297,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4966378482228626,
+      "grad_norm": 1.0019875764846802,
+      "learning_rate": 0.0002688929214863192,
+      "loss": 3.237133026123047,
+      "step": 2585
+    },
+    {
+      "epoch": 0.49759846301633043,
+      "grad_norm": 1.1600396633148193,
+      "learning_rate": 0.0002687474979248299,
+      "loss": 3.2444324493408203,
+      "step": 2590
+    },
+    {
+      "epoch": 0.49855907780979825,
+      "grad_norm": 1.2407069206237793,
+      "learning_rate": 0.00026860177475122133,
+      "loss": 3.2405441284179686,
+      "step": 2595
+    },
+    {
+      "epoch": 0.49951969260326606,
+      "grad_norm": 1.313706874847412,
+      "learning_rate": 0.00026845575233316806,
+      "loss": 3.2400081634521483,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5004803073967339,
+      "grad_norm": 1.102402925491333,
+      "learning_rate": 0.00026830943103909954,
+      "loss": 3.236441421508789,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5014409221902018,
+      "grad_norm": 1.4062278270721436,
+      "learning_rate": 0.00026816281123819946,
+      "loss": 3.24122314453125,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5024015369836695,
+      "grad_norm": 1.1859440803527832,
+      "learning_rate": 0.00026801589330040465,
+      "loss": 3.238700103759766,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5033621517771374,
+      "grad_norm": 0.8828426599502563,
+      "learning_rate": 0.000267868677596404,
+      "loss": 3.2366287231445314,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5043227665706052,
+      "grad_norm": 1.1908822059631348,
+      "learning_rate": 0.000267721164497638,
+      "loss": 3.2335018157958983,
+      "step": 2625
+    },
+    {
+      "epoch": 0.505283381364073,
+      "grad_norm": 1.3185487985610962,
+      "learning_rate": 0.00026757335437629725,
+      "loss": 3.2354408264160157,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5062439961575408,
+      "grad_norm": 1.1835713386535645,
+      "learning_rate": 0.00026742524760532183,
+      "loss": 3.235929489135742,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5072046109510087,
+      "grad_norm": 1.6619583368301392,
+      "learning_rate": 0.00026727684455840037,
+      "loss": 3.236317443847656,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5081652257444764,
+      "grad_norm": 1.0703696012496948,
+      "learning_rate": 0.00026712814560996896,
+      "loss": 3.2386764526367187,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5091258405379443,
+      "grad_norm": 1.149133324623108,
+      "learning_rate": 0.0002669791511352103,
+      "loss": 3.23681526184082,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5100864553314121,
+      "grad_norm": 1.1290984153747559,
+      "learning_rate": 0.0002668298615100527,
+      "loss": 3.2398101806640627,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5110470701248799,
+      "grad_norm": 1.0490314960479736,
+      "learning_rate": 0.0002666802771111693,
+      "loss": 3.238353729248047,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5120076849183477,
+      "grad_norm": 0.849686861038208,
+      "learning_rate": 0.0002665303983159767,
+      "loss": 3.2377159118652346,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5129682997118156,
+      "grad_norm": 1.13809072971344,
+      "learning_rate": 0.00026638022550263467,
+      "loss": 3.2373073577880858,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5139289145052833,
+      "grad_norm": 1.1344239711761475,
+      "learning_rate": 0.0002662297590500445,
+      "loss": 3.233965301513672,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5148895292987512,
+      "grad_norm": 1.3232028484344482,
+      "learning_rate": 0.0002660789993378486,
+      "loss": 3.236408233642578,
+      "step": 2680
+    },
+    {
+      "epoch": 0.515850144092219,
+      "grad_norm": 1.2224704027175903,
+      "learning_rate": 0.0002659279467464291,
+      "loss": 3.238209915161133,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5168107588856868,
+      "grad_norm": 1.106103777885437,
+      "learning_rate": 0.0002657766016569072,
+      "loss": 3.2342296600341798,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5177713736791547,
+      "grad_norm": 0.8532865643501282,
+      "learning_rate": 0.00026562496445114205,
+      "loss": 3.238364410400391,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5187319884726225,
+      "grad_norm": 0.8444566130638123,
+      "learning_rate": 0.00026547303551172996,
+      "loss": 3.236472320556641,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5196926032660903,
+      "grad_norm": 0.8221146464347839,
+      "learning_rate": 0.0002653208152220032,
+      "loss": 3.2367286682128906,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5206532180595581,
+      "grad_norm": 1.1430490016937256,
+      "learning_rate": 0.00026516830396602914,
+      "loss": 3.2309627532958984,
+      "step": 2710
+    },
+    {
+      "epoch": 0.521613832853026,
+      "grad_norm": 1.3973969221115112,
+      "learning_rate": 0.00026501550212860937,
+      "loss": 3.2348331451416015,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5225744476464937,
+      "grad_norm": 1.2749481201171875,
+      "learning_rate": 0.0002648624100952786,
+      "loss": 3.2363750457763674,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5235350624399616,
+      "grad_norm": 1.2370373010635376,
+      "learning_rate": 0.00026470902825230367,
+      "loss": 3.2376075744628907,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5244956772334294,
+      "grad_norm": 1.0846977233886719,
+      "learning_rate": 0.0002645553569866828,
+      "loss": 3.2337451934814454,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5254562920268973,
+      "grad_norm": 1.026764154434204,
+      "learning_rate": 0.00026440139668614427,
+      "loss": 3.237920379638672,
+      "step": 2735
+    },
+    {
+      "epoch": 0.526416906820365,
+      "grad_norm": 1.045480489730835,
+      "learning_rate": 0.00026424714773914574,
+      "loss": 3.231882858276367,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5273775216138329,
+      "grad_norm": 0.9559823274612427,
+      "learning_rate": 0.0002640926105348732,
+      "loss": 3.23189697265625,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5283381364073007,
+      "grad_norm": 1.3470317125320435,
+      "learning_rate": 0.00026393778546323976,
+      "loss": 3.2387470245361327,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5292987512007685,
+      "grad_norm": 1.1389282941818237,
+      "learning_rate": 0.00026378267291488506,
+      "loss": 3.2349002838134764,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5302593659942363,
+      "grad_norm": 1.0445148944854736,
+      "learning_rate": 0.00026362727328117384,
+      "loss": 3.2324363708496096,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5312199807877042,
+      "grad_norm": 1.0291074514389038,
+      "learning_rate": 0.00026347158695419546,
+      "loss": 3.2345924377441406,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5321805955811719,
+      "grad_norm": 1.23056161403656,
+      "learning_rate": 0.00026331561432676244,
+      "loss": 3.2309837341308594,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5331412103746398,
+      "grad_norm": 0.658456563949585,
+      "learning_rate": 0.0002631593557924097,
+      "loss": 3.232289123535156,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5341018251681076,
+      "grad_norm": 1.1292603015899658,
+      "learning_rate": 0.0002630028117453936,
+      "loss": 3.2309852600097657,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5350624399615754,
+      "grad_norm": 0.8553731441497803,
+      "learning_rate": 0.00026284598258069074,
+      "loss": 3.235504150390625,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5360230547550432,
+      "grad_norm": 0.8991851210594177,
+      "learning_rate": 0.00026268886869399723,
+      "loss": 3.233042907714844,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5369836695485111,
+      "grad_norm": 0.916896402835846,
+      "learning_rate": 0.0002625314704817276,
+      "loss": 3.2310935974121096,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5379442843419788,
+      "grad_norm": 1.1668161153793335,
+      "learning_rate": 0.0002623737883410136,
+      "loss": 3.232282257080078,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5389048991354467,
+      "grad_norm": 1.223141074180603,
+      "learning_rate": 0.00026221582266970346,
+      "loss": 3.2328109741210938,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5398655139289145,
+      "grad_norm": 1.0517135858535767,
+      "learning_rate": 0.00026205757386636085,
+      "loss": 3.2323383331298827,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5408261287223823,
+      "grad_norm": 0.7666857242584229,
+      "learning_rate": 0.0002618990423302636,
+      "loss": 3.234081268310547,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5417867435158501,
+      "grad_norm": 1.253327488899231,
+      "learning_rate": 0.0002617402284614031,
+      "loss": 3.233143997192383,
+      "step": 2820
+    },
+    {
+      "epoch": 0.542747358309318,
+      "grad_norm": 0.7495427131652832,
+      "learning_rate": 0.00026158113266048317,
+      "loss": 3.236143112182617,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5437079731027857,
+      "grad_norm": 1.1047379970550537,
+      "learning_rate": 0.0002614217553289186,
+      "loss": 3.234874725341797,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5446685878962536,
+      "grad_norm": 1.2154449224472046,
+      "learning_rate": 0.0002612620968688349,
+      "loss": 3.2348342895507813,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5456292026897214,
+      "grad_norm": 1.1259676218032837,
+      "learning_rate": 0.0002611021576830667,
+      "loss": 3.232221221923828,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5465898174831892,
+      "grad_norm": 1.2192593812942505,
+      "learning_rate": 0.00026094193817515697,
+      "loss": 3.2312171936035154,
+      "step": 2845
+    },
+    {
+      "epoch": 0.547550432276657,
+      "grad_norm": 1.3673197031021118,
+      "learning_rate": 0.00026078143874935604,
+      "loss": 3.229291534423828,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5485110470701249,
+      "grad_norm": 1.3008506298065186,
+      "learning_rate": 0.00026062065981062024,
+      "loss": 3.2330032348632813,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5494716618635928,
+      "grad_norm": 1.1764352321624756,
+      "learning_rate": 0.00026045960176461155,
+      "loss": 3.232608413696289,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5504322766570605,
+      "grad_norm": 1.0877325534820557,
+      "learning_rate": 0.0002602982650176958,
+      "loss": 3.232305908203125,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5513928914505284,
+      "grad_norm": 1.1681047677993774,
+      "learning_rate": 0.0002601366499769422,
+      "loss": 3.233287811279297,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5523535062439962,
+      "grad_norm": 1.0753793716430664,
+      "learning_rate": 0.00025997475705012203,
+      "loss": 3.2348102569580077,
+      "step": 2875
+    },
+    {
+      "epoch": 0.553314121037464,
+      "grad_norm": 0.9505957365036011,
+      "learning_rate": 0.0002598125866457078,
+      "loss": 3.2324668884277346,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5542747358309318,
+      "grad_norm": 1.2333307266235352,
+      "learning_rate": 0.0002596501391728721,
+      "loss": 3.235234832763672,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5552353506243997,
+      "grad_norm": 1.0360201597213745,
+      "learning_rate": 0.00025948741504148656,
+      "loss": 3.234451675415039,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5561959654178674,
+      "grad_norm": 0.9275029301643372,
+      "learning_rate": 0.0002593244146621208,
+      "loss": 3.2318180084228514,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5571565802113353,
+      "grad_norm": 0.8183401823043823,
+      "learning_rate": 0.00025916113844604157,
+      "loss": 3.2336944580078124,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5581171950048031,
+      "grad_norm": 0.7255450487136841,
+      "learning_rate": 0.0002589975868052114,
+      "loss": 3.2311965942382814,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5590778097982709,
+      "grad_norm": 1.1232422590255737,
+      "learning_rate": 0.000258833760152288,
+      "loss": 3.2349414825439453,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5600384245917387,
+      "grad_norm": 1.311683177947998,
+      "learning_rate": 0.00025866965890062276,
+      "loss": 3.233409881591797,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5609990393852066,
+      "grad_norm": 1.104836106300354,
+      "learning_rate": 0.0002585052834642599,
+      "loss": 3.228401947021484,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5619596541786743,
+      "grad_norm": 1.170188069343567,
+      "learning_rate": 0.0002583406342579357,
+      "loss": 3.2292003631591797,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5629202689721422,
+      "grad_norm": 1.1829124689102173,
+      "learning_rate": 0.0002581757116970768,
+      "loss": 3.2359298706054687,
+      "step": 2930
+    },
+    {
+      "epoch": 0.56388088376561,
+      "grad_norm": 0.885960042476654,
+      "learning_rate": 0.00025801051619779987,
+      "loss": 3.2320457458496095,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5648414985590778,
+      "grad_norm": 1.0792263746261597,
+      "learning_rate": 0.0002578450481769099,
+      "loss": 3.2303844451904298,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5658021133525456,
+      "grad_norm": 1.2442518472671509,
+      "learning_rate": 0.0002576793080519,
+      "loss": 3.2286468505859376,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5667627281460135,
+      "grad_norm": 0.976241409778595,
+      "learning_rate": 0.00025751329624094926,
+      "loss": 3.226253128051758,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5677233429394812,
+      "grad_norm": 1.2932735681533813,
+      "learning_rate": 0.00025734701316292263,
+      "loss": 3.2291339874267577,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5686839577329491,
+      "grad_norm": 1.17381751537323,
+      "learning_rate": 0.0002571804592373693,
+      "loss": 3.2318046569824217,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5696445725264169,
+      "grad_norm": 1.1502513885498047,
+      "learning_rate": 0.00025701363488452196,
+      "loss": 3.230632781982422,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5706051873198847,
+      "grad_norm": 1.2376641035079956,
+      "learning_rate": 0.00025684654052529556,
+      "loss": 3.2285648345947267,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5715658021133525,
+      "grad_norm": 1.2463771104812622,
+      "learning_rate": 0.0002566791765812862,
+      "loss": 3.232989501953125,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5725264169068204,
+      "grad_norm": 1.148836612701416,
+      "learning_rate": 0.0002565115434747705,
+      "loss": 3.2307807922363283,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5734870317002881,
+      "grad_norm": 1.313881516456604,
+      "learning_rate": 0.00025634364162870386,
+      "loss": 3.2307106018066407,
+      "step": 2985
+    },
+    {
+      "epoch": 0.574447646493756,
+      "grad_norm": 1.0571279525756836,
+      "learning_rate": 0.0002561754714667198,
+      "loss": 3.230280303955078,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5754082612872238,
+      "grad_norm": 0.9680122137069702,
+      "learning_rate": 0.000256007033413129,
+      "loss": 3.2283321380615235,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5763688760806917,
+      "grad_norm": 1.11472749710083,
+      "learning_rate": 0.00025583832789291794,
+      "loss": 3.229805755615234,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5773294908741594,
+      "grad_norm": 1.0987354516983032,
+      "learning_rate": 0.0002556693553317479,
+      "loss": 3.2307418823242187,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5782901056676273,
+      "grad_norm": 0.8845497369766235,
+      "learning_rate": 0.0002555001161559539,
+      "loss": 3.226715850830078,
+      "step": 3010
+    },
+    {
+      "epoch": 0.579250720461095,
+      "grad_norm": 1.37449049949646,
+      "learning_rate": 0.000255330610792544,
+      "loss": 3.230916976928711,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5802113352545629,
+      "grad_norm": 1.2311089038848877,
+      "learning_rate": 0.00025516083966919744,
+      "loss": 3.228615570068359,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5811719500480308,
+      "grad_norm": 1.0169094800949097,
+      "learning_rate": 0.00025499080321426417,
+      "loss": 3.2289390563964844,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5821325648414986,
+      "grad_norm": 1.280392050743103,
+      "learning_rate": 0.00025482050185676367,
+      "loss": 3.227009963989258,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5830931796349664,
+      "grad_norm": 0.9485029578208923,
+      "learning_rate": 0.0002546499360263837,
+      "loss": 3.231037139892578,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5840537944284342,
+      "grad_norm": 1.0025532245635986,
+      "learning_rate": 0.0002544791061534794,
+      "loss": 3.231890869140625,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5850144092219021,
+      "grad_norm": 0.9485952854156494,
+      "learning_rate": 0.000254308012669072,
+      "loss": 3.2329727172851563,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5859750240153698,
+      "grad_norm": 0.9976146817207336,
+      "learning_rate": 0.00025413665600484796,
+      "loss": 3.228594207763672,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5869356388088377,
+      "grad_norm": 0.9106408953666687,
+      "learning_rate": 0.0002539650365931577,
+      "loss": 3.225025177001953,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5878962536023055,
+      "grad_norm": 0.823363721370697,
+      "learning_rate": 0.00025379315486701454,
+      "loss": 3.227878189086914,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5888568683957733,
+      "grad_norm": 0.9073953032493591,
+      "learning_rate": 0.00025362101126009387,
+      "loss": 3.2288444519042967,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5898174831892411,
+      "grad_norm": 1.0130420923233032,
+      "learning_rate": 0.0002534486062067316,
+      "loss": 3.228134536743164,
+      "step": 3070
+    },
+    {
+      "epoch": 0.590778097982709,
+      "grad_norm": 0.8417114615440369,
+      "learning_rate": 0.00025327594014192326,
+      "loss": 3.2314247131347655,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5917387127761767,
+      "grad_norm": 0.8221418261528015,
+      "learning_rate": 0.0002531030135013232,
+      "loss": 3.226838302612305,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5926993275696446,
+      "grad_norm": 1.0958610773086548,
+      "learning_rate": 0.0002529298267212429,
+      "loss": 3.2264358520507814,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5936599423631124,
+      "grad_norm": 1.3347340822219849,
+      "learning_rate": 0.00025275638023865055,
+      "loss": 3.2262367248535155,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5946205571565802,
+      "grad_norm": 0.8695631623268127,
+      "learning_rate": 0.0002525826744911693,
+      "loss": 3.228030776977539,
+      "step": 3095
+    },
+    {
+      "epoch": 0.595581171950048,
+      "grad_norm": 0.8885836601257324,
+      "learning_rate": 0.00025240870991707665,
+      "loss": 3.229400634765625,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5965417867435159,
+      "grad_norm": 0.8682327270507812,
+      "learning_rate": 0.000252234486955303,
+      "loss": 3.2293159484863283,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5975024015369836,
+      "grad_norm": 1.130051851272583,
+      "learning_rate": 0.00025206000604543083,
+      "loss": 3.2304660797119142,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5984630163304515,
+      "grad_norm": 1.0737762451171875,
+      "learning_rate": 0.00025188526762769326,
+      "loss": 3.2278984069824217,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5994236311239193,
+      "grad_norm": 1.096584677696228,
+      "learning_rate": 0.0002517102721429734,
+      "loss": 3.2275615692138673,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6003842459173871,
+      "grad_norm": 1.1471421718597412,
+      "learning_rate": 0.00025153502003280267,
+      "loss": 3.225168991088867,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6013448607108549,
+      "grad_norm": 0.8774202466011047,
+      "learning_rate": 0.0002513595117393602,
+      "loss": 3.225673294067383,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6023054755043228,
+      "grad_norm": 1.1272785663604736,
+      "learning_rate": 0.00025118374770547136,
+      "loss": 3.228855514526367,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6032660902977905,
+      "grad_norm": 1.0108392238616943,
+      "learning_rate": 0.0002510077283746069,
+      "loss": 3.226605224609375,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6042267050912584,
+      "grad_norm": 0.9391249418258667,
+      "learning_rate": 0.00025083145419088165,
+      "loss": 3.2272396087646484,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6051873198847262,
+      "grad_norm": 1.0647284984588623,
+      "learning_rate": 0.0002506549255990534,
+      "loss": 3.2265804290771483,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6061479346781941,
+      "grad_norm": 1.143602967262268,
+      "learning_rate": 0.00025047814304452197,
+      "loss": 3.2242530822753905,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6071085494716618,
+      "grad_norm": 1.098870873451233,
+      "learning_rate": 0.0002503011069733278,
+      "loss": 3.225702667236328,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6080691642651297,
+      "grad_norm": 1.1901029348373413,
+      "learning_rate": 0.00025012381783215117,
+      "loss": 3.2249305725097654,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6090297790585975,
+      "grad_norm": 1.254163384437561,
+      "learning_rate": 0.0002499462760683106,
+      "loss": 3.2239353179931642,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6099903938520653,
+      "grad_norm": 1.3494505882263184,
+      "learning_rate": 0.00024976848212976237,
+      "loss": 3.225982666015625,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6109510086455331,
+      "grad_norm": 0.9600538015365601,
+      "learning_rate": 0.00024959043646509866,
+      "loss": 3.2219696044921875,
+      "step": 3180
+    },
+    {
+      "epoch": 0.611911623439001,
+      "grad_norm": 1.1406611204147339,
+      "learning_rate": 0.00024941213952354703,
+      "loss": 3.225663757324219,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6128722382324687,
+      "grad_norm": 1.276875376701355,
+      "learning_rate": 0.000249233591754969,
+      "loss": 3.226406478881836,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6138328530259366,
+      "grad_norm": 1.0921087265014648,
+      "learning_rate": 0.00024905479360985884,
+      "loss": 3.227063751220703,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6147934678194045,
+      "grad_norm": 1.0305696725845337,
+      "learning_rate": 0.0002488757455393426,
+      "loss": 3.2252223968505858,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6157540826128722,
+      "grad_norm": 1.0612218379974365,
+      "learning_rate": 0.0002486964479951769,
+      "loss": 3.2239383697509765,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6167146974063401,
+      "grad_norm": 1.011340856552124,
+      "learning_rate": 0.00024851690142974796,
+      "loss": 3.2294452667236326,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6176753121998079,
+      "grad_norm": 1.0465643405914307,
+      "learning_rate": 0.0002483371062960701,
+      "loss": 3.2301746368408204,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6186359269932757,
+      "grad_norm": 0.8956519961357117,
+      "learning_rate": 0.0002481570630477849,
+      "loss": 3.226789093017578,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6195965417867435,
+      "grad_norm": 0.8143766522407532,
+      "learning_rate": 0.00024797677213915997,
+      "loss": 3.2227046966552733,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6205571565802114,
+      "grad_norm": 1.0714408159255981,
+      "learning_rate": 0.00024779623402508766,
+      "loss": 3.225188064575195,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6215177713736791,
+      "grad_norm": 1.0253653526306152,
+      "learning_rate": 0.0002476154491610843,
+      "loss": 3.225335693359375,
+      "step": 3235
+    },
+    {
+      "epoch": 0.622478386167147,
+      "grad_norm": 1.0260114669799805,
+      "learning_rate": 0.00024743441800328866,
+      "loss": 3.2250225067138674,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6234390009606148,
+      "grad_norm": 1.0167043209075928,
+      "learning_rate": 0.0002472531410084607,
+      "loss": 3.222640609741211,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6243996157540826,
+      "grad_norm": 1.0281862020492554,
+      "learning_rate": 0.00024707161863398114,
+      "loss": 3.223117446899414,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6253602305475504,
+      "grad_norm": 1.1766680479049683,
+      "learning_rate": 0.0002468898513378494,
+      "loss": 3.2250850677490233,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6263208453410183,
+      "grad_norm": 0.9479981064796448,
+      "learning_rate": 0.0002467078395786831,
+      "loss": 3.2229190826416017,
+      "step": 3260
+    },
+    {
+      "epoch": 0.627281460134486,
+      "grad_norm": 0.9690991640090942,
+      "learning_rate": 0.00024652558381571646,
+      "loss": 3.2238792419433593,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6282420749279539,
+      "grad_norm": 1.127388596534729,
+      "learning_rate": 0.00024634308450879953,
+      "loss": 3.22578125,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6292026897214217,
+      "grad_norm": 1.0700501203536987,
+      "learning_rate": 0.0002461603421183968,
+      "loss": 3.222941589355469,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6301633045148896,
+      "grad_norm": 1.1317108869552612,
+      "learning_rate": 0.00024597735710558613,
+      "loss": 3.228685760498047,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6311239193083573,
+      "grad_norm": 1.0256825685501099,
+      "learning_rate": 0.0002457941299320574,
+      "loss": 3.2236000061035157,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6320845341018252,
+      "grad_norm": 1.0949205160140991,
+      "learning_rate": 0.0002456106610601115,
+      "loss": 3.223644256591797,
+      "step": 3290
+    },
+    {
+      "epoch": 0.633045148895293,
+      "grad_norm": 1.147100567817688,
+      "learning_rate": 0.0002454269509526593,
+      "loss": 3.222336196899414,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6340057636887608,
+      "grad_norm": 1.0118300914764404,
+      "learning_rate": 0.0002452430000732203,
+      "loss": 3.2236583709716795,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6349663784822286,
+      "grad_norm": 1.4460432529449463,
+      "learning_rate": 0.00024505880888592134,
+      "loss": 3.225933074951172,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6359269932756965,
+      "grad_norm": 1.0727514028549194,
+      "learning_rate": 0.0002448743778554957,
+      "loss": 3.222777557373047,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6368876080691642,
+      "grad_norm": 0.9514535069465637,
+      "learning_rate": 0.0002446897074472819,
+      "loss": 3.2218833923339845,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6378482228626321,
+      "grad_norm": 0.8115454912185669,
+      "learning_rate": 0.0002445047981272221,
+      "loss": 3.2228424072265627,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6388088376560999,
+      "grad_norm": 1.2015719413757324,
+      "learning_rate": 0.00024431965036186175,
+      "loss": 3.2233604431152343,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6397694524495677,
+      "grad_norm": 1.0239508152008057,
+      "learning_rate": 0.00024413426461834746,
+      "loss": 3.223210906982422,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6407300672430355,
+      "grad_norm": 0.8938039541244507,
+      "learning_rate": 0.0002439486413644265,
+      "loss": 3.2242591857910154,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6416906820365034,
+      "grad_norm": 1.0734831094741821,
+      "learning_rate": 0.00024376278106844548,
+      "loss": 3.221482849121094,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6426512968299711,
+      "grad_norm": 0.9349012970924377,
+      "learning_rate": 0.00024357668419934894,
+      "loss": 3.2233955383300783,
+      "step": 3345
+    },
+    {
+      "epoch": 0.643611911623439,
+      "grad_norm": 1.0495328903198242,
+      "learning_rate": 0.00024339035122667837,
+      "loss": 3.2246463775634764,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6445725264169068,
+      "grad_norm": 1.152726173400879,
+      "learning_rate": 0.00024320378262057102,
+      "loss": 3.221533203125,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6455331412103746,
+      "grad_norm": 1.1071127653121948,
+      "learning_rate": 0.00024301697885175847,
+      "loss": 3.2202079772949217,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6464937560038425,
+      "grad_norm": 1.143314003944397,
+      "learning_rate": 0.00024282994039156598,
+      "loss": 3.228096771240234,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6474543707973103,
+      "grad_norm": 1.0042517185211182,
+      "learning_rate": 0.00024264266771191066,
+      "loss": 3.2260887145996096,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6484149855907781,
+      "grad_norm": 0.9083545804023743,
+      "learning_rate": 0.00024245516128530077,
+      "loss": 3.2202564239501954,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6493756003842459,
+      "grad_norm": 0.8550533056259155,
+      "learning_rate": 0.0002422674215848342,
+      "loss": 3.22034912109375,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6503362151777138,
+      "grad_norm": 0.9592113494873047,
+      "learning_rate": 0.0002420794490841975,
+      "loss": 3.22401123046875,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6512968299711815,
+      "grad_norm": 0.8267397880554199,
+      "learning_rate": 0.00024189124425766454,
+      "loss": 3.2178253173828124,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6522574447646494,
+      "grad_norm": 0.9734513759613037,
+      "learning_rate": 0.00024170280758009546,
+      "loss": 3.2190723419189453,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6532180595581172,
+      "grad_norm": 1.0415617227554321,
+      "learning_rate": 0.00024151413952693524,
+      "loss": 3.2226608276367186,
+      "step": 3400
+    },
+    {
+      "epoch": 0.654178674351585,
+      "grad_norm": 0.9090771675109863,
+      "learning_rate": 0.00024132524057421287,
+      "loss": 3.221820831298828,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6551392891450528,
+      "grad_norm": 0.8799076676368713,
+      "learning_rate": 0.0002411361111985396,
+      "loss": 3.2193443298339846,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6560999039385207,
+      "grad_norm": 1.0817033052444458,
+      "learning_rate": 0.00024094675187710847,
+      "loss": 3.2185131072998048,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6570605187319885,
+      "grad_norm": 0.9831625819206238,
+      "learning_rate": 0.0002407571630876923,
+      "loss": 3.2214286804199217,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6580211335254563,
+      "grad_norm": 0.987235963344574,
+      "learning_rate": 0.00024056734530864313,
+      "loss": 3.219521713256836,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6589817483189241,
+      "grad_norm": 1.047991394996643,
+      "learning_rate": 0.00024037729901889068,
+      "loss": 3.2219139099121095,
+      "step": 3430
+    },
+    {
+      "epoch": 0.659942363112392,
+      "grad_norm": 0.8672451376914978,
+      "learning_rate": 0.00024018702469794125,
+      "loss": 3.2201393127441404,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6609029779058597,
+      "grad_norm": 1.1878256797790527,
+      "learning_rate": 0.0002399965228258764,
+      "loss": 3.2203353881835937,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6618635926993276,
+      "grad_norm": 1.1251111030578613,
+      "learning_rate": 0.00023980579388335193,
+      "loss": 3.21944580078125,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6628242074927954,
+      "grad_norm": 0.9487177729606628,
+      "learning_rate": 0.0002396148383515966,
+      "loss": 3.2208953857421876,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6637848222862632,
+      "grad_norm": 0.9404230713844299,
+      "learning_rate": 0.00023942365671241072,
+      "loss": 3.22216796875,
+      "step": 3455
+    },
+    {
+      "epoch": 0.664745437079731,
+      "grad_norm": 1.1221790313720703,
+      "learning_rate": 0.00023923224944816522,
+      "loss": 3.220103454589844,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6657060518731989,
+      "grad_norm": 1.1422487497329712,
+      "learning_rate": 0.00023904061704180027,
+      "loss": 3.2190887451171877,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.9574413299560547,
+      "learning_rate": 0.00023884875997682404,
+      "loss": 3.215955352783203,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6676272814601345,
+      "grad_norm": 1.1348673105239868,
+      "learning_rate": 0.00023865667873731168,
+      "loss": 3.2160018920898437,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6685878962536023,
+      "grad_norm": 1.0261040925979614,
+      "learning_rate": 0.00023846437380790368,
+      "loss": 3.221507263183594,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6695485110470701,
+      "grad_norm": 1.009157657623291,
+      "learning_rate": 0.0002382718456738053,
+      "loss": 3.2182361602783205,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6705091258405379,
+      "grad_norm": 0.9334181547164917,
+      "learning_rate": 0.00023807909482078475,
+      "loss": 3.221875762939453,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6714697406340058,
+      "grad_norm": 1.0333514213562012,
+      "learning_rate": 0.0002378861217351721,
+      "loss": 3.2202327728271483,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6724303554274735,
+      "grad_norm": 0.8510705828666687,
+      "learning_rate": 0.00023769292690385831,
+      "loss": 3.2205406188964845,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6733909702209414,
+      "grad_norm": 1.0097713470458984,
+      "learning_rate": 0.0002374995108142938,
+      "loss": 3.2197288513183593,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6743515850144092,
+      "grad_norm": 1.3777557611465454,
+      "learning_rate": 0.0002373058739544871,
+      "loss": 3.2197029113769533,
+      "step": 3510
+    },
+    {
+      "epoch": 0.675312199807877,
+      "grad_norm": 1.1440997123718262,
+      "learning_rate": 0.0002371120168130039,
+      "loss": 3.2237571716308593,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6762728146013448,
+      "grad_norm": 1.1652058362960815,
+      "learning_rate": 0.0002369179398789657,
+      "loss": 3.216419982910156,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6772334293948127,
+      "grad_norm": 0.7069166302680969,
+      "learning_rate": 0.00023672364364204853,
+      "loss": 3.2144538879394533,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6781940441882806,
+      "grad_norm": 1.2215242385864258,
+      "learning_rate": 0.00023652912859248166,
+      "loss": 3.221971130371094,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6791546589817483,
+      "grad_norm": 0.978817880153656,
+      "learning_rate": 0.00023633439522104658,
+      "loss": 3.21834716796875,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6801152737752162,
+      "grad_norm": 1.0850077867507935,
+      "learning_rate": 0.0002361394440190755,
+      "loss": 3.216111755371094,
+      "step": 3540
+    },
+    {
+      "epoch": 0.681075888568684,
+      "grad_norm": 1.0549561977386475,
+      "learning_rate": 0.00023594427547845033,
+      "loss": 3.2226284027099608,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6820365033621518,
+      "grad_norm": 1.0915031433105469,
+      "learning_rate": 0.00023574889009160134,
+      "loss": 3.220775604248047,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6829971181556196,
+      "grad_norm": 1.2648931741714478,
+      "learning_rate": 0.00023555328835150587,
+      "loss": 3.219307708740234,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6839577329490875,
+      "grad_norm": 0.8500558733940125,
+      "learning_rate": 0.00023535747075168726,
+      "loss": 3.2166213989257812,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6849183477425552,
+      "grad_norm": 0.9885119795799255,
+      "learning_rate": 0.00023516143778621327,
+      "loss": 3.219788360595703,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6858789625360231,
+      "grad_norm": 1.1124404668807983,
+      "learning_rate": 0.00023496518994969528,
+      "loss": 3.218476104736328,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6868395773294909,
+      "grad_norm": 1.0537521839141846,
+      "learning_rate": 0.00023476872773728666,
+      "loss": 3.217354583740234,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6878001921229587,
+      "grad_norm": 1.0711948871612549,
+      "learning_rate": 0.00023457205164468173,
+      "loss": 3.2162689208984374,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6887608069164265,
+      "grad_norm": 1.0931755304336548,
+      "learning_rate": 0.0002343751621681145,
+      "loss": 3.217786407470703,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6897214217098944,
+      "grad_norm": 0.9976537227630615,
+      "learning_rate": 0.00023417805980435736,
+      "loss": 3.216781997680664,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6906820365033621,
+      "grad_norm": 0.8797841668128967,
+      "learning_rate": 0.00023398074505071964,
+      "loss": 3.21632080078125,
+      "step": 3595
+    },
+    {
+      "epoch": 0.69164265129683,
+      "grad_norm": 1.0468313694000244,
+      "learning_rate": 0.0002337832184050468,
+      "loss": 3.2173301696777346,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6926032660902978,
+      "grad_norm": 1.2071698904037476,
+      "learning_rate": 0.0002335854803657188,
+      "loss": 3.2191944122314453,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6935638808837656,
+      "grad_norm": 1.0584746599197388,
+      "learning_rate": 0.00023338753143164906,
+      "loss": 3.2196414947509764,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6945244956772334,
+      "grad_norm": 0.9132346510887146,
+      "learning_rate": 0.00023318937210228295,
+      "loss": 3.215652847290039,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6954851104707013,
+      "grad_norm": 0.9715785384178162,
+      "learning_rate": 0.00023299100287759686,
+      "loss": 3.215423583984375,
+      "step": 3620
+    },
+    {
+      "epoch": 0.696445725264169,
+      "grad_norm": 1.1859434843063354,
+      "learning_rate": 0.00023279242425809667,
+      "loss": 3.211594009399414,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6974063400576369,
+      "grad_norm": 1.0725688934326172,
+      "learning_rate": 0.00023259363674481666,
+      "loss": 3.2200057983398436,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6983669548511047,
+      "grad_norm": 1.0033189058303833,
+      "learning_rate": 0.00023239464083931802,
+      "loss": 3.217108154296875,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6993275696445725,
+      "grad_norm": 1.0625147819519043,
+      "learning_rate": 0.00023219543704368792,
+      "loss": 3.2169322967529297,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7002881844380403,
+      "grad_norm": 0.9788894653320312,
+      "learning_rate": 0.00023199602586053793,
+      "loss": 3.2101593017578125,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7012487992315082,
+      "grad_norm": 0.9568902254104614,
+      "learning_rate": 0.0002317964077930029,
+      "loss": 3.215932846069336,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7022094140249759,
+      "grad_norm": 1.3549705743789673,
+      "learning_rate": 0.00023159658334473974,
+      "loss": 3.2176162719726564,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7031700288184438,
+      "grad_norm": 0.9975321292877197,
+      "learning_rate": 0.00023139655301992603,
+      "loss": 3.218878173828125,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7041306436119116,
+      "grad_norm": 0.8793084621429443,
+      "learning_rate": 0.00023119631732325882,
+      "loss": 3.2155433654785157,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7050912584053795,
+      "grad_norm": 1.2263166904449463,
+      "learning_rate": 0.0002309958767599532,
+      "loss": 3.2179286956787108,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7060518731988472,
+      "grad_norm": 1.0385061502456665,
+      "learning_rate": 0.0002307952318357414,
+      "loss": 3.2134330749511717,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7070124879923151,
+      "grad_norm": 1.019902229309082,
+      "learning_rate": 0.0002305943830568711,
+      "loss": 3.2176845550537108,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7079731027857828,
+      "grad_norm": 0.9984716773033142,
+      "learning_rate": 0.00023039333093010433,
+      "loss": 3.218429946899414,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7089337175792507,
+      "grad_norm": 1.0424875020980835,
+      "learning_rate": 0.00023019207596271634,
+      "loss": 3.2165111541748046,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7098943323727186,
+      "grad_norm": 1.0895493030548096,
+      "learning_rate": 0.00022999061866249397,
+      "loss": 3.2162498474121093,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7108549471661864,
+      "grad_norm": 1.0710501670837402,
+      "learning_rate": 0.00022978895953773464,
+      "loss": 3.2143184661865236,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7118155619596542,
+      "grad_norm": 1.1927725076675415,
+      "learning_rate": 0.00022958709909724512,
+      "loss": 3.2180755615234373,
+      "step": 3705
+    },
+    {
+      "epoch": 0.712776176753122,
+      "grad_norm": 0.9517913460731506,
+      "learning_rate": 0.00022938503785033991,
+      "loss": 3.2159660339355467,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7137367915465899,
+      "grad_norm": 1.0126185417175293,
+      "learning_rate": 0.00022918277630684032,
+      "loss": 3.2123146057128906,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7146974063400576,
+      "grad_norm": 1.0625087022781372,
+      "learning_rate": 0.0002289803149770729,
+      "loss": 3.215711212158203,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7156580211335255,
+      "grad_norm": 1.2499691247940063,
+      "learning_rate": 0.00022877765437186843,
+      "loss": 3.215605545043945,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7166186359269933,
+      "grad_norm": 1.0240933895111084,
+      "learning_rate": 0.0002285747950025604,
+      "loss": 3.2169639587402346,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7175792507204611,
+      "grad_norm": 1.0153264999389648,
+      "learning_rate": 0.00022837173738098375,
+      "loss": 3.2149864196777345,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7185398655139289,
+      "grad_norm": 0.7960460782051086,
+      "learning_rate": 0.00022816848201947376,
+      "loss": 3.217276382446289,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7195004803073968,
+      "grad_norm": 0.886752724647522,
+      "learning_rate": 0.0002279650294308645,
+      "loss": 3.2126708984375,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7204610951008645,
+      "grad_norm": 0.9281694889068604,
+      "learning_rate": 0.00022776138012848776,
+      "loss": 3.2142318725585937,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7214217098943324,
+      "grad_norm": 0.9548509120941162,
+      "learning_rate": 0.00022755753462617162,
+      "loss": 3.2146896362304687,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7223823246878002,
+      "grad_norm": 1.099318265914917,
+      "learning_rate": 0.0002273534934382392,
+      "loss": 3.2121913909912108,
+      "step": 3760
+    },
+    {
+      "epoch": 0.723342939481268,
+      "grad_norm": 1.0480355024337769,
+      "learning_rate": 0.00022714925707950734,
+      "loss": 3.2157440185546875,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7243035542747358,
+      "grad_norm": 0.8996202349662781,
+      "learning_rate": 0.0002269448260652853,
+      "loss": 3.2162132263183594,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7252641690682037,
+      "grad_norm": 1.0007187128067017,
+      "learning_rate": 0.00022674020091137357,
+      "loss": 3.216075897216797,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7262247838616714,
+      "grad_norm": 1.0066853761672974,
+      "learning_rate": 0.0002265353821340623,
+      "loss": 3.2114944458007812,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7271853986551393,
+      "grad_norm": 0.8113794922828674,
+      "learning_rate": 0.00022633037025013034,
+      "loss": 3.212854766845703,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7281460134486071,
+      "grad_norm": 0.8382619619369507,
+      "learning_rate": 0.00022612516577684373,
+      "loss": 3.2154186248779295,
+      "step": 3790
+    },
+    {
+      "epoch": 0.729106628242075,
+      "grad_norm": 0.7889550924301147,
+      "learning_rate": 0.0002259197692319544,
+      "loss": 3.212788391113281,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7300672430355427,
+      "grad_norm": 0.988488495349884,
+      "learning_rate": 0.00022571418113369885,
+      "loss": 3.2164848327636717,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7310278578290106,
+      "grad_norm": 1.2173198461532593,
+      "learning_rate": 0.00022550840200079696,
+      "loss": 3.216341018676758,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7319884726224783,
+      "grad_norm": 0.9247362017631531,
+      "learning_rate": 0.00022530243235245067,
+      "loss": 3.214260482788086,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7329490874159462,
+      "grad_norm": 1.0379600524902344,
+      "learning_rate": 0.0002250962727083424,
+      "loss": 3.211573028564453,
+      "step": 3815
+    },
+    {
+      "epoch": 0.733909702209414,
+      "grad_norm": 1.0794739723205566,
+      "learning_rate": 0.00022488992358863416,
+      "loss": 3.2143383026123047,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7348703170028819,
+      "grad_norm": 0.7588670253753662,
+      "learning_rate": 0.00022468338551396598,
+      "loss": 3.2139129638671875,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7358309317963496,
+      "grad_norm": 0.9358298182487488,
+      "learning_rate": 0.00022447665900545452,
+      "loss": 3.2157051086425783,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7367915465898175,
+      "grad_norm": 1.0112359523773193,
+      "learning_rate": 0.00022426974458469206,
+      "loss": 3.2114688873291017,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7377521613832853,
+      "grad_norm": 1.0056138038635254,
+      "learning_rate": 0.00022406264277374487,
+      "loss": 3.211741638183594,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7387127761767531,
+      "grad_norm": 1.0002721548080444,
+      "learning_rate": 0.00022385535409515203,
+      "loss": 3.214643096923828,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7396733909702209,
+      "grad_norm": 1.0511516332626343,
+      "learning_rate": 0.0002236478790719242,
+      "loss": 3.2131919860839844,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7406340057636888,
+      "grad_norm": 1.0013073682785034,
+      "learning_rate": 0.00022344021822754212,
+      "loss": 3.2114883422851563,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7415946205571565,
+      "grad_norm": 0.9666048288345337,
+      "learning_rate": 0.0002232323720859554,
+      "loss": 3.2139522552490236,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7425552353506244,
+      "grad_norm": 0.9765871167182922,
+      "learning_rate": 0.0002230243411715812,
+      "loss": 3.216571044921875,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7435158501440923,
+      "grad_norm": 0.8558294773101807,
+      "learning_rate": 0.00022281612600930282,
+      "loss": 3.211492156982422,
+      "step": 3870
+    },
+    {
+      "epoch": 0.74447646493756,
+      "grad_norm": 0.9790803790092468,
+      "learning_rate": 0.00022260772712446848,
+      "loss": 3.212940979003906,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7454370797310279,
+      "grad_norm": 1.115885853767395,
+      "learning_rate": 0.00022239914504289002,
+      "loss": 3.215241241455078,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7463976945244957,
+      "grad_norm": 0.8012208342552185,
+      "learning_rate": 0.00022219038029084134,
+      "loss": 3.2129531860351563,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7473583093179635,
+      "grad_norm": 0.9198108315467834,
+      "learning_rate": 0.00022198143339505736,
+      "loss": 3.2131282806396486,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7483189241114313,
+      "grad_norm": 0.9384759068489075,
+      "learning_rate": 0.00022177230488273255,
+      "loss": 3.2148651123046874,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7492795389048992,
+      "grad_norm": 0.9330629706382751,
+      "learning_rate": 0.00022156299528151957,
+      "loss": 3.2130111694335937,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7502401536983669,
+      "grad_norm": 1.0961406230926514,
+      "learning_rate": 0.00022135350511952805,
+      "loss": 3.209431457519531,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7512007684918348,
+      "grad_norm": 1.0598336458206177,
+      "learning_rate": 0.0002211438349253231,
+      "loss": 3.2150421142578125,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7521613832853026,
+      "grad_norm": 0.8614126443862915,
+      "learning_rate": 0.00022093398522792418,
+      "loss": 3.2113304138183594,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7531219980787704,
+      "grad_norm": 1.2163376808166504,
+      "learning_rate": 0.0002207239565568036,
+      "loss": 3.213220977783203,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7540826128722382,
+      "grad_norm": 1.068865180015564,
+      "learning_rate": 0.00022051374944188524,
+      "loss": 3.2152915954589845,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7550432276657061,
+      "grad_norm": 0.8628993034362793,
+      "learning_rate": 0.0002203033644135432,
+      "loss": 3.2100101470947267,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7560038424591738,
+      "grad_norm": 1.0070104598999023,
+      "learning_rate": 0.00022009280200260053,
+      "loss": 3.214314651489258,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7569644572526417,
+      "grad_norm": 1.1911405324935913,
+      "learning_rate": 0.00021988206274032776,
+      "loss": 3.2116607666015624,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7579250720461095,
+      "grad_norm": 1.0605741739273071,
+      "learning_rate": 0.00021967114715844163,
+      "loss": 3.210451126098633,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7588856868395774,
+      "grad_norm": 1.010385274887085,
+      "learning_rate": 0.0002194600557891039,
+      "loss": 3.2117488861083983,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7598463016330451,
+      "grad_norm": 0.9285851716995239,
+      "learning_rate": 0.00021924878916491962,
+      "loss": 3.2095130920410155,
+      "step": 3955
+    },
+    {
+      "epoch": 0.760806916426513,
+      "grad_norm": 0.9447981715202332,
+      "learning_rate": 0.00021903734781893625,
+      "loss": 3.214548873901367,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7617675312199808,
+      "grad_norm": 1.0844841003417969,
+      "learning_rate": 0.000218825732284642,
+      "loss": 3.2142662048339843,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7627281460134486,
+      "grad_norm": 0.9654716849327087,
+      "learning_rate": 0.00021861394309596446,
+      "loss": 3.214899444580078,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7636887608069164,
+      "grad_norm": 0.8219507336616516,
+      "learning_rate": 0.00021840198078726964,
+      "loss": 3.2075408935546874,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7646493756003843,
+      "grad_norm": 1.0514135360717773,
+      "learning_rate": 0.00021818984589336006,
+      "loss": 3.2124794006347654,
+      "step": 3980
+    },
+    {
+      "epoch": 0.765609990393852,
+      "grad_norm": 0.764786958694458,
+      "learning_rate": 0.0002179775389494739,
+      "loss": 3.210370635986328,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7665706051873199,
+      "grad_norm": 0.9463275671005249,
+      "learning_rate": 0.0002177650604912833,
+      "loss": 3.214555358886719,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7675312199807877,
+      "grad_norm": 0.9087357521057129,
+      "learning_rate": 0.0002175524110548932,
+      "loss": 3.211328125,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7684918347742555,
+      "grad_norm": 0.8839055299758911,
+      "learning_rate": 0.00021733959117684008,
+      "loss": 3.208209991455078,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7694524495677233,
+      "grad_norm": 1.0021069049835205,
+      "learning_rate": 0.00021712660139409015,
+      "loss": 3.213267517089844,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7704130643611912,
+      "grad_norm": 1.089827060699463,
+      "learning_rate": 0.0002169134422440386,
+      "loss": 3.209787368774414,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7713736791546589,
+      "grad_norm": 0.9058536887168884,
+      "learning_rate": 0.00021670011426450772,
+      "loss": 3.2052375793457033,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7723342939481268,
+      "grad_norm": 0.8882672190666199,
+      "learning_rate": 0.000216486617993746,
+      "loss": 3.2103240966796873,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7732949087415946,
+      "grad_norm": 0.7981977462768555,
+      "learning_rate": 0.00021627295397042635,
+      "loss": 3.2096282958984377,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7742555235350624,
+      "grad_norm": 0.9157530665397644,
+      "learning_rate": 0.00021605912273364513,
+      "loss": 3.213287353515625,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7752161383285303,
+      "grad_norm": 0.7963248491287231,
+      "learning_rate": 0.00021584512482292038,
+      "loss": 3.2098575592041017,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7761767531219981,
+      "grad_norm": 1.0376338958740234,
+      "learning_rate": 0.00021563096077819083,
+      "loss": 3.2072002410888674,
+      "step": 4040
+    },
+    {
+      "epoch": 0.777137367915466,
+      "grad_norm": 1.159191608428955,
+      "learning_rate": 0.00021541663113981433,
+      "loss": 3.2127113342285156,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7780979827089337,
+      "grad_norm": 1.213563323020935,
+      "learning_rate": 0.0002152021364485665,
+      "loss": 3.2094375610351564,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7790585975024016,
+      "grad_norm": 1.1174225807189941,
+      "learning_rate": 0.00021498747724563953,
+      "loss": 3.2087074279785157,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7800192122958693,
+      "grad_norm": 1.0365922451019287,
+      "learning_rate": 0.00021477265407264051,
+      "loss": 3.2079147338867187,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7809798270893372,
+      "grad_norm": 1.0354453325271606,
+      "learning_rate": 0.00021455766747159044,
+      "loss": 3.2089813232421873,
+      "step": 4065
+    },
+    {
+      "epoch": 0.781940441882805,
+      "grad_norm": 1.0720642805099487,
+      "learning_rate": 0.0002143425179849226,
+      "loss": 3.207681655883789,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7829010566762729,
+      "grad_norm": 0.9447735548019409,
+      "learning_rate": 0.00021412720615548105,
+      "loss": 3.2113746643066405,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7838616714697406,
+      "grad_norm": 0.9376322627067566,
+      "learning_rate": 0.00021391173252651978,
+      "loss": 3.2086830139160156,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7848222862632085,
+      "grad_norm": 1.0265626907348633,
+      "learning_rate": 0.00021369609764170075,
+      "loss": 3.2098388671875,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7857829010566763,
+      "grad_norm": 1.1230889558792114,
+      "learning_rate": 0.00021348030204509303,
+      "loss": 3.208446502685547,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7867435158501441,
+      "grad_norm": 0.8800603747367859,
+      "learning_rate": 0.00021326434628117088,
+      "loss": 3.208513641357422,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7877041306436119,
+      "grad_norm": 1.0674163103103638,
+      "learning_rate": 0.00021304823089481298,
+      "loss": 3.2110557556152344,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7886647454370798,
+      "grad_norm": 0.8815765976905823,
+      "learning_rate": 0.00021283195643130058,
+      "loss": 3.210008239746094,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7896253602305475,
+      "grad_norm": 1.0902719497680664,
+      "learning_rate": 0.00021261552343631633,
+      "loss": 3.2107887268066406,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7905859750240154,
+      "grad_norm": 0.859643816947937,
+      "learning_rate": 0.00021239893245594287,
+      "loss": 3.210626220703125,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7915465898174832,
+      "grad_norm": 0.9734321236610413,
+      "learning_rate": 0.00021218218403666148,
+      "loss": 3.207859420776367,
+      "step": 4120
+    },
+    {
+      "epoch": 0.792507204610951,
+      "grad_norm": 0.8096666932106018,
+      "learning_rate": 0.00021196527872535068,
+      "loss": 3.2104820251464843,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7934678194044188,
+      "grad_norm": 1.0135759115219116,
+      "learning_rate": 0.0002117482170692847,
+      "loss": 3.2095088958740234,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7944284341978867,
+      "grad_norm": 0.9503042101860046,
+      "learning_rate": 0.00021153099961613257,
+      "loss": 3.2078937530517577,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7953890489913544,
+      "grad_norm": 0.8183991312980652,
+      "learning_rate": 0.00021131362691395608,
+      "loss": 3.2082897186279298,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7963496637848223,
+      "grad_norm": 0.9838517308235168,
+      "learning_rate": 0.00021109609951120887,
+      "loss": 3.2070068359375,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7973102785782901,
+      "grad_norm": 1.0721310377120972,
+      "learning_rate": 0.00021087841795673488,
+      "loss": 3.206983947753906,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7982708933717579,
+      "grad_norm": 1.0762015581130981,
+      "learning_rate": 0.00021066058279976704,
+      "loss": 3.2114967346191405,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7992315081652257,
+      "grad_norm": 0.7675787806510925,
+      "learning_rate": 0.0002104425945899258,
+      "loss": 3.2034461975097654,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8001921229586936,
+      "grad_norm": 1.0134726762771606,
+      "learning_rate": 0.00021022445387721767,
+      "loss": 3.207468032836914,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8011527377521613,
+      "grad_norm": 0.9967617988586426,
+      "learning_rate": 0.00021000616121203422,
+      "loss": 3.2145057678222657,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8021133525456292,
+      "grad_norm": 1.0330275297164917,
+      "learning_rate": 0.00020978771714515016,
+      "loss": 3.205029296875,
+      "step": 4175
+    },
+    {
+      "epoch": 0.803073967339097,
+      "grad_norm": 0.8626118898391724,
+      "learning_rate": 0.00020956912222772222,
+      "loss": 3.2084095001220705,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8040345821325648,
+      "grad_norm": 1.186023235321045,
+      "learning_rate": 0.0002093503770112879,
+      "loss": 3.209725189208984,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8049951969260326,
+      "grad_norm": 0.943437397480011,
+      "learning_rate": 0.00020913148204776378,
+      "loss": 3.209857177734375,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8059558117195005,
+      "grad_norm": 1.167366623878479,
+      "learning_rate": 0.00020891243788944428,
+      "loss": 3.2103233337402344,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8069164265129684,
+      "grad_norm": 0.9998597502708435,
+      "learning_rate": 0.00020869324508900026,
+      "loss": 3.211322784423828,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8078770413064361,
+      "grad_norm": 0.9104480147361755,
+      "learning_rate": 0.00020847390419947778,
+      "loss": 3.2085220336914064,
+      "step": 4205
+    },
+    {
+      "epoch": 0.808837656099904,
+      "grad_norm": 1.0278260707855225,
+      "learning_rate": 0.0002082544157742963,
+      "loss": 3.2045223236083986,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8097982708933718,
+      "grad_norm": 0.9279727339744568,
+      "learning_rate": 0.0002080347803672476,
+      "loss": 3.2089134216308595,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8107588856868396,
+      "grad_norm": 0.9136155247688293,
+      "learning_rate": 0.00020781499853249441,
+      "loss": 3.2083240509033204,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8117195004803074,
+      "grad_norm": 0.8539382219314575,
+      "learning_rate": 0.00020759507082456885,
+      "loss": 3.204279327392578,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8126801152737753,
+      "grad_norm": 1.1511064767837524,
+      "learning_rate": 0.00020737499779837105,
+      "loss": 3.2057140350341795,
+      "step": 4230
+    },
+    {
+      "epoch": 0.813640730067243,
+      "grad_norm": 0.8153745532035828,
+      "learning_rate": 0.00020715478000916783,
+      "loss": 3.2046875,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8146013448607109,
+      "grad_norm": 0.8916031122207642,
+      "learning_rate": 0.00020693441801259135,
+      "loss": 3.210984802246094,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8155619596541787,
+      "grad_norm": 0.9601288437843323,
+      "learning_rate": 0.00020671391236463742,
+      "loss": 3.2044132232666014,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8165225744476465,
+      "grad_norm": 0.9480450749397278,
+      "learning_rate": 0.00020649326362166449,
+      "loss": 3.2047096252441407,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8174831892411143,
+      "grad_norm": 1.3009711503982544,
+      "learning_rate": 0.0002062724723403919,
+      "loss": 3.210066223144531,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8184438040345822,
+      "grad_norm": 0.8482257127761841,
+      "learning_rate": 0.00020605153907789876,
+      "loss": 3.2040420532226563,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8194044188280499,
+      "grad_norm": 1.1215349435806274,
+      "learning_rate": 0.00020583046439162232,
+      "loss": 3.209978485107422,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8203650336215178,
+      "grad_norm": 1.0416927337646484,
+      "learning_rate": 0.00020560924883935674,
+      "loss": 3.20634765625,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8213256484149856,
+      "grad_norm": 0.9002500176429749,
+      "learning_rate": 0.00020538789297925154,
+      "loss": 3.210291290283203,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8222862632084534,
+      "grad_norm": 0.7959718108177185,
+      "learning_rate": 0.00020516639736981027,
+      "loss": 3.205914306640625,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8232468780019212,
+      "grad_norm": 0.9458216428756714,
+      "learning_rate": 0.000204944762569889,
+      "loss": 3.2048385620117186,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8242074927953891,
+      "grad_norm": 1.0240002870559692,
+      "learning_rate": 0.00020472298913869514,
+      "loss": 3.2106887817382814,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8251681075888568,
+      "grad_norm": 0.9878635406494141,
+      "learning_rate": 0.00020450107763578582,
+      "loss": 3.206460952758789,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8261287223823247,
+      "grad_norm": 0.961821973323822,
+      "learning_rate": 0.00020427902862106645,
+      "loss": 3.208537292480469,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8270893371757925,
+      "grad_norm": 1.0129098892211914,
+      "learning_rate": 0.00020405684265478955,
+      "loss": 3.2066085815429686,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8280499519692603,
+      "grad_norm": 1.2535499334335327,
+      "learning_rate": 0.00020383452029755308,
+      "loss": 3.206789016723633,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8290105667627281,
+      "grad_norm": 1.0553781986236572,
+      "learning_rate": 0.00020361206211029915,
+      "loss": 3.208069610595703,
+      "step": 4315
+    },
+    {
+      "epoch": 0.829971181556196,
+      "grad_norm": 1.033327341079712,
+      "learning_rate": 0.00020338946865431257,
+      "loss": 3.208452606201172,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8309317963496637,
+      "grad_norm": 1.1007755994796753,
+      "learning_rate": 0.00020316674049121954,
+      "loss": 3.205517578125,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8318924111431316,
+      "grad_norm": 1.0913561582565308,
+      "learning_rate": 0.00020294387818298596,
+      "loss": 3.205488586425781,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8328530259365994,
+      "grad_norm": 1.073475956916809,
+      "learning_rate": 0.00020272088229191638,
+      "loss": 3.2070037841796877,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8338136407300673,
+      "grad_norm": 0.8359492421150208,
+      "learning_rate": 0.00020249775338065224,
+      "loss": 3.2025604248046875,
+      "step": 4340
+    },
+    {
+      "epoch": 0.834774255523535,
+      "grad_norm": 1.1163265705108643,
+      "learning_rate": 0.0002022744920121707,
+      "loss": 3.205812454223633,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8357348703170029,
+      "grad_norm": 1.0195002555847168,
+      "learning_rate": 0.0002020510987497832,
+      "loss": 3.2072071075439452,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8366954851104706,
+      "grad_norm": 1.06333327293396,
+      "learning_rate": 0.0002018275741571337,
+      "loss": 3.2045135498046875,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8376560999039385,
+      "grad_norm": 1.07416570186615,
+      "learning_rate": 0.00020160391879819775,
+      "loss": 3.2041862487792967,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8386167146974063,
+      "grad_norm": 0.8247685432434082,
+      "learning_rate": 0.00020138013323728072,
+      "loss": 3.204395294189453,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8395773294908742,
+      "grad_norm": 0.7512625455856323,
+      "learning_rate": 0.00020115621803901658,
+      "loss": 3.2026763916015626,
+      "step": 4370
+    },
+    {
+      "epoch": 0.840537944284342,
+      "grad_norm": 0.6788026094436646,
+      "learning_rate": 0.0002009321737683664,
+      "loss": 3.202665328979492,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8414985590778098,
+      "grad_norm": 1.178909182548523,
+      "learning_rate": 0.00020070800099061676,
+      "loss": 3.203816604614258,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8424591738712777,
+      "grad_norm": 1.056028127670288,
+      "learning_rate": 0.00020048370027137864,
+      "loss": 3.203491973876953,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8434197886647454,
+      "grad_norm": 0.9850196838378906,
+      "learning_rate": 0.0002002592721765857,
+      "loss": 3.203214645385742,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8443804034582133,
+      "grad_norm": 0.812002956867218,
+      "learning_rate": 0.00020003471727249317,
+      "loss": 3.2064640045166017,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8453410182516811,
+      "grad_norm": 0.7780818343162537,
+      "learning_rate": 0.00019981003612567594,
+      "loss": 3.2039905548095704,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8463016330451489,
+      "grad_norm": 0.7282761931419373,
+      "learning_rate": 0.0001995852293030277,
+      "loss": 3.2027183532714845,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8472622478386167,
+      "grad_norm": 1.597205400466919,
+      "learning_rate": 0.0001993602973717591,
+      "loss": 3.2074440002441404,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8482228626320846,
+      "grad_norm": 0.7743385434150696,
+      "learning_rate": 0.0001991352408993965,
+      "loss": 3.20322265625,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8491834774255523,
+      "grad_norm": 0.8802571296691895,
+      "learning_rate": 0.00019891006045378047,
+      "loss": 3.2041389465332033,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8501440922190202,
+      "grad_norm": 0.9674323797225952,
+      "learning_rate": 0.00019868475660306435,
+      "loss": 3.2038337707519533,
+      "step": 4425
+    },
+    {
+      "epoch": 0.851104707012488,
+      "grad_norm": 0.8163370490074158,
+      "learning_rate": 0.00019845932991571284,
+      "loss": 3.204595947265625,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8520653218059558,
+      "grad_norm": 0.8586170077323914,
+      "learning_rate": 0.00019823378096050067,
+      "loss": 3.204967498779297,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8530259365994236,
+      "grad_norm": 0.9709478616714478,
+      "learning_rate": 0.00019800811030651095,
+      "loss": 3.200875091552734,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8539865513928915,
+      "grad_norm": 0.8611642718315125,
+      "learning_rate": 0.00019778231852313386,
+      "loss": 3.2068603515625,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8549471661863592,
+      "grad_norm": 0.8717654347419739,
+      "learning_rate": 0.00019755640618006532,
+      "loss": 3.2012969970703127,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8559077809798271,
+      "grad_norm": 0.9735947251319885,
+      "learning_rate": 0.0001973303738473053,
+      "loss": 3.2008651733398437,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8568683957732949,
+      "grad_norm": 0.9653201699256897,
+      "learning_rate": 0.0001971042220951565,
+      "loss": 3.2013267517089843,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8578290105667628,
+      "grad_norm": 0.8183594346046448,
+      "learning_rate": 0.00019687795149422315,
+      "loss": 3.204022979736328,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8587896253602305,
+      "grad_norm": 0.8697395920753479,
+      "learning_rate": 0.00019665156261540898,
+      "loss": 3.202830505371094,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8597502401536984,
+      "grad_norm": 0.829225480556488,
+      "learning_rate": 0.00019642505602991654,
+      "loss": 3.2021392822265624,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8607108549471661,
+      "grad_norm": 1.1149920225143433,
+      "learning_rate": 0.0001961984323092451,
+      "loss": 3.205643081665039,
+      "step": 4480
+    },
+    {
+      "epoch": 0.861671469740634,
+      "grad_norm": 0.8502190113067627,
+      "learning_rate": 0.00019597169202518954,
+      "loss": 3.2061065673828124,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8626320845341018,
+      "grad_norm": 0.7430519461631775,
+      "learning_rate": 0.0001957448357498389,
+      "loss": 3.2033714294433593,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8635926993275697,
+      "grad_norm": 0.8692427277565002,
+      "learning_rate": 0.0001955178640555748,
+      "loss": 3.2003097534179688,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8645533141210374,
+      "grad_norm": 0.809934139251709,
+      "learning_rate": 0.00019529077751507016,
+      "loss": 3.2053035736083983,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8655139289145053,
+      "grad_norm": 0.6470674276351929,
+      "learning_rate": 0.00019506357670128749,
+      "loss": 3.2052642822265627,
+      "step": 4505
+    },
+    {
+      "epoch": 0.866474543707973,
+      "grad_norm": 0.9643762111663818,
+      "learning_rate": 0.00019483626218747794,
+      "loss": 3.203343963623047,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8674351585014409,
+      "grad_norm": 1.0639313459396362,
+      "learning_rate": 0.00019460883454717922,
+      "loss": 3.2024681091308596,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8683957732949087,
+      "grad_norm": 0.8861836194992065,
+      "learning_rate": 0.00019438129435421462,
+      "loss": 3.201047897338867,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8693563880883766,
+      "grad_norm": 1.0320836305618286,
+      "learning_rate": 0.00019415364218269146,
+      "loss": 3.2003639221191404,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8703170028818443,
+      "grad_norm": 0.9835689663887024,
+      "learning_rate": 0.00019392587860699942,
+      "loss": 3.2048595428466795,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8712776176753122,
+      "grad_norm": 0.7003117799758911,
+      "learning_rate": 0.00019369800420180943,
+      "loss": 3.204084014892578,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8722382324687801,
+      "grad_norm": 0.9389500021934509,
+      "learning_rate": 0.00019347001954207193,
+      "loss": 3.2038955688476562,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8731988472622478,
+      "grad_norm": 1.1108181476593018,
+      "learning_rate": 0.00019324192520301566,
+      "loss": 3.200461196899414,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8741594620557157,
+      "grad_norm": 0.84727942943573,
+      "learning_rate": 0.00019301372176014605,
+      "loss": 3.2037403106689455,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8751200768491835,
+      "grad_norm": 0.776715874671936,
+      "learning_rate": 0.00019278540978924378,
+      "loss": 3.1999557495117186,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8760806916426513,
+      "grad_norm": 0.873356282711029,
+      "learning_rate": 0.0001925569898663633,
+      "loss": 3.201426696777344,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8770413064361191,
+      "grad_norm": 1.0719802379608154,
+      "learning_rate": 0.00019232846256783163,
+      "loss": 3.2020469665527345,
+      "step": 4565
+    },
+    {
+      "epoch": 0.878001921229587,
+      "grad_norm": 0.8912343978881836,
+      "learning_rate": 0.00019209982847024655,
+      "loss": 3.200605010986328,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8789625360230547,
+      "grad_norm": 0.8289459347724915,
+      "learning_rate": 0.00019187108815047523,
+      "loss": 3.2027835845947266,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8799231508165226,
+      "grad_norm": 0.7044758200645447,
+      "learning_rate": 0.00019164224218565313,
+      "loss": 3.2026336669921873,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8808837656099904,
+      "grad_norm": 0.8316428661346436,
+      "learning_rate": 0.00019141329115318203,
+      "loss": 3.204214096069336,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8818443804034583,
+      "grad_norm": 0.9814732074737549,
+      "learning_rate": 0.00019118423563072885,
+      "loss": 3.2025177001953127,
+      "step": 4590
+    },
+    {
+      "epoch": 0.882804995196926,
+      "grad_norm": 0.6341441869735718,
+      "learning_rate": 0.0001909550761962242,
+      "loss": 3.2031497955322266,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8837656099903939,
+      "grad_norm": 0.977326512336731,
+      "learning_rate": 0.00019072581342786084,
+      "loss": 3.199177551269531,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8847262247838616,
+      "grad_norm": 0.7784291505813599,
+      "learning_rate": 0.00019049644790409225,
+      "loss": 3.198267936706543,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8856868395773295,
+      "grad_norm": 1.1206945180892944,
+      "learning_rate": 0.00019026698020363107,
+      "loss": 3.205250549316406,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8866474543707973,
+      "grad_norm": 0.8689852356910706,
+      "learning_rate": 0.00019003741090544804,
+      "loss": 3.2022071838378907,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8876080691642652,
+      "grad_norm": 1.1133830547332764,
+      "learning_rate": 0.00018980774058876995,
+      "loss": 3.2035804748535157,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8885686839577329,
+      "grad_norm": 0.8632625937461853,
+      "learning_rate": 0.00018957796983307858,
+      "loss": 3.2034885406494142,
+      "step": 4625
+    },
+    {
+      "epoch": 0.8895292987512008,
+      "grad_norm": 0.8747223615646362,
+      "learning_rate": 0.0001893480992181091,
+      "loss": 3.2007545471191405,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8904899135446686,
+      "grad_norm": 1.075469732284546,
+      "learning_rate": 0.00018911812932384872,
+      "loss": 3.2042964935302733,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8914505283381364,
+      "grad_norm": 1.0559624433517456,
+      "learning_rate": 0.000188888060730535,
+      "loss": 3.199691963195801,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8924111431316042,
+      "grad_norm": 1.0888924598693848,
+      "learning_rate": 0.00018865789401865458,
+      "loss": 3.202299118041992,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8933717579250721,
+      "grad_norm": 0.7396308779716492,
+      "learning_rate": 0.0001884276297689418,
+      "loss": 3.19757080078125,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8943323727185398,
+      "grad_norm": 0.8118330240249634,
+      "learning_rate": 0.00018819726856237685,
+      "loss": 3.2041015625,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8952929875120077,
+      "grad_norm": 0.9625673890113831,
+      "learning_rate": 0.00018796681098018486,
+      "loss": 3.2014694213867188,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8962536023054755,
+      "grad_norm": 1.0025596618652344,
+      "learning_rate": 0.00018773625760383375,
+      "loss": 3.2010726928710938,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8972142170989433,
+      "grad_norm": 1.0931344032287598,
+      "learning_rate": 0.00018750560901503353,
+      "loss": 3.2006298065185548,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8981748318924111,
+      "grad_norm": 1.0470645427703857,
+      "learning_rate": 0.00018727486579573407,
+      "loss": 3.1973636627197264,
+      "step": 4675
+    },
+    {
+      "epoch": 0.899135446685879,
+      "grad_norm": 0.9046833515167236,
+      "learning_rate": 0.00018704402852812431,
+      "loss": 3.2017478942871094,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9000960614793467,
+      "grad_norm": 0.9769369959831238,
+      "learning_rate": 0.00018681309779463033,
+      "loss": 3.2017196655273437,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9010566762728146,
+      "grad_norm": 0.8879526853561401,
+      "learning_rate": 0.00018658207417791405,
+      "loss": 3.2009010314941406,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9020172910662824,
+      "grad_norm": 0.7592387795448303,
+      "learning_rate": 0.00018635095826087175,
+      "loss": 3.2003753662109373,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9029779058597502,
+      "grad_norm": 0.9972590804100037,
+      "learning_rate": 0.00018611975062663263,
+      "loss": 3.1992095947265624,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9039385206532181,
+      "grad_norm": 0.9059445261955261,
+      "learning_rate": 0.0001858884518585572,
+      "loss": 3.195656585693359,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9048991354466859,
+      "grad_norm": 0.8296246528625488,
+      "learning_rate": 0.00018565706254023605,
+      "loss": 3.196135711669922,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9058597502401537,
+      "grad_norm": 0.9912506937980652,
+      "learning_rate": 0.00018542558325548814,
+      "loss": 3.1977779388427736,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9068203650336215,
+      "grad_norm": 0.7257289290428162,
+      "learning_rate": 0.00018519401458835948,
+      "loss": 3.197937774658203,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9077809798270894,
+      "grad_norm": 1.0594534873962402,
+      "learning_rate": 0.00018496235712312154,
+      "loss": 3.193804168701172,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9087415946205571,
+      "grad_norm": 1.0351169109344482,
+      "learning_rate": 0.00018473061144426986,
+      "loss": 3.196752166748047,
+      "step": 4730
+    },
+    {
+      "epoch": 0.909702209414025,
+      "grad_norm": 0.8392277359962463,
+      "learning_rate": 0.0001844987781365226,
+      "loss": 3.2042442321777345,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9106628242074928,
+      "grad_norm": 0.6720691323280334,
+      "learning_rate": 0.00018426685778481897,
+      "loss": 3.2033058166503907,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9116234390009607,
+      "grad_norm": 1.009544014930725,
+      "learning_rate": 0.00018403485097431778,
+      "loss": 3.200358200073242,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9125840537944284,
+      "grad_norm": 0.8520932793617249,
+      "learning_rate": 0.00018380275829039602,
+      "loss": 3.201191711425781,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9135446685878963,
+      "grad_norm": 0.9514408111572266,
+      "learning_rate": 0.0001835705803186474,
+      "loss": 3.194804000854492,
+      "step": 4755
+    },
+    {
+      "epoch": 0.914505283381364,
+      "grad_norm": 0.8069214224815369,
+      "learning_rate": 0.00018333831764488065,
+      "loss": 3.1965736389160155,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9154658981748319,
+      "grad_norm": 0.8542043566703796,
+      "learning_rate": 0.00018310597085511844,
+      "loss": 3.1960010528564453,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9164265129682997,
+      "grad_norm": 0.7166752815246582,
+      "learning_rate": 0.00018287354053559546,
+      "loss": 3.1996959686279296,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9173871277617676,
+      "grad_norm": 1.2815790176391602,
+      "learning_rate": 0.0001826410272727574,
+      "loss": 3.200403594970703,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9183477425552353,
+      "grad_norm": 0.9036345481872559,
+      "learning_rate": 0.00018240843165325882,
+      "loss": 3.203938674926758,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9193083573487032,
+      "grad_norm": 0.9565702080726624,
+      "learning_rate": 0.00018217575426396256,
+      "loss": 3.197901153564453,
+      "step": 4785
+    },
+    {
+      "epoch": 0.920268972142171,
+      "grad_norm": 0.8386558890342712,
+      "learning_rate": 0.00018194299569193747,
+      "loss": 3.194584274291992,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9212295869356388,
+      "grad_norm": 1.0965979099273682,
+      "learning_rate": 0.0001817101565244573,
+      "loss": 3.20220947265625,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9221902017291066,
+      "grad_norm": 1.1644456386566162,
+      "learning_rate": 0.00018147723734899916,
+      "loss": 3.1982275009155274,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9231508165225745,
+      "grad_norm": 1.0380982160568237,
+      "learning_rate": 0.00018124423875324198,
+      "loss": 3.1980897903442385,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9241114313160422,
+      "grad_norm": 0.823890209197998,
+      "learning_rate": 0.00018101116132506522,
+      "loss": 3.1994667053222656,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9250720461095101,
+      "grad_norm": 0.5867325663566589,
+      "learning_rate": 0.00018077800565254702,
+      "loss": 3.1975902557373046,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9260326609029779,
+      "grad_norm": 1.1908833980560303,
+      "learning_rate": 0.00018054477232396312,
+      "loss": 3.198208808898926,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9269932756964457,
+      "grad_norm": 1.0711902379989624,
+      "learning_rate": 0.00018031146192778517,
+      "loss": 3.1953447341918944,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9279538904899135,
+      "grad_norm": 0.9716858863830566,
+      "learning_rate": 0.0001800780750526792,
+      "loss": 3.1982431411743164,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9289145052833814,
+      "grad_norm": 0.9620060920715332,
+      "learning_rate": 0.00017984461228750422,
+      "loss": 3.196470260620117,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9298751200768491,
+      "grad_norm": 1.015932321548462,
+      "learning_rate": 0.00017961107422131075,
+      "loss": 3.199555206298828,
+      "step": 4840
+    },
+    {
+      "epoch": 0.930835734870317,
+      "grad_norm": 0.9855188727378845,
+      "learning_rate": 0.00017937746144333934,
+      "loss": 3.2012577056884766,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9317963496637848,
+      "grad_norm": 0.8330841660499573,
+      "learning_rate": 0.0001791437745430189,
+      "loss": 3.1955251693725586,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9327569644572526,
+      "grad_norm": 0.8580245971679688,
+      "learning_rate": 0.00017891001410996554,
+      "loss": 3.1982753753662108,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9337175792507204,
+      "grad_norm": 1.1361298561096191,
+      "learning_rate": 0.00017867618073398074,
+      "loss": 3.202219009399414,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9346781940441883,
+      "grad_norm": 0.9495155215263367,
+      "learning_rate": 0.00017844227500505016,
+      "loss": 3.1957366943359373,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9356388088376562,
+      "grad_norm": 0.8877089619636536,
+      "learning_rate": 0.00017820829751334194,
+      "loss": 3.198602867126465,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9365994236311239,
+      "grad_norm": 0.7768687605857849,
+      "learning_rate": 0.0001779742488492052,
+      "loss": 3.1972635269165037,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9375600384245918,
+      "grad_norm": 1.0583457946777344,
+      "learning_rate": 0.00017774012960316887,
+      "loss": 3.1948143005371095,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9385206532180596,
+      "grad_norm": 0.9406375288963318,
+      "learning_rate": 0.00017750594036593974,
+      "loss": 3.1969860076904295,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9394812680115274,
+      "grad_norm": 1.0013177394866943,
+      "learning_rate": 0.00017727168172840133,
+      "loss": 3.1973243713378907,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9404418828049952,
+      "grad_norm": 0.7509773373603821,
+      "learning_rate": 0.00017703735428161216,
+      "loss": 3.198769760131836,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9414024975984631,
+      "grad_norm": 0.8109461069107056,
+      "learning_rate": 0.00017680295861680445,
+      "loss": 3.1947399139404298,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9423631123919308,
+      "grad_norm": 0.8327275514602661,
+      "learning_rate": 0.0001765684953253825,
+      "loss": 3.196604919433594,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9433237271853987,
+      "grad_norm": 0.8777264356613159,
+      "learning_rate": 0.00017633396499892125,
+      "loss": 3.197701644897461,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9442843419788665,
+      "grad_norm": 1.045627474784851,
+      "learning_rate": 0.00017609936822916475,
+      "loss": 3.204631805419922,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9452449567723343,
+      "grad_norm": 1.0442372560501099,
+      "learning_rate": 0.0001758647056080248,
+      "loss": 3.2004959106445314,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9462055715658021,
+      "grad_norm": 0.7651126384735107,
+      "learning_rate": 0.00017562997772757916,
+      "loss": 3.1948537826538086,
+      "step": 4925
+    },
+    {
+      "epoch": 0.94716618635927,
+      "grad_norm": 0.790465235710144,
+      "learning_rate": 0.00017539518518007043,
+      "loss": 3.196913719177246,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9481268011527377,
+      "grad_norm": 0.9765505194664001,
+      "learning_rate": 0.00017516032855790424,
+      "loss": 3.1958955764770507,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9490874159462056,
+      "grad_norm": 1.0204505920410156,
+      "learning_rate": 0.00017492540845364798,
+      "loss": 3.197770690917969,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9500480307396734,
+      "grad_norm": 0.7875639796257019,
+      "learning_rate": 0.00017469042546002913,
+      "loss": 3.1972192764282226,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9510086455331412,
+      "grad_norm": 0.8441899418830872,
+      "learning_rate": 0.00017445538016993393,
+      "loss": 3.1970443725585938,
+      "step": 4950
+    },
+    {
+      "epoch": 0.951969260326609,
+      "grad_norm": 0.9154574275016785,
+      "learning_rate": 0.00017422027317640575,
+      "loss": 3.1967355728149416,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9529298751200769,
+      "grad_norm": 0.6774746775627136,
+      "learning_rate": 0.00017398510507264363,
+      "loss": 3.1978105545043944,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9538904899135446,
+      "grad_norm": 0.9463170170783997,
+      "learning_rate": 0.0001737498764520009,
+      "loss": 3.195131301879883,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9548511047070125,
+      "grad_norm": 1.0156409740447998,
+      "learning_rate": 0.00017351458790798338,
+      "loss": 3.197483253479004,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9558117195004803,
+      "grad_norm": 0.8678054213523865,
+      "learning_rate": 0.0001732792400342483,
+      "loss": 3.1987661361694335,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9567723342939481,
+      "grad_norm": 0.8004742860794067,
+      "learning_rate": 0.00017304383342460242,
+      "loss": 3.1965717315673827,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9577329490874159,
+      "grad_norm": 1.0178972482681274,
+      "learning_rate": 0.0001728083686730008,
+      "loss": 3.197255325317383,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9586935638808838,
+      "grad_norm": 0.7633374333381653,
+      "learning_rate": 0.00017257284637354524,
+      "loss": 3.1934268951416014,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9596541786743515,
+      "grad_norm": 0.735908031463623,
+      "learning_rate": 0.00017233726712048253,
+      "loss": 3.194230079650879,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9606147934678194,
+      "grad_norm": 0.936837911605835,
+      "learning_rate": 0.0001721016315082034,
+      "loss": 3.1968971252441407,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9615754082612872,
+      "grad_norm": 0.8966473937034607,
+      "learning_rate": 0.00017186594013124057,
+      "loss": 3.2022254943847654,
+      "step": 5005
+    },
+    {
+      "epoch": 0.962536023054755,
+      "grad_norm": 0.8859227299690247,
+      "learning_rate": 0.00017163019358426763,
+      "loss": 3.19702205657959,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9634966378482228,
+      "grad_norm": 0.6885091662406921,
+      "learning_rate": 0.00017139439246209728,
+      "loss": 3.1955480575561523,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9644572526416907,
+      "grad_norm": 0.6677486896514893,
+      "learning_rate": 0.00017115853735967995,
+      "loss": 3.1977615356445312,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9654178674351584,
+      "grad_norm": 1.0436347723007202,
+      "learning_rate": 0.00017092262887210232,
+      "loss": 3.195005416870117,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9663784822286263,
+      "grad_norm": 0.9044774174690247,
+      "learning_rate": 0.0001706866675945856,
+      "loss": 3.1969493865966796,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9673390970220941,
+      "grad_norm": 0.9574539661407471,
+      "learning_rate": 0.00017045065412248434,
+      "loss": 3.194741058349609,
+      "step": 5035
+    },
+    {
+      "epoch": 0.968299711815562,
+      "grad_norm": 0.9477970004081726,
+      "learning_rate": 0.00017021458905128477,
+      "loss": 3.1963451385498045,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9692603266090298,
+      "grad_norm": 0.8571361899375916,
+      "learning_rate": 0.00016997847297660324,
+      "loss": 3.194466972351074,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9702209414024976,
+      "grad_norm": 0.8901039361953735,
+      "learning_rate": 0.00016974230649418487,
+      "loss": 3.1929143905639648,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9711815561959655,
+      "grad_norm": 1.027086615562439,
+      "learning_rate": 0.00016950609019990187,
+      "loss": 3.191559982299805,
+      "step": 5055
+    },
+    {
+      "epoch": 0.9721421709894332,
+      "grad_norm": 0.8591023683547974,
+      "learning_rate": 0.00016926982468975225,
+      "loss": 3.199761962890625,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9731027857829011,
+      "grad_norm": 1.0826267004013062,
+      "learning_rate": 0.00016903351055985806,
+      "loss": 3.1961299896240236,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9740634005763689,
+      "grad_norm": 1.0031397342681885,
+      "learning_rate": 0.0001687971484064642,
+      "loss": 3.1959823608398437,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9750240153698367,
+      "grad_norm": 0.7752434015274048,
+      "learning_rate": 0.00016856073882593646,
+      "loss": 3.1921688079833985,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9759846301633045,
+      "grad_norm": 0.911412239074707,
+      "learning_rate": 0.00016832428241476063,
+      "loss": 3.191498947143555,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9769452449567724,
+      "grad_norm": 0.9647287726402283,
+      "learning_rate": 0.00016808777976954042,
+      "loss": 3.1929500579833983,
+      "step": 5085
+    },
+    {
+      "epoch": 0.9779058597502401,
+      "grad_norm": 0.7081168293952942,
+      "learning_rate": 0.00016785123148699624,
+      "loss": 3.1945589065551756,
+      "step": 5090
+    },
+    {
+      "epoch": 0.978866474543708,
+      "grad_norm": 0.8361444473266602,
+      "learning_rate": 0.00016761463816396374,
+      "loss": 3.1967111587524415,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9798270893371758,
+      "grad_norm": 0.8908597230911255,
+      "learning_rate": 0.00016737800039739212,
+      "loss": 3.1946334838867188,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9807877041306436,
+      "grad_norm": 0.8121721744537354,
+      "learning_rate": 0.0001671413187843427,
+      "loss": 3.195412826538086,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9817483189241114,
+      "grad_norm": 0.6848476529121399,
+      "learning_rate": 0.00016690459392198752,
+      "loss": 3.191677284240723,
+      "step": 5110
+    },
+    {
+      "epoch": 0.9827089337175793,
+      "grad_norm": 0.6763195991516113,
+      "learning_rate": 0.00016666782640760766,
+      "loss": 3.1958782196044924,
+      "step": 5115
+    },
+    {
+      "epoch": 0.983669548511047,
+      "grad_norm": 0.8343232870101929,
+      "learning_rate": 0.00016643101683859186,
+      "loss": 3.197767639160156,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9846301633045149,
+      "grad_norm": 0.9125416874885559,
+      "learning_rate": 0.00016619416581243498,
+      "loss": 3.1921886444091796,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9855907780979827,
+      "grad_norm": 0.8966071009635925,
+      "learning_rate": 0.00016595727392673643,
+      "loss": 3.1952947616577148,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9865513928914506,
+      "grad_norm": 0.6765633225440979,
+      "learning_rate": 0.00016572034177919875,
+      "loss": 3.1933788299560546,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9875120076849183,
+      "grad_norm": 0.814222514629364,
+      "learning_rate": 0.00016548336996762608,
+      "loss": 3.1904666900634764,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9884726224783862,
+      "grad_norm": 0.9140084981918335,
+      "learning_rate": 0.00016524635908992253,
+      "loss": 3.1964300155639647,
+      "step": 5145
+    },
+    {
+      "epoch": 0.989433237271854,
+      "grad_norm": 0.9220796227455139,
+      "learning_rate": 0.00016500930974409092,
+      "loss": 3.191013526916504,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9903938520653218,
+      "grad_norm": 0.7357509732246399,
+      "learning_rate": 0.00016477222252823107,
+      "loss": 3.192600059509277,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9913544668587896,
+      "grad_norm": 0.9331154227256775,
+      "learning_rate": 0.00016453509804053833,
+      "loss": 3.195221519470215,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9923150816522575,
+      "grad_norm": 0.8373104929924011,
+      "learning_rate": 0.00016429793687930215,
+      "loss": 3.195628356933594,
+      "step": 5165
+    },
+    {
+      "epoch": 0.9932756964457252,
+      "grad_norm": 0.8984804153442383,
+      "learning_rate": 0.0001640607396429044,
+      "loss": 3.192228889465332,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9942363112391931,
+      "grad_norm": 0.8978838324546814,
+      "learning_rate": 0.00016382350692981812,
+      "loss": 3.1939905166625975,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9951969260326609,
+      "grad_norm": 0.6365435719490051,
+      "learning_rate": 0.00016358623933860567,
+      "loss": 3.1925621032714844,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9961575408261287,
+      "grad_norm": 0.7753834128379822,
+      "learning_rate": 0.00016334893746791762,
+      "loss": 3.192973327636719,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9971181556195965,
+      "grad_norm": 0.9735800623893738,
+      "learning_rate": 0.0001631116019164909,
+      "loss": 3.1959808349609373,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9980787704130644,
+      "grad_norm": 0.6787250638008118,
+      "learning_rate": 0.00016287423328314746,
+      "loss": 3.193808746337891,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9990393852065321,
+      "grad_norm": 0.7312700152397156,
+      "learning_rate": 0.00016263683216679274,
+      "loss": 3.1927206039428713,
+      "step": 5200
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7704550623893738,
+      "learning_rate": 0.00016239939916641404,
+      "loss": 3.1963579177856447,
+      "step": 5205
+    },
+    {
+      "epoch": 1.0009606147934678,
+      "grad_norm": 0.7807424664497375,
+      "learning_rate": 0.00016216193488107926,
+      "loss": 3.1920791625976563,
+      "step": 5210
+    },
+    {
+      "epoch": 1.0019212295869357,
+      "grad_norm": 0.9543339014053345,
+      "learning_rate": 0.0001619244399099351,
+      "loss": 3.1910972595214844,
+      "step": 5215
+    },
+    {
+      "epoch": 1.0028818443804035,
+      "grad_norm": 0.7766204476356506,
+      "learning_rate": 0.00016168691485220573,
+      "loss": 3.194061851501465,
+      "step": 5220
+    },
+    {
+      "epoch": 1.0038424591738713,
+      "grad_norm": 0.8086463809013367,
+      "learning_rate": 0.00016144936030719126,
+      "loss": 3.1898380279541017,
+      "step": 5225
+    },
+    {
+      "epoch": 1.004803073967339,
+      "grad_norm": 1.1027806997299194,
+      "learning_rate": 0.00016121177687426617,
+      "loss": 3.193951416015625,
+      "step": 5230
+    },
+    {
+      "epoch": 1.005763688760807,
+      "grad_norm": 0.996457576751709,
+      "learning_rate": 0.00016097416515287787,
+      "loss": 3.188690757751465,
+      "step": 5235
+    },
+    {
+      "epoch": 1.0067243035542748,
+      "grad_norm": 0.7975702881813049,
+      "learning_rate": 0.00016073652574254504,
+      "loss": 3.1924358367919923,
+      "step": 5240
+    },
+    {
+      "epoch": 1.0076849183477425,
+      "grad_norm": 0.7981067299842834,
+      "learning_rate": 0.00016049885924285638,
+      "loss": 3.1881412506103515,
+      "step": 5245
+    },
+    {
+      "epoch": 1.0086455331412103,
+      "grad_norm": 0.8105031251907349,
+      "learning_rate": 0.00016026116625346876,
+      "loss": 3.192718505859375,
+      "step": 5250
+    },
+    {
+      "epoch": 1.0096061479346783,
+      "grad_norm": 0.7112568616867065,
+      "learning_rate": 0.00016002344737410602,
+      "loss": 3.187815856933594,
+      "step": 5255
+    },
+    {
+      "epoch": 1.010566762728146,
+      "grad_norm": 0.8126310706138611,
+      "learning_rate": 0.0001597857032045573,
+      "loss": 3.191788673400879,
+      "step": 5260
+    },
+    {
+      "epoch": 1.0115273775216138,
+      "grad_norm": 0.733974814414978,
+      "learning_rate": 0.00015954793434467545,
+      "loss": 3.1942237854003905,
+      "step": 5265
+    },
+    {
+      "epoch": 1.0124879923150816,
+      "grad_norm": 0.6596633791923523,
+      "learning_rate": 0.0001593101413943758,
+      "loss": 3.189275932312012,
+      "step": 5270
+    },
+    {
+      "epoch": 1.0134486071085496,
+      "grad_norm": 0.8669559955596924,
+      "learning_rate": 0.0001590723249536343,
+      "loss": 3.1921119689941406,
+      "step": 5275
+    },
+    {
+      "epoch": 1.0144092219020173,
+      "grad_norm": 1.2266993522644043,
+      "learning_rate": 0.00015883448562248624,
+      "loss": 3.191496467590332,
+      "step": 5280
+    },
+    {
+      "epoch": 1.015369836695485,
+      "grad_norm": 0.8078704476356506,
+      "learning_rate": 0.00015859662400102462,
+      "loss": 3.1911800384521483,
+      "step": 5285
+    },
+    {
+      "epoch": 1.0163304514889528,
+      "grad_norm": 0.6551411151885986,
+      "learning_rate": 0.00015835874068939878,
+      "loss": 3.188446044921875,
+      "step": 5290
+    },
+    {
+      "epoch": 1.0172910662824208,
+      "grad_norm": 0.8124401569366455,
+      "learning_rate": 0.0001581208362878126,
+      "loss": 3.1922882080078123,
+      "step": 5295
+    },
+    {
+      "epoch": 1.0182516810758886,
+      "grad_norm": 0.8051562309265137,
+      "learning_rate": 0.00015788291139652344,
+      "loss": 3.190486717224121,
+      "step": 5300
+    },
+    {
+      "epoch": 1.0192122958693564,
+      "grad_norm": 0.9743736386299133,
+      "learning_rate": 0.00015764496661584013,
+      "loss": 3.1894975662231446,
+      "step": 5305
+    },
+    {
+      "epoch": 1.0201729106628241,
+      "grad_norm": 0.7459861040115356,
+      "learning_rate": 0.00015740700254612177,
+      "loss": 3.193043518066406,
+      "step": 5310
+    },
+    {
+      "epoch": 1.021133525456292,
+      "grad_norm": 1.2092807292938232,
+      "learning_rate": 0.0001571690197877761,
+      "loss": 3.1933567047119142,
+      "step": 5315
+    },
+    {
+      "epoch": 1.0220941402497599,
+      "grad_norm": 0.7785642147064209,
+      "learning_rate": 0.00015693101894125806,
+      "loss": 3.1891590118408204,
+      "step": 5320
+    },
+    {
+      "epoch": 1.0230547550432276,
+      "grad_norm": 0.7812630534172058,
+      "learning_rate": 0.00015669300060706823,
+      "loss": 3.190948486328125,
+      "step": 5325
+    },
+    {
+      "epoch": 1.0240153698366954,
+      "grad_norm": 0.9370684027671814,
+      "learning_rate": 0.0001564549653857512,
+      "loss": 3.1883705139160154,
+      "step": 5330
+    },
+    {
+      "epoch": 1.0249759846301634,
+      "grad_norm": 0.8281055092811584,
+      "learning_rate": 0.00015621691387789432,
+      "loss": 3.1863418579101563,
+      "step": 5335
+    },
+    {
+      "epoch": 1.0259365994236311,
+      "grad_norm": 1.0316886901855469,
+      "learning_rate": 0.000155978846684126,
+      "loss": 3.19388370513916,
+      "step": 5340
+    },
+    {
+      "epoch": 1.026897214217099,
+      "grad_norm": 0.758702278137207,
+      "learning_rate": 0.00015574076440511407,
+      "loss": 3.1906862258911133,
+      "step": 5345
+    },
+    {
+      "epoch": 1.0278578290105667,
+      "grad_norm": 0.5779961347579956,
+      "learning_rate": 0.00015550266764156466,
+      "loss": 3.190965461730957,
+      "step": 5350
+    },
+    {
+      "epoch": 1.0288184438040346,
+      "grad_norm": 0.9982998371124268,
+      "learning_rate": 0.00015526455699422023,
+      "loss": 3.190749740600586,
+      "step": 5355
+    },
+    {
+      "epoch": 1.0297790585975024,
+      "grad_norm": 1.105873465538025,
+      "learning_rate": 0.00015502643306385846,
+      "loss": 3.1928503036499025,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0307396733909702,
+      "grad_norm": 0.9287274479866028,
+      "learning_rate": 0.00015478829645129046,
+      "loss": 3.1860063552856444,
+      "step": 5365
+    },
+    {
+      "epoch": 1.031700288184438,
+      "grad_norm": 0.7572025656700134,
+      "learning_rate": 0.00015455014775735923,
+      "loss": 3.1904781341552733,
+      "step": 5370
+    },
+    {
+      "epoch": 1.032660902977906,
+      "grad_norm": 0.8866919279098511,
+      "learning_rate": 0.0001543119875829385,
+      "loss": 3.1870819091796876,
+      "step": 5375
+    },
+    {
+      "epoch": 1.0336215177713737,
+      "grad_norm": 0.7839595675468445,
+      "learning_rate": 0.00015407381652893066,
+      "loss": 3.1915761947631838,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0345821325648414,
+      "grad_norm": 0.7243456244468689,
+      "learning_rate": 0.00015383563519626582,
+      "loss": 3.1893665313720705,
+      "step": 5385
+    },
+    {
+      "epoch": 1.0355427473583094,
+      "grad_norm": 1.0272892713546753,
+      "learning_rate": 0.0001535974441858999,
+      "loss": 3.187778854370117,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0365033621517772,
+      "grad_norm": 1.069861650466919,
+      "learning_rate": 0.00015335924409881323,
+      "loss": 3.1904586791992187,
+      "step": 5395
+    },
+    {
+      "epoch": 1.037463976945245,
+      "grad_norm": 0.8298363089561462,
+      "learning_rate": 0.00015312103553600913,
+      "loss": 3.1916831970214843,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0384245917387127,
+      "grad_norm": 0.7426701784133911,
+      "learning_rate": 0.00015288281909851213,
+      "loss": 3.1897172927856445,
+      "step": 5405
+    },
+    {
+      "epoch": 1.0393852065321807,
+      "grad_norm": 0.8290877938270569,
+      "learning_rate": 0.00015264459538736686,
+      "loss": 3.1861778259277345,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0403458213256485,
+      "grad_norm": 0.5183548927307129,
+      "learning_rate": 0.00015240636500363607,
+      "loss": 3.185122489929199,
+      "step": 5415
+    },
+    {
+      "epoch": 1.0413064361191162,
+      "grad_norm": 0.9525085687637329,
+      "learning_rate": 0.00015216812854839957,
+      "loss": 3.187310791015625,
+      "step": 5420
+    },
+    {
+      "epoch": 1.042267050912584,
+      "grad_norm": 0.9072858691215515,
+      "learning_rate": 0.0001519298866227523,
+      "loss": 3.1853969573974608,
+      "step": 5425
+    },
+    {
+      "epoch": 1.043227665706052,
+      "grad_norm": 0.8549609780311584,
+      "learning_rate": 0.00015169163982780304,
+      "loss": 3.1908397674560547,
+      "step": 5430
+    },
+    {
+      "epoch": 1.0441882804995197,
+      "grad_norm": 0.5929837822914124,
+      "learning_rate": 0.00015145338876467303,
+      "loss": 3.1879791259765624,
+      "step": 5435
+    },
+    {
+      "epoch": 1.0451488952929875,
+      "grad_norm": 0.8749310970306396,
+      "learning_rate": 0.00015121513403449402,
+      "loss": 3.1909906387329103,
+      "step": 5440
+    },
+    {
+      "epoch": 1.0461095100864553,
+      "grad_norm": 0.6831730604171753,
+      "learning_rate": 0.00015097687623840726,
+      "loss": 3.188145637512207,
+      "step": 5445
+    },
+    {
+      "epoch": 1.0470701248799232,
+      "grad_norm": 0.7838398814201355,
+      "learning_rate": 0.00015073861597756144,
+      "loss": 3.1877973556518553,
+      "step": 5450
+    },
+    {
+      "epoch": 1.048030739673391,
+      "grad_norm": 0.6628211140632629,
+      "learning_rate": 0.00015050035385311182,
+      "loss": 3.1896636962890623,
+      "step": 5455
+    },
+    {
+      "epoch": 1.0489913544668588,
+      "grad_norm": 0.737322211265564,
+      "learning_rate": 0.00015026209046621803,
+      "loss": 3.1869300842285155,
+      "step": 5460
+    },
+    {
+      "epoch": 1.0499519692603265,
+      "grad_norm": 1.0209139585494995,
+      "learning_rate": 0.0001500238264180431,
+      "loss": 3.187608528137207,
+      "step": 5465
+    },
+    {
+      "epoch": 1.0509125840537945,
+      "grad_norm": 0.7242445945739746,
+      "learning_rate": 0.00014978556230975173,
+      "loss": 3.188776969909668,
+      "step": 5470
+    },
+    {
+      "epoch": 1.0518731988472623,
+      "grad_norm": 0.968614399433136,
+      "learning_rate": 0.0001495472987425086,
+      "loss": 3.1897449493408203,
+      "step": 5475
+    },
+    {
+      "epoch": 1.05283381364073,
+      "grad_norm": 1.0760611295700073,
+      "learning_rate": 0.00014930903631747724,
+      "loss": 3.1843488693237303,
+      "step": 5480
+    },
+    {
+      "epoch": 1.0537944284341978,
+      "grad_norm": 0.8381465673446655,
+      "learning_rate": 0.00014907077563581809,
+      "loss": 3.189168930053711,
+      "step": 5485
+    },
+    {
+      "epoch": 1.0547550432276658,
+      "grad_norm": 0.5522369742393494,
+      "learning_rate": 0.00014883251729868737,
+      "loss": 3.1856075286865235,
+      "step": 5490
+    },
+    {
+      "epoch": 1.0557156580211335,
+      "grad_norm": 0.8688974380493164,
+      "learning_rate": 0.00014859426190723523,
+      "loss": 3.1877534866333006,
+      "step": 5495
+    },
+    {
+      "epoch": 1.0566762728146013,
+      "grad_norm": 1.0312248468399048,
+      "learning_rate": 0.00014835601006260455,
+      "loss": 3.187721824645996,
+      "step": 5500
+    },
+    {
+      "epoch": 1.057636887608069,
+      "grad_norm": 1.0165605545043945,
+      "learning_rate": 0.00014811776236592914,
+      "loss": 3.1891069412231445,
+      "step": 5505
+    },
+    {
+      "epoch": 1.058597502401537,
+      "grad_norm": 0.6776289939880371,
+      "learning_rate": 0.00014787951941833236,
+      "loss": 3.188289451599121,
+      "step": 5510
+    },
+    {
+      "epoch": 1.0595581171950048,
+      "grad_norm": 0.6294692158699036,
+      "learning_rate": 0.00014764128182092557,
+      "loss": 3.191664123535156,
+      "step": 5515
+    },
+    {
+      "epoch": 1.0605187319884726,
+      "grad_norm": 0.8317521810531616,
+      "learning_rate": 0.00014740305017480674,
+      "loss": 3.1893966674804686,
+      "step": 5520
+    },
+    {
+      "epoch": 1.0614793467819403,
+      "grad_norm": 0.9894767999649048,
+      "learning_rate": 0.0001471648250810588,
+      "loss": 3.1885601043701173,
+      "step": 5525
+    },
+    {
+      "epoch": 1.0624399615754083,
+      "grad_norm": 0.7484564185142517,
+      "learning_rate": 0.00014692660714074796,
+      "loss": 3.18869571685791,
+      "step": 5530
+    },
+    {
+      "epoch": 1.063400576368876,
+      "grad_norm": 0.8479475975036621,
+      "learning_rate": 0.00014668839695492264,
+      "loss": 3.184678649902344,
+      "step": 5535
+    },
+    {
+      "epoch": 1.0643611911623438,
+      "grad_norm": 0.6609562039375305,
+      "learning_rate": 0.00014645019512461157,
+      "loss": 3.1846710205078126,
+      "step": 5540
+    },
+    {
+      "epoch": 1.0653218059558118,
+      "grad_norm": 0.955204963684082,
+      "learning_rate": 0.0001462120022508223,
+      "loss": 3.19210205078125,
+      "step": 5545
+    },
+    {
+      "epoch": 1.0662824207492796,
+      "grad_norm": 0.7516992092132568,
+      "learning_rate": 0.00014597381893453998,
+      "loss": 3.188296318054199,
+      "step": 5550
+    },
+    {
+      "epoch": 1.0672430355427474,
+      "grad_norm": 0.7409364581108093,
+      "learning_rate": 0.0001457356457767255,
+      "loss": 3.1862789154052735,
+      "step": 5555
+    },
+    {
+      "epoch": 1.0682036503362151,
+      "grad_norm": 0.6613131165504456,
+      "learning_rate": 0.00014549748337831423,
+      "loss": 3.185582733154297,
+      "step": 5560
+    },
+    {
+      "epoch": 1.069164265129683,
+      "grad_norm": 0.6710831522941589,
+      "learning_rate": 0.00014525933234021428,
+      "loss": 3.1856903076171874,
+      "step": 5565
+    },
+    {
+      "epoch": 1.0701248799231509,
+      "grad_norm": 0.6263184547424316,
+      "learning_rate": 0.00014502119326330516,
+      "loss": 3.186991310119629,
+      "step": 5570
+    },
+    {
+      "epoch": 1.0710854947166186,
+      "grad_norm": 0.838955283164978,
+      "learning_rate": 0.0001447830667484362,
+      "loss": 3.1853305816650392,
+      "step": 5575
+    },
+    {
+      "epoch": 1.0720461095100864,
+      "grad_norm": 0.6884387135505676,
+      "learning_rate": 0.00014454495339642504,
+      "loss": 3.189616584777832,
+      "step": 5580
+    },
+    {
+      "epoch": 1.0730067243035544,
+      "grad_norm": 0.8635831475257874,
+      "learning_rate": 0.00014430685380805605,
+      "loss": 3.185613822937012,
+      "step": 5585
+    },
+    {
+      "epoch": 1.0739673390970221,
+      "grad_norm": 0.842369794845581,
+      "learning_rate": 0.00014406876858407894,
+      "loss": 3.1873950958251953,
+      "step": 5590
+    },
+    {
+      "epoch": 1.07492795389049,
+      "grad_norm": 0.6380950212478638,
+      "learning_rate": 0.00014383069832520721,
+      "loss": 3.1872686386108398,
+      "step": 5595
+    },
+    {
+      "epoch": 1.0758885686839577,
+      "grad_norm": 0.6167466044425964,
+      "learning_rate": 0.00014359264363211647,
+      "loss": 3.1889183044433596,
+      "step": 5600
+    },
+    {
+      "epoch": 1.0768491834774256,
+      "grad_norm": 0.8871692419052124,
+      "learning_rate": 0.00014335460510544318,
+      "loss": 3.190326118469238,
+      "step": 5605
+    },
+    {
+      "epoch": 1.0778097982708934,
+      "grad_norm": 0.8428598642349243,
+      "learning_rate": 0.00014311658334578296,
+      "loss": 3.190188407897949,
+      "step": 5610
+    },
+    {
+      "epoch": 1.0787704130643612,
+      "grad_norm": 0.7600328922271729,
+      "learning_rate": 0.00014287857895368906,
+      "loss": 3.1856243133544924,
+      "step": 5615
+    },
+    {
+      "epoch": 1.079731027857829,
+      "grad_norm": 0.7061536908149719,
+      "learning_rate": 0.00014264059252967108,
+      "loss": 3.1858592987060548,
+      "step": 5620
+    },
+    {
+      "epoch": 1.080691642651297,
+      "grad_norm": 0.6776458024978638,
+      "learning_rate": 0.0001424026246741931,
+      "loss": 3.188393402099609,
+      "step": 5625
+    },
+    {
+      "epoch": 1.0816522574447647,
+      "grad_norm": 0.9490513205528259,
+      "learning_rate": 0.00014216467598767248,
+      "loss": 3.1876445770263673,
+      "step": 5630
+    },
+    {
+      "epoch": 1.0826128722382324,
+      "grad_norm": 0.8668431043624878,
+      "learning_rate": 0.0001419267470704781,
+      "loss": 3.1881153106689455,
+      "step": 5635
+    },
+    {
+      "epoch": 1.0835734870317002,
+      "grad_norm": 0.7393624782562256,
+      "learning_rate": 0.00014168883852292915,
+      "loss": 3.186983108520508,
+      "step": 5640
+    },
+    {
+      "epoch": 1.0845341018251682,
+      "grad_norm": 0.7788834571838379,
+      "learning_rate": 0.00014145095094529318,
+      "loss": 3.187802314758301,
+      "step": 5645
+    },
+    {
+      "epoch": 1.085494716618636,
+      "grad_norm": 0.6699324250221252,
+      "learning_rate": 0.00014121308493778503,
+      "loss": 3.1878355026245115,
+      "step": 5650
+    },
+    {
+      "epoch": 1.0864553314121037,
+      "grad_norm": 0.7733721733093262,
+      "learning_rate": 0.00014097524110056494,
+      "loss": 3.1843936920166014,
+      "step": 5655
+    },
+    {
+      "epoch": 1.0874159462055715,
+      "grad_norm": 0.711729884147644,
+      "learning_rate": 0.0001407374200337374,
+      "loss": 3.1883363723754883,
+      "step": 5660
+    },
+    {
+      "epoch": 1.0883765609990395,
+      "grad_norm": 0.6013932824134827,
+      "learning_rate": 0.0001404996223373494,
+      "loss": 3.1905609130859376,
+      "step": 5665
+    },
+    {
+      "epoch": 1.0893371757925072,
+      "grad_norm": 0.7252037525177002,
+      "learning_rate": 0.0001402618486113888,
+      "loss": 3.1879077911376954,
+      "step": 5670
+    },
+    {
+      "epoch": 1.090297790585975,
+      "grad_norm": 1.1090692281723022,
+      "learning_rate": 0.0001400240994557832,
+      "loss": 3.1889415740966798,
+      "step": 5675
+    },
+    {
+      "epoch": 1.0912584053794427,
+      "grad_norm": 1.0110430717468262,
+      "learning_rate": 0.0001397863754703981,
+      "loss": 3.185939407348633,
+      "step": 5680
+    },
+    {
+      "epoch": 1.0922190201729107,
+      "grad_norm": 0.8803393840789795,
+      "learning_rate": 0.00013954867725503542,
+      "loss": 3.1873733520507814,
+      "step": 5685
+    },
+    {
+      "epoch": 1.0931796349663785,
+      "grad_norm": 0.6381970047950745,
+      "learning_rate": 0.00013931100540943227,
+      "loss": 3.1853010177612306,
+      "step": 5690
+    },
+    {
+      "epoch": 1.0941402497598463,
+      "grad_norm": 0.5555366277694702,
+      "learning_rate": 0.00013907336053325896,
+      "loss": 3.184972381591797,
+      "step": 5695
+    },
+    {
+      "epoch": 1.0951008645533142,
+      "grad_norm": 1.0896075963974,
+      "learning_rate": 0.00013883574322611801,
+      "loss": 3.186721992492676,
+      "step": 5700
+    },
+    {
+      "epoch": 1.096061479346782,
+      "grad_norm": 0.8639760613441467,
+      "learning_rate": 0.00013859815408754218,
+      "loss": 3.1882835388183595,
+      "step": 5705
+    },
+    {
+      "epoch": 1.0970220941402498,
+      "grad_norm": 0.7950151562690735,
+      "learning_rate": 0.00013836059371699332,
+      "loss": 3.187421417236328,
+      "step": 5710
+    },
+    {
+      "epoch": 1.0979827089337175,
+      "grad_norm": 0.6552711725234985,
+      "learning_rate": 0.00013812306271386048,
+      "loss": 3.184224319458008,
+      "step": 5715
+    },
+    {
+      "epoch": 1.0989433237271853,
+      "grad_norm": 0.7803718447685242,
+      "learning_rate": 0.00013788556167745886,
+      "loss": 3.1838443756103514,
+      "step": 5720
+    },
+    {
+      "epoch": 1.0999039385206533,
+      "grad_norm": 0.8713034987449646,
+      "learning_rate": 0.00013764809120702797,
+      "loss": 3.188005828857422,
+      "step": 5725
+    },
+    {
+      "epoch": 1.100864553314121,
+      "grad_norm": 0.638888955116272,
+      "learning_rate": 0.00013741065190173008,
+      "loss": 3.183907890319824,
+      "step": 5730
+    },
+    {
+      "epoch": 1.1018251681075888,
+      "grad_norm": 0.8170698285102844,
+      "learning_rate": 0.00013717324436064897,
+      "loss": 3.184910011291504,
+      "step": 5735
+    },
+    {
+      "epoch": 1.1027857829010568,
+      "grad_norm": 0.8812254667282104,
+      "learning_rate": 0.0001369358691827882,
+      "loss": 3.1871992111206056,
+      "step": 5740
+    },
+    {
+      "epoch": 1.1037463976945245,
+      "grad_norm": 0.6346907615661621,
+      "learning_rate": 0.0001366985269670697,
+      "loss": 3.1861663818359376,
+      "step": 5745
+    },
+    {
+      "epoch": 1.1047070124879923,
+      "grad_norm": 0.930957019329071,
+      "learning_rate": 0.00013646121831233226,
+      "loss": 3.1821567535400392,
+      "step": 5750
+    },
+    {
+      "epoch": 1.10566762728146,
+      "grad_norm": 0.7975543141365051,
+      "learning_rate": 0.00013622394381732986,
+      "loss": 3.1874359130859373,
+      "step": 5755
+    },
+    {
+      "epoch": 1.106628242074928,
+      "grad_norm": 0.6229592561721802,
+      "learning_rate": 0.0001359867040807305,
+      "loss": 3.1849109649658205,
+      "step": 5760
+    },
+    {
+      "epoch": 1.1075888568683958,
+      "grad_norm": 0.7393524050712585,
+      "learning_rate": 0.0001357494997011143,
+      "loss": 3.1865949630737305,
+      "step": 5765
+    },
+    {
+      "epoch": 1.1085494716618636,
+      "grad_norm": 0.7825970649719238,
+      "learning_rate": 0.0001355123312769723,
+      "loss": 3.179779815673828,
+      "step": 5770
+    },
+    {
+      "epoch": 1.1095100864553313,
+      "grad_norm": 1.018021821975708,
+      "learning_rate": 0.00013527519940670474,
+      "loss": 3.1880813598632813,
+      "step": 5775
+    },
+    {
+      "epoch": 1.1104707012487993,
+      "grad_norm": 0.7721784114837646,
+      "learning_rate": 0.00013503810468861967,
+      "loss": 3.1860368728637694,
+      "step": 5780
+    },
+    {
+      "epoch": 1.111431316042267,
+      "grad_norm": 0.7420386075973511,
+      "learning_rate": 0.00013480104772093136,
+      "loss": 3.181845855712891,
+      "step": 5785
+    },
+    {
+      "epoch": 1.1123919308357348,
+      "grad_norm": 0.7086904644966125,
+      "learning_rate": 0.0001345640291017588,
+      "loss": 3.181869125366211,
+      "step": 5790
+    },
+    {
+      "epoch": 1.1133525456292026,
+      "grad_norm": 0.5291205048561096,
+      "learning_rate": 0.00013432704942912445,
+      "loss": 3.1842918395996094,
+      "step": 5795
+    },
+    {
+      "epoch": 1.1143131604226706,
+      "grad_norm": 0.6153315901756287,
+      "learning_rate": 0.0001340901093009522,
+      "loss": 3.180573654174805,
+      "step": 5800
+    },
+    {
+      "epoch": 1.1152737752161384,
+      "grad_norm": 0.8189062476158142,
+      "learning_rate": 0.00013385320931506636,
+      "loss": 3.181578826904297,
+      "step": 5805
+    },
+    {
+      "epoch": 1.1162343900096061,
+      "grad_norm": 0.8425748348236084,
+      "learning_rate": 0.00013361635006918987,
+      "loss": 3.1840728759765624,
+      "step": 5810
+    },
+    {
+      "epoch": 1.1171950048030739,
+      "grad_norm": 0.6964288353919983,
+      "learning_rate": 0.00013337953216094293,
+      "loss": 3.183258056640625,
+      "step": 5815
+    },
+    {
+      "epoch": 1.1181556195965419,
+      "grad_norm": 0.8112155795097351,
+      "learning_rate": 0.0001331427561878414,
+      "loss": 3.1822595596313477,
+      "step": 5820
+    },
+    {
+      "epoch": 1.1191162343900096,
+      "grad_norm": 0.5789711475372314,
+      "learning_rate": 0.0001329060227472953,
+      "loss": 3.1870792388916014,
+      "step": 5825
+    },
+    {
+      "epoch": 1.1200768491834774,
+      "grad_norm": 0.7857502102851868,
+      "learning_rate": 0.00013266933243660748,
+      "loss": 3.1816539764404297,
+      "step": 5830
+    },
+    {
+      "epoch": 1.1210374639769451,
+      "grad_norm": 0.8584936857223511,
+      "learning_rate": 0.0001324326858529718,
+      "loss": 3.1840972900390625,
+      "step": 5835
+    },
+    {
+      "epoch": 1.1219980787704131,
+      "grad_norm": 0.7861515879631042,
+      "learning_rate": 0.00013219608359347194,
+      "loss": 3.1865585327148436,
+      "step": 5840
+    },
+    {
+      "epoch": 1.122958693563881,
+      "grad_norm": 0.6675422787666321,
+      "learning_rate": 0.0001319595262550796,
+      "loss": 3.1822277069091798,
+      "step": 5845
+    },
+    {
+      "epoch": 1.1239193083573487,
+      "grad_norm": 0.7415298223495483,
+      "learning_rate": 0.00013172301443465327,
+      "loss": 3.18204345703125,
+      "step": 5850
+    },
+    {
+      "epoch": 1.1248799231508164,
+      "grad_norm": 0.7668789029121399,
+      "learning_rate": 0.00013148654872893647,
+      "loss": 3.1823986053466795,
+      "step": 5855
+    },
+    {
+      "epoch": 1.1258405379442844,
+      "grad_norm": 1.0284775495529175,
+      "learning_rate": 0.00013125012973455645,
+      "loss": 3.186284637451172,
+      "step": 5860
+    },
+    {
+      "epoch": 1.1268011527377522,
+      "grad_norm": 0.9501076936721802,
+      "learning_rate": 0.00013101375804802268,
+      "loss": 3.180257034301758,
+      "step": 5865
+    },
+    {
+      "epoch": 1.12776176753122,
+      "grad_norm": 0.6620021462440491,
+      "learning_rate": 0.00013077743426572508,
+      "loss": 3.1850921630859377,
+      "step": 5870
+    },
+    {
+      "epoch": 1.1287223823246877,
+      "grad_norm": 0.6377175450325012,
+      "learning_rate": 0.0001305411589839328,
+      "loss": 3.1847787857055665,
+      "step": 5875
+    },
+    {
+      "epoch": 1.1296829971181557,
+      "grad_norm": 0.8956018686294556,
+      "learning_rate": 0.00013030493279879267,
+      "loss": 3.1820547103881838,
+      "step": 5880
+    },
+    {
+      "epoch": 1.1306436119116234,
+      "grad_norm": 0.9648638367652893,
+      "learning_rate": 0.00013006875630632757,
+      "loss": 3.1855083465576173,
+      "step": 5885
+    },
+    {
+      "epoch": 1.1316042267050912,
+      "grad_norm": 0.7749320864677429,
+      "learning_rate": 0.000129832630102435,
+      "loss": 3.18371639251709,
+      "step": 5890
+    },
+    {
+      "epoch": 1.1325648414985592,
+      "grad_norm": 0.7162895202636719,
+      "learning_rate": 0.00012959655478288556,
+      "loss": 3.1814332962036134,
+      "step": 5895
+    },
+    {
+      "epoch": 1.133525456292027,
+      "grad_norm": 0.7946479320526123,
+      "learning_rate": 0.00012936053094332158,
+      "loss": 3.183310699462891,
+      "step": 5900
+    },
+    {
+      "epoch": 1.1344860710854947,
+      "grad_norm": 0.9230923652648926,
+      "learning_rate": 0.00012912455917925535,
+      "loss": 3.1882291793823243,
+      "step": 5905
+    },
+    {
+      "epoch": 1.1354466858789625,
+      "grad_norm": 0.6934447288513184,
+      "learning_rate": 0.0001288886400860679,
+      "loss": 3.1866472244262694,
+      "step": 5910
+    },
+    {
+      "epoch": 1.1364073006724302,
+      "grad_norm": 0.5946722626686096,
+      "learning_rate": 0.00012865277425900724,
+      "loss": 3.1836269378662108,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1373679154658982,
+      "grad_norm": 0.6142674088478088,
+      "learning_rate": 0.00012841696229318712,
+      "loss": 3.1850711822509767,
+      "step": 5920
+    },
+    {
+      "epoch": 1.138328530259366,
+      "grad_norm": 0.5882613062858582,
+      "learning_rate": 0.00012818120478358522,
+      "loss": 3.1808708190917967,
+      "step": 5925
+    },
+    {
+      "epoch": 1.1392891450528337,
+      "grad_norm": 1.1265987157821655,
+      "learning_rate": 0.000127945502325042,
+      "loss": 3.1862953186035154,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1402497598463017,
+      "grad_norm": 0.8293631076812744,
+      "learning_rate": 0.00012770985551225898,
+      "loss": 3.1817962646484377,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1412103746397695,
+      "grad_norm": 0.6100813150405884,
+      "learning_rate": 0.00012747426493979713,
+      "loss": 3.182262420654297,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1421709894332372,
+      "grad_norm": 0.6388895511627197,
+      "learning_rate": 0.00012723873120207575,
+      "loss": 3.183791732788086,
+      "step": 5945
+    },
+    {
+      "epoch": 1.143131604226705,
+      "grad_norm": 1.1153887510299683,
+      "learning_rate": 0.00012700325489337056,
+      "loss": 3.1867645263671873,
+      "step": 5950
+    },
+    {
+      "epoch": 1.144092219020173,
+      "grad_norm": 0.6275951862335205,
+      "learning_rate": 0.00012676783660781245,
+      "loss": 3.183899688720703,
+      "step": 5955
+    },
+    {
+      "epoch": 1.1450528338136408,
+      "grad_norm": 0.6649777293205261,
+      "learning_rate": 0.00012653247693938588,
+      "loss": 3.1856327056884766,
+      "step": 5960
+    },
+    {
+      "epoch": 1.1460134486071085,
+      "grad_norm": 0.7300867438316345,
+      "learning_rate": 0.00012629717648192747,
+      "loss": 3.1813177108764648,
+      "step": 5965
+    },
+    {
+      "epoch": 1.1469740634005763,
+      "grad_norm": 0.8946056365966797,
+      "learning_rate": 0.00012606193582912446,
+      "loss": 3.182071304321289,
+      "step": 5970
+    },
+    {
+      "epoch": 1.1479346781940443,
+      "grad_norm": 0.8307191729545593,
+      "learning_rate": 0.00012582675557451306,
+      "loss": 3.1808849334716798,
+      "step": 5975
+    },
+    {
+      "epoch": 1.148895292987512,
+      "grad_norm": 0.8783102631568909,
+      "learning_rate": 0.00012559163631147723,
+      "loss": 3.184329032897949,
+      "step": 5980
+    },
+    {
+      "epoch": 1.1498559077809798,
+      "grad_norm": 0.7081605792045593,
+      "learning_rate": 0.00012535657863324695,
+      "loss": 3.1825115203857424,
+      "step": 5985
+    },
+    {
+      "epoch": 1.1508165225744476,
+      "grad_norm": 0.5934545993804932,
+      "learning_rate": 0.0001251215831328969,
+      "loss": 3.1839290618896485,
+      "step": 5990
+    },
+    {
+      "epoch": 1.1517771373679155,
+      "grad_norm": 0.8041390180587769,
+      "learning_rate": 0.00012488665040334479,
+      "loss": 3.183102607727051,
+      "step": 5995
+    },
+    {
+      "epoch": 1.1527377521613833,
+      "grad_norm": 0.6526947617530823,
+      "learning_rate": 0.00012465178103735,
+      "loss": 3.1817670822143556,
+      "step": 6000
+    },
+    {
+      "epoch": 1.153698366954851,
+      "grad_norm": 0.6892653703689575,
+      "learning_rate": 0.00012441697562751215,
+      "loss": 3.180023193359375,
+      "step": 6005
+    },
+    {
+      "epoch": 1.154658981748319,
+      "grad_norm": 0.6279175877571106,
+      "learning_rate": 0.00012418223476626925,
+      "loss": 3.180769157409668,
+      "step": 6010
+    },
+    {
+      "epoch": 1.1556195965417868,
+      "grad_norm": 0.8991310596466064,
+      "learning_rate": 0.00012394755904589664,
+      "loss": 3.185033416748047,
+      "step": 6015
+    },
+    {
+      "epoch": 1.1565802113352546,
+      "grad_norm": 0.6720035672187805,
+      "learning_rate": 0.00012371294905850518,
+      "loss": 3.1769649505615236,
+      "step": 6020
+    },
+    {
+      "epoch": 1.1575408261287223,
+      "grad_norm": 0.8395649790763855,
+      "learning_rate": 0.00012347840539603998,
+      "loss": 3.1827545166015625,
+      "step": 6025
+    },
+    {
+      "epoch": 1.15850144092219,
+      "grad_norm": 0.8189507722854614,
+      "learning_rate": 0.00012324392865027873,
+      "loss": 3.1804759979248045,
+      "step": 6030
+    },
+    {
+      "epoch": 1.159462055715658,
+      "grad_norm": 0.9116944074630737,
+      "learning_rate": 0.00012300951941283036,
+      "loss": 3.185091018676758,
+      "step": 6035
+    },
+    {
+      "epoch": 1.1604226705091258,
+      "grad_norm": 0.6200742125511169,
+      "learning_rate": 0.00012277517827513341,
+      "loss": 3.1775371551513674,
+      "step": 6040
+    },
+    {
+      "epoch": 1.1613832853025936,
+      "grad_norm": 0.6865299344062805,
+      "learning_rate": 0.00012254090582845457,
+      "loss": 3.1813201904296875,
+      "step": 6045
+    },
+    {
+      "epoch": 1.1623439000960616,
+      "grad_norm": 0.7450284957885742,
+      "learning_rate": 0.00012230670266388728,
+      "loss": 3.18134765625,
+      "step": 6050
+    },
+    {
+      "epoch": 1.1633045148895294,
+      "grad_norm": 0.814132571220398,
+      "learning_rate": 0.00012207256937235014,
+      "loss": 3.182879638671875,
+      "step": 6055
+    },
+    {
+      "epoch": 1.1642651296829971,
+      "grad_norm": 0.8822999000549316,
+      "learning_rate": 0.00012183850654458553,
+      "loss": 3.1848114013671873,
+      "step": 6060
+    },
+    {
+      "epoch": 1.1652257444764649,
+      "grad_norm": 0.5908654928207397,
+      "learning_rate": 0.00012160451477115783,
+      "loss": 3.1844045639038088,
+      "step": 6065
+    },
+    {
+      "epoch": 1.1661863592699326,
+      "grad_norm": 0.5196489095687866,
+      "learning_rate": 0.00012137059464245242,
+      "loss": 3.181051254272461,
+      "step": 6070
+    },
+    {
+      "epoch": 1.1671469740634006,
+      "grad_norm": 0.7493326663970947,
+      "learning_rate": 0.00012113674674867379,
+      "loss": 3.180977630615234,
+      "step": 6075
+    },
+    {
+      "epoch": 1.1681075888568684,
+      "grad_norm": 0.8588430881500244,
+      "learning_rate": 0.00012090297167984409,
+      "loss": 3.183562088012695,
+      "step": 6080
+    },
+    {
+      "epoch": 1.1690682036503361,
+      "grad_norm": 0.707949697971344,
+      "learning_rate": 0.00012066927002580185,
+      "loss": 3.186185836791992,
+      "step": 6085
+    },
+    {
+      "epoch": 1.1700288184438041,
+      "grad_norm": 0.7008398771286011,
+      "learning_rate": 0.00012043564237620028,
+      "loss": 3.183705520629883,
+      "step": 6090
+    },
+    {
+      "epoch": 1.170989433237272,
+      "grad_norm": 0.5148990750312805,
+      "learning_rate": 0.00012020208932050595,
+      "loss": 3.180450439453125,
+      "step": 6095
+    },
+    {
+      "epoch": 1.1719500480307397,
+      "grad_norm": 1.0000126361846924,
+      "learning_rate": 0.00011996861144799712,
+      "loss": 3.1828609466552735,
+      "step": 6100
+    },
+    {
+      "epoch": 1.1729106628242074,
+      "grad_norm": 0.9941655993461609,
+      "learning_rate": 0.0001197352093477625,
+      "loss": 3.1851844787597656,
+      "step": 6105
+    },
+    {
+      "epoch": 1.1738712776176754,
+      "grad_norm": 0.6262122988700867,
+      "learning_rate": 0.00011950188360869947,
+      "loss": 3.186477279663086,
+      "step": 6110
+    },
+    {
+      "epoch": 1.1748318924111432,
+      "grad_norm": 0.7170460224151611,
+      "learning_rate": 0.00011926863481951279,
+      "loss": 3.181361770629883,
+      "step": 6115
+    },
+    {
+      "epoch": 1.175792507204611,
+      "grad_norm": 0.5651366114616394,
+      "learning_rate": 0.00011903546356871315,
+      "loss": 3.180009460449219,
+      "step": 6120
+    },
+    {
+      "epoch": 1.1767531219980787,
+      "grad_norm": 0.7348089814186096,
+      "learning_rate": 0.00011880237044461546,
+      "loss": 3.1843181610107423,
+      "step": 6125
+    },
+    {
+      "epoch": 1.1777137367915467,
+      "grad_norm": 0.8490554690361023,
+      "learning_rate": 0.00011856935603533759,
+      "loss": 3.1855024337768554,
+      "step": 6130
+    },
+    {
+      "epoch": 1.1786743515850144,
+      "grad_norm": 0.688126802444458,
+      "learning_rate": 0.0001183364209287989,
+      "loss": 3.1825450897216796,
+      "step": 6135
+    },
+    {
+      "epoch": 1.1796349663784822,
+      "grad_norm": 0.6043757796287537,
+      "learning_rate": 0.00011810356571271847,
+      "loss": 3.1808750152587892,
+      "step": 6140
+    },
+    {
+      "epoch": 1.18059558117195,
+      "grad_norm": 0.6653470396995544,
+      "learning_rate": 0.00011787079097461394,
+      "loss": 3.185806655883789,
+      "step": 6145
+    },
+    {
+      "epoch": 1.181556195965418,
+      "grad_norm": 0.5727336406707764,
+      "learning_rate": 0.0001176380973017998,
+      "loss": 3.182196617126465,
+      "step": 6150
+    },
+    {
+      "epoch": 1.1825168107588857,
+      "grad_norm": 0.4174487292766571,
+      "learning_rate": 0.00011740548528138613,
+      "loss": 3.1803747177124024,
+      "step": 6155
+    },
+    {
+      "epoch": 1.1834774255523535,
+      "grad_norm": 0.8878217339515686,
+      "learning_rate": 0.00011717295550027685,
+      "loss": 3.1840145111083986,
+      "step": 6160
+    },
+    {
+      "epoch": 1.1844380403458212,
+      "grad_norm": 0.8910025954246521,
+      "learning_rate": 0.00011694050854516853,
+      "loss": 3.181249237060547,
+      "step": 6165
+    },
+    {
+      "epoch": 1.1853986551392892,
+      "grad_norm": 0.7205166220664978,
+      "learning_rate": 0.00011670814500254858,
+      "loss": 3.1778148651123046,
+      "step": 6170
+    },
+    {
+      "epoch": 1.186359269932757,
+      "grad_norm": 0.6521825790405273,
+      "learning_rate": 0.00011647586545869418,
+      "loss": 3.17816276550293,
+      "step": 6175
+    },
+    {
+      "epoch": 1.1873198847262247,
+      "grad_norm": 0.6897614002227783,
+      "learning_rate": 0.00011624367049967037,
+      "loss": 3.1780323028564452,
+      "step": 6180
+    },
+    {
+      "epoch": 1.1882804995196925,
+      "grad_norm": 0.5899947881698608,
+      "learning_rate": 0.00011601156071132883,
+      "loss": 3.181020164489746,
+      "step": 6185
+    },
+    {
+      "epoch": 1.1892411143131605,
+      "grad_norm": 0.6569263935089111,
+      "learning_rate": 0.00011577953667930641,
+      "loss": 3.1823570251464846,
+      "step": 6190
+    },
+    {
+      "epoch": 1.1902017291066282,
+      "grad_norm": 0.6824354529380798,
+      "learning_rate": 0.00011554759898902355,
+      "loss": 3.1782264709472656,
+      "step": 6195
+    },
+    {
+      "epoch": 1.191162343900096,
+      "grad_norm": 0.5225342512130737,
+      "learning_rate": 0.00011531574822568278,
+      "loss": 3.1797101974487303,
+      "step": 6200
+    },
+    {
+      "epoch": 1.192122958693564,
+      "grad_norm": 0.5605508685112,
+      "learning_rate": 0.00011508398497426745,
+      "loss": 3.1792762756347654,
+      "step": 6205
+    },
+    {
+      "epoch": 1.1930835734870318,
+      "grad_norm": 0.8098820447921753,
+      "learning_rate": 0.00011485230981953997,
+      "loss": 3.183140182495117,
+      "step": 6210
+    },
+    {
+      "epoch": 1.1940441882804995,
+      "grad_norm": 0.6241822838783264,
+      "learning_rate": 0.00011462072334604053,
+      "loss": 3.1846290588378907,
+      "step": 6215
+    },
+    {
+      "epoch": 1.1950048030739673,
+      "grad_norm": 0.6163104772567749,
+      "learning_rate": 0.00011438922613808554,
+      "loss": 3.184892272949219,
+      "step": 6220
+    },
+    {
+      "epoch": 1.195965417867435,
+      "grad_norm": 0.6423901915550232,
+      "learning_rate": 0.00011415781877976628,
+      "loss": 3.180209922790527,
+      "step": 6225
+    },
+    {
+      "epoch": 1.196926032660903,
+      "grad_norm": 0.743615448474884,
+      "learning_rate": 0.00011392650185494712,
+      "loss": 3.181998634338379,
+      "step": 6230
+    },
+    {
+      "epoch": 1.1978866474543708,
+      "grad_norm": 0.7482926845550537,
+      "learning_rate": 0.00011369527594726452,
+      "loss": 3.178305244445801,
+      "step": 6235
+    },
+    {
+      "epoch": 1.1988472622478386,
+      "grad_norm": 0.6290075778961182,
+      "learning_rate": 0.0001134641416401251,
+      "loss": 3.182933807373047,
+      "step": 6240
+    },
+    {
+      "epoch": 1.1998078770413065,
+      "grad_norm": 0.5498515963554382,
+      "learning_rate": 0.00011323309951670444,
+      "loss": 3.1798477172851562,
+      "step": 6245
+    },
+    {
+      "epoch": 1.2007684918347743,
+      "grad_norm": 0.6985566020011902,
+      "learning_rate": 0.00011300215015994554,
+      "loss": 3.1814619064331056,
+      "step": 6250
+    },
+    {
+      "epoch": 1.201729106628242,
+      "grad_norm": 0.8303685784339905,
+      "learning_rate": 0.00011277129415255727,
+      "loss": 3.1828535079956053,
+      "step": 6255
+    },
+    {
+      "epoch": 1.2026897214217098,
+      "grad_norm": 0.7280902862548828,
+      "learning_rate": 0.00011254053207701308,
+      "loss": 3.17972354888916,
+      "step": 6260
+    },
+    {
+      "epoch": 1.2036503362151778,
+      "grad_norm": 0.6531400084495544,
+      "learning_rate": 0.00011230986451554932,
+      "loss": 3.1785831451416016,
+      "step": 6265
+    },
+    {
+      "epoch": 1.2046109510086456,
+      "grad_norm": 0.6152121424674988,
+      "learning_rate": 0.00011207929205016386,
+      "loss": 3.177712631225586,
+      "step": 6270
+    },
+    {
+      "epoch": 1.2055715658021133,
+      "grad_norm": 0.5569039583206177,
+      "learning_rate": 0.00011184881526261483,
+      "loss": 3.1826234817504884,
+      "step": 6275
+    },
+    {
+      "epoch": 1.206532180595581,
+      "grad_norm": 1.1272635459899902,
+      "learning_rate": 0.00011161843473441867,
+      "loss": 3.179414749145508,
+      "step": 6280
+    },
+    {
+      "epoch": 1.207492795389049,
+      "grad_norm": 0.6206833720207214,
+      "learning_rate": 0.0001113881510468492,
+      "loss": 3.184890365600586,
+      "step": 6285
+    },
+    {
+      "epoch": 1.2084534101825168,
+      "grad_norm": 0.6447250247001648,
+      "learning_rate": 0.00011115796478093569,
+      "loss": 3.1766143798828126,
+      "step": 6290
+    },
+    {
+      "epoch": 1.2094140249759846,
+      "grad_norm": 0.7120223641395569,
+      "learning_rate": 0.00011092787651746178,
+      "loss": 3.1796751022338867,
+      "step": 6295
+    },
+    {
+      "epoch": 1.2103746397694524,
+      "grad_norm": 0.620508074760437,
+      "learning_rate": 0.00011069788683696366,
+      "loss": 3.1746284484863283,
+      "step": 6300
+    },
+    {
+      "epoch": 1.2113352545629203,
+      "grad_norm": 0.6917296051979065,
+      "learning_rate": 0.00011046799631972903,
+      "loss": 3.1808353424072267,
+      "step": 6305
+    },
+    {
+      "epoch": 1.2122958693563881,
+      "grad_norm": 0.7654911875724792,
+      "learning_rate": 0.00011023820554579512,
+      "loss": 3.1830776214599608,
+      "step": 6310
+    },
+    {
+      "epoch": 1.2132564841498559,
+      "grad_norm": 0.7507694959640503,
+      "learning_rate": 0.00011000851509494771,
+      "loss": 3.181525230407715,
+      "step": 6315
+    },
+    {
+      "epoch": 1.2142170989433236,
+      "grad_norm": 0.6086501479148865,
+      "learning_rate": 0.00010977892554671935,
+      "loss": 3.1778955459594727,
+      "step": 6320
+    },
+    {
+      "epoch": 1.2151777137367916,
+      "grad_norm": 0.5555998682975769,
+      "learning_rate": 0.00010954943748038798,
+      "loss": 3.1771930694580077,
+      "step": 6325
+    },
+    {
+      "epoch": 1.2161383285302594,
+      "grad_norm": 0.6753772497177124,
+      "learning_rate": 0.00010932005147497559,
+      "loss": 3.178631591796875,
+      "step": 6330
+    },
+    {
+      "epoch": 1.2170989433237271,
+      "grad_norm": 0.5674150586128235,
+      "learning_rate": 0.00010909076810924654,
+      "loss": 3.183444786071777,
+      "step": 6335
+    },
+    {
+      "epoch": 1.218059558117195,
+      "grad_norm": 0.8342090249061584,
+      "learning_rate": 0.00010886158796170633,
+      "loss": 3.181511116027832,
+      "step": 6340
+    },
+    {
+      "epoch": 1.219020172910663,
+      "grad_norm": 0.9122212529182434,
+      "learning_rate": 0.00010863251161060002,
+      "loss": 3.1852695465087892,
+      "step": 6345
+    },
+    {
+      "epoch": 1.2199807877041307,
+      "grad_norm": 0.44202694296836853,
+      "learning_rate": 0.00010840353963391065,
+      "loss": 3.1827775955200197,
+      "step": 6350
+    },
+    {
+      "epoch": 1.2209414024975984,
+      "grad_norm": 0.5556421279907227,
+      "learning_rate": 0.0001081746726093581,
+      "loss": 3.176057815551758,
+      "step": 6355
+    },
+    {
+      "epoch": 1.2219020172910664,
+      "grad_norm": 0.5319374203681946,
+      "learning_rate": 0.00010794591111439729,
+      "loss": 3.177410125732422,
+      "step": 6360
+    },
+    {
+      "epoch": 1.2228626320845342,
+      "grad_norm": 0.7626725435256958,
+      "learning_rate": 0.000107717255726217,
+      "loss": 3.182592582702637,
+      "step": 6365
+    },
+    {
+      "epoch": 1.223823246878002,
+      "grad_norm": 0.7391705513000488,
+      "learning_rate": 0.00010748870702173815,
+      "loss": 3.178932952880859,
+      "step": 6370
+    },
+    {
+      "epoch": 1.2247838616714697,
+      "grad_norm": 0.4964456260204315,
+      "learning_rate": 0.00010726026557761264,
+      "loss": 3.1790546417236327,
+      "step": 6375
+    },
+    {
+      "epoch": 1.2257444764649374,
+      "grad_norm": 0.5090710520744324,
+      "learning_rate": 0.00010703193197022159,
+      "loss": 3.1741233825683595,
+      "step": 6380
+    },
+    {
+      "epoch": 1.2267050912584054,
+      "grad_norm": 0.8102352023124695,
+      "learning_rate": 0.00010680370677567423,
+      "loss": 3.184102249145508,
+      "step": 6385
+    },
+    {
+      "epoch": 1.2276657060518732,
+      "grad_norm": 0.7811996340751648,
+      "learning_rate": 0.00010657559056980603,
+      "loss": 3.1772319793701174,
+      "step": 6390
+    },
+    {
+      "epoch": 1.228626320845341,
+      "grad_norm": 0.7701186537742615,
+      "learning_rate": 0.00010634758392817762,
+      "loss": 3.1799732208251954,
+      "step": 6395
+    },
+    {
+      "epoch": 1.229586935638809,
+      "grad_norm": 0.685007631778717,
+      "learning_rate": 0.00010611968742607318,
+      "loss": 3.176517105102539,
+      "step": 6400
+    },
+    {
+      "epoch": 1.2305475504322767,
+      "grad_norm": 0.5316605567932129,
+      "learning_rate": 0.00010589190163849885,
+      "loss": 3.1800348281860353,
+      "step": 6405
+    },
+    {
+      "epoch": 1.2315081652257445,
+      "grad_norm": 0.6539739370346069,
+      "learning_rate": 0.00010566422714018167,
+      "loss": 3.1817600250244142,
+      "step": 6410
+    },
+    {
+      "epoch": 1.2324687800192122,
+      "grad_norm": 0.5445916056632996,
+      "learning_rate": 0.00010543666450556774,
+      "loss": 3.1751848220825196,
+      "step": 6415
+    },
+    {
+      "epoch": 1.23342939481268,
+      "grad_norm": 0.6994644403457642,
+      "learning_rate": 0.00010520921430882088,
+      "loss": 3.182989311218262,
+      "step": 6420
+    },
+    {
+      "epoch": 1.234390009606148,
+      "grad_norm": 0.6963391304016113,
+      "learning_rate": 0.00010498187712382132,
+      "loss": 3.1813974380493164,
+      "step": 6425
+    },
+    {
+      "epoch": 1.2353506243996157,
+      "grad_norm": 0.8707190752029419,
+      "learning_rate": 0.00010475465352416402,
+      "loss": 3.1847021102905275,
+      "step": 6430
+    },
+    {
+      "epoch": 1.2363112391930835,
+      "grad_norm": 0.628430962562561,
+      "learning_rate": 0.00010452754408315754,
+      "loss": 3.180718994140625,
+      "step": 6435
+    },
+    {
+      "epoch": 1.2372718539865515,
+      "grad_norm": 0.4727374017238617,
+      "learning_rate": 0.00010430054937382219,
+      "loss": 3.179873466491699,
+      "step": 6440
+    },
+    {
+      "epoch": 1.2382324687800192,
+      "grad_norm": 0.6456632614135742,
+      "learning_rate": 0.000104073669968889,
+      "loss": 3.1764720916748046,
+      "step": 6445
+    },
+    {
+      "epoch": 1.239193083573487,
+      "grad_norm": 0.9211506843566895,
+      "learning_rate": 0.0001038469064407979,
+      "loss": 3.175563430786133,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2401536983669548,
+      "grad_norm": 0.7014700770378113,
+      "learning_rate": 0.00010362025936169664,
+      "loss": 3.1777694702148436,
+      "step": 6455
+    },
+    {
+      "epoch": 1.2411143131604228,
+      "grad_norm": 0.612933874130249,
+      "learning_rate": 0.00010339372930343896,
+      "loss": 3.177769088745117,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2420749279538905,
+      "grad_norm": 0.47281619906425476,
+      "learning_rate": 0.00010316731683758346,
+      "loss": 3.177565574645996,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2430355427473583,
+      "grad_norm": 0.6775256395339966,
+      "learning_rate": 0.00010294102253539204,
+      "loss": 3.1776294708251953,
+      "step": 6470
+    },
+    {
+      "epoch": 1.243996157540826,
+      "grad_norm": 0.7892163395881653,
+      "learning_rate": 0.00010271484696782837,
+      "loss": 3.175709342956543,
+      "step": 6475
+    },
+    {
+      "epoch": 1.244956772334294,
+      "grad_norm": 0.6570128202438354,
+      "learning_rate": 0.00010248879070555668,
+      "loss": 3.1776430130004885,
+      "step": 6480
+    },
+    {
+      "epoch": 1.2459173871277618,
+      "grad_norm": 0.5200539231300354,
+      "learning_rate": 0.00010226285431894013,
+      "loss": 3.176724433898926,
+      "step": 6485
+    },
+    {
+      "epoch": 1.2468780019212296,
+      "grad_norm": 0.7223069667816162,
+      "learning_rate": 0.00010203703837803934,
+      "loss": 3.176532745361328,
+      "step": 6490
+    },
+    {
+      "epoch": 1.2478386167146973,
+      "grad_norm": 0.8052407503128052,
+      "learning_rate": 0.00010181134345261115,
+      "loss": 3.176080322265625,
+      "step": 6495
+    },
+    {
+      "epoch": 1.2487992315081653,
+      "grad_norm": 0.48731186985969543,
+      "learning_rate": 0.00010158577011210695,
+      "loss": 3.173239517211914,
+      "step": 6500
+    },
+    {
+      "epoch": 1.249759846301633,
+      "grad_norm": 0.48055845499038696,
+      "learning_rate": 0.00010136031892567145,
+      "loss": 3.1757953643798826,
+      "step": 6505
+    },
+    {
+      "epoch": 1.2507204610951008,
+      "grad_norm": 0.9334460496902466,
+      "learning_rate": 0.00010113499046214113,
+      "loss": 3.173488998413086,
+      "step": 6510
+    },
+    {
+      "epoch": 1.2516810758885688,
+      "grad_norm": 0.747382402420044,
+      "learning_rate": 0.00010090978529004284,
+      "loss": 3.1791568756103517,
+      "step": 6515
+    },
+    {
+      "epoch": 1.2526416906820366,
+      "grad_norm": 0.5684623718261719,
+      "learning_rate": 0.00010068470397759226,
+      "loss": 3.1786075592041017,
+      "step": 6520
+    },
+    {
+      "epoch": 1.2536023054755043,
+      "grad_norm": 0.7101691365242004,
+      "learning_rate": 0.0001004597470926927,
+      "loss": 3.17834587097168,
+      "step": 6525
+    },
+    {
+      "epoch": 1.254562920268972,
+      "grad_norm": 0.6113804578781128,
+      "learning_rate": 0.0001002349152029334,
+      "loss": 3.1760730743408203,
+      "step": 6530
+    },
+    {
+      "epoch": 1.2555235350624399,
+      "grad_norm": 0.634519100189209,
+      "learning_rate": 0.00010001020887558837,
+      "loss": 3.1759071350097656,
+      "step": 6535
+    },
+    {
+      "epoch": 1.2564841498559078,
+      "grad_norm": 0.6819270253181458,
+      "learning_rate": 9.978562867761465e-05,
+      "loss": 3.1800254821777343,
+      "step": 6540
+    },
+    {
+      "epoch": 1.2574447646493756,
+      "grad_norm": 0.6920965313911438,
+      "learning_rate": 9.95611751756511e-05,
+      "loss": 3.179019546508789,
+      "step": 6545
+    },
+    {
+      "epoch": 1.2584053794428434,
+      "grad_norm": 0.7164304256439209,
+      "learning_rate": 9.9336848936017e-05,
+      "loss": 3.174900436401367,
+      "step": 6550
+    },
+    {
+      "epoch": 1.2593659942363113,
+      "grad_norm": 0.5404880046844482,
+      "learning_rate": 9.911265052471046e-05,
+      "loss": 3.1754596710205076,
+      "step": 6555
+    },
+    {
+      "epoch": 1.260326609029779,
+      "grad_norm": 0.6277545690536499,
+      "learning_rate": 9.888858050740703e-05,
+      "loss": 3.1767545700073243,
+      "step": 6560
+    },
+    {
+      "epoch": 1.2612872238232469,
+      "grad_norm": 0.4282405972480774,
+      "learning_rate": 9.866463944945839e-05,
+      "loss": 3.175259590148926,
+      "step": 6565
+    },
+    {
+      "epoch": 1.2622478386167146,
+      "grad_norm": 0.740606963634491,
+      "learning_rate": 9.844082791589075e-05,
+      "loss": 3.1768466949462892,
+      "step": 6570
+    },
+    {
+      "epoch": 1.2632084534101824,
+      "grad_norm": 0.6880332231521606,
+      "learning_rate": 9.821714647140363e-05,
+      "loss": 3.1770793914794924,
+      "step": 6575
+    },
+    {
+      "epoch": 1.2641690682036504,
+      "grad_norm": 0.5514328479766846,
+      "learning_rate": 9.799359568036821e-05,
+      "loss": 3.180289649963379,
+      "step": 6580
+    },
+    {
+      "epoch": 1.2651296829971181,
+      "grad_norm": 0.569648027420044,
+      "learning_rate": 9.777017610682612e-05,
+      "loss": 3.1760627746582033,
+      "step": 6585
+    },
+    {
+      "epoch": 1.266090297790586,
+      "grad_norm": 0.6740429401397705,
+      "learning_rate": 9.754688831448781e-05,
+      "loss": 3.173795700073242,
+      "step": 6590
+    },
+    {
+      "epoch": 1.267050912584054,
+      "grad_norm": 0.6410177946090698,
+      "learning_rate": 9.732373286673135e-05,
+      "loss": 3.178542709350586,
+      "step": 6595
+    },
+    {
+      "epoch": 1.2680115273775217,
+      "grad_norm": 0.5730255246162415,
+      "learning_rate": 9.710071032660076e-05,
+      "loss": 3.1776718139648437,
+      "step": 6600
+    },
+    {
+      "epoch": 1.2689721421709894,
+      "grad_norm": 0.6753981113433838,
+      "learning_rate": 9.687782125680484e-05,
+      "loss": 3.1791826248168946,
+      "step": 6605
+    },
+    {
+      "epoch": 1.2699327569644572,
+      "grad_norm": 0.5752228498458862,
+      "learning_rate": 9.66550662197155e-05,
+      "loss": 3.173518753051758,
+      "step": 6610
+    },
+    {
+      "epoch": 1.270893371757925,
+      "grad_norm": 0.5868836045265198,
+      "learning_rate": 9.643244577736667e-05,
+      "loss": 3.173557090759277,
+      "step": 6615
+    },
+    {
+      "epoch": 1.271853986551393,
+      "grad_norm": 0.5383107662200928,
+      "learning_rate": 9.620996049145247e-05,
+      "loss": 3.1773754119873048,
+      "step": 6620
+    },
+    {
+      "epoch": 1.2728146013448607,
+      "grad_norm": 0.5838056802749634,
+      "learning_rate": 9.598761092332616e-05,
+      "loss": 3.1778125762939453,
+      "step": 6625
+    },
+    {
+      "epoch": 1.2737752161383284,
+      "grad_norm": 0.5867598056793213,
+      "learning_rate": 9.576539763399847e-05,
+      "loss": 3.183711814880371,
+      "step": 6630
+    },
+    {
+      "epoch": 1.2747358309317964,
+      "grad_norm": 0.6111621856689453,
+      "learning_rate": 9.554332118413635e-05,
+      "loss": 3.1793869018554686,
+      "step": 6635
+    },
+    {
+      "epoch": 1.2756964457252642,
+      "grad_norm": 0.8043511509895325,
+      "learning_rate": 9.532138213406143e-05,
+      "loss": 3.175552749633789,
+      "step": 6640
+    },
+    {
+      "epoch": 1.276657060518732,
+      "grad_norm": 0.8225224018096924,
+      "learning_rate": 9.509958104374877e-05,
+      "loss": 3.1786468505859373,
+      "step": 6645
+    },
+    {
+      "epoch": 1.2776176753121997,
+      "grad_norm": 0.6611747741699219,
+      "learning_rate": 9.487791847282517e-05,
+      "loss": 3.1788875579833986,
+      "step": 6650
+    },
+    {
+      "epoch": 1.2785782901056677,
+      "grad_norm": 0.49061068892478943,
+      "learning_rate": 9.465639498056815e-05,
+      "loss": 3.1779201507568358,
+      "step": 6655
+    },
+    {
+      "epoch": 1.2795389048991355,
+      "grad_norm": 0.6303700804710388,
+      "learning_rate": 9.443501112590413e-05,
+      "loss": 3.176443862915039,
+      "step": 6660
+    },
+    {
+      "epoch": 1.2804995196926032,
+      "grad_norm": 0.7213342189788818,
+      "learning_rate": 9.421376746740733e-05,
+      "loss": 3.1703159332275392,
+      "step": 6665
+    },
+    {
+      "epoch": 1.2814601344860712,
+      "grad_norm": 0.6647571921348572,
+      "learning_rate": 9.399266456329815e-05,
+      "loss": 3.1765552520751954,
+      "step": 6670
+    },
+    {
+      "epoch": 1.282420749279539,
+      "grad_norm": 0.6062517762184143,
+      "learning_rate": 9.377170297144196e-05,
+      "loss": 3.1765026092529296,
+      "step": 6675
+    },
+    {
+      "epoch": 1.2833813640730067,
+      "grad_norm": 0.4922938048839569,
+      "learning_rate": 9.355088324934741e-05,
+      "loss": 3.176412582397461,
+      "step": 6680
+    },
+    {
+      "epoch": 1.2843419788664745,
+      "grad_norm": 0.5771006941795349,
+      "learning_rate": 9.333020595416548e-05,
+      "loss": 3.1751476287841798,
+      "step": 6685
+    },
+    {
+      "epoch": 1.2853025936599423,
+      "grad_norm": 0.6197171211242676,
+      "learning_rate": 9.310967164268749e-05,
+      "loss": 3.174680709838867,
+      "step": 6690
+    },
+    {
+      "epoch": 1.2862632084534102,
+      "grad_norm": 0.5435570478439331,
+      "learning_rate": 9.28892808713442e-05,
+      "loss": 3.1765819549560548,
+      "step": 6695
+    },
+    {
+      "epoch": 1.287223823246878,
+      "grad_norm": 0.5521527528762817,
+      "learning_rate": 9.266903419620411e-05,
+      "loss": 3.173064041137695,
+      "step": 6700
+    },
+    {
+      "epoch": 1.2881844380403458,
+      "grad_norm": 0.5107563138008118,
+      "learning_rate": 9.24489321729722e-05,
+      "loss": 3.1766895294189452,
+      "step": 6705
+    },
+    {
+      "epoch": 1.2891450528338138,
+      "grad_norm": 0.66566401720047,
+      "learning_rate": 9.222897535698841e-05,
+      "loss": 3.175777053833008,
+      "step": 6710
+    },
+    {
+      "epoch": 1.2901056676272815,
+      "grad_norm": 0.501441478729248,
+      "learning_rate": 9.20091643032264e-05,
+      "loss": 3.1773338317871094,
+      "step": 6715
+    },
+    {
+      "epoch": 1.2910662824207493,
+      "grad_norm": 0.7562515735626221,
+      "learning_rate": 9.178949956629195e-05,
+      "loss": 3.1776708602905273,
+      "step": 6720
+    },
+    {
+      "epoch": 1.292026897214217,
+      "grad_norm": 0.7048882246017456,
+      "learning_rate": 9.156998170042182e-05,
+      "loss": 3.1779850006103514,
+      "step": 6725
+    },
+    {
+      "epoch": 1.2929875120076848,
+      "grad_norm": 0.5082115530967712,
+      "learning_rate": 9.1350611259482e-05,
+      "loss": 3.174315071105957,
+      "step": 6730
+    },
+    {
+      "epoch": 1.2939481268011528,
+      "grad_norm": 0.5481780171394348,
+      "learning_rate": 9.113138879696667e-05,
+      "loss": 3.174916648864746,
+      "step": 6735
+    },
+    {
+      "epoch": 1.2949087415946205,
+      "grad_norm": 0.5947979688644409,
+      "learning_rate": 9.091231486599656e-05,
+      "loss": 3.177842903137207,
+      "step": 6740
+    },
+    {
+      "epoch": 1.2958693563880883,
+      "grad_norm": 0.5829041600227356,
+      "learning_rate": 9.069339001931775e-05,
+      "loss": 3.1755516052246096,
+      "step": 6745
+    },
+    {
+      "epoch": 1.2968299711815563,
+      "grad_norm": 0.7779812812805176,
+      "learning_rate": 9.047461480929996e-05,
+      "loss": 3.1758398056030273,
+      "step": 6750
+    },
+    {
+      "epoch": 1.297790585975024,
+      "grad_norm": 0.5685781836509705,
+      "learning_rate": 9.025598978793564e-05,
+      "loss": 3.1764198303222657,
+      "step": 6755
+    },
+    {
+      "epoch": 1.2987512007684918,
+      "grad_norm": 0.5285525918006897,
+      "learning_rate": 9.003751550683803e-05,
+      "loss": 3.1752038955688477,
+      "step": 6760
+    },
+    {
+      "epoch": 1.2997118155619596,
+      "grad_norm": 0.6162580251693726,
+      "learning_rate": 8.981919251724023e-05,
+      "loss": 3.172860336303711,
+      "step": 6765
+    },
+    {
+      "epoch": 1.3006724303554273,
+      "grad_norm": 0.6464502215385437,
+      "learning_rate": 8.960102136999346e-05,
+      "loss": 3.174723815917969,
+      "step": 6770
+    },
+    {
+      "epoch": 1.3016330451488953,
+      "grad_norm": 0.472971111536026,
+      "learning_rate": 8.938300261556599e-05,
+      "loss": 3.176887321472168,
+      "step": 6775
+    },
+    {
+      "epoch": 1.302593659942363,
+      "grad_norm": 0.46003440022468567,
+      "learning_rate": 8.916513680404145e-05,
+      "loss": 3.176510047912598,
+      "step": 6780
+    },
+    {
+      "epoch": 1.3035542747358309,
+      "grad_norm": 0.7004385590553284,
+      "learning_rate": 8.894742448511766e-05,
+      "loss": 3.1788423538208006,
+      "step": 6785
+    },
+    {
+      "epoch": 1.3045148895292988,
+      "grad_norm": 0.6769760847091675,
+      "learning_rate": 8.872986620810508e-05,
+      "loss": 3.175991439819336,
+      "step": 6790
+    },
+    {
+      "epoch": 1.3054755043227666,
+      "grad_norm": 0.593636691570282,
+      "learning_rate": 8.851246252192566e-05,
+      "loss": 3.1727279663085937,
+      "step": 6795
+    },
+    {
+      "epoch": 1.3064361191162344,
+      "grad_norm": 0.580614447593689,
+      "learning_rate": 8.82952139751111e-05,
+      "loss": 3.178467559814453,
+      "step": 6800
+    },
+    {
+      "epoch": 1.3073967339097021,
+      "grad_norm": 0.519120991230011,
+      "learning_rate": 8.807812111580186e-05,
+      "loss": 3.1754281997680662,
+      "step": 6805
+    },
+    {
+      "epoch": 1.30835734870317,
+      "grad_norm": 0.5653340816497803,
+      "learning_rate": 8.786118449174541e-05,
+      "loss": 3.1773509979248047,
+      "step": 6810
+    },
+    {
+      "epoch": 1.3093179634966379,
+      "grad_norm": 0.7600631713867188,
+      "learning_rate": 8.764440465029513e-05,
+      "loss": 3.1753978729248047,
+      "step": 6815
+    },
+    {
+      "epoch": 1.3102785782901056,
+      "grad_norm": 0.5709202885627747,
+      "learning_rate": 8.742778213840888e-05,
+      "loss": 3.176059341430664,
+      "step": 6820
+    },
+    {
+      "epoch": 1.3112391930835736,
+      "grad_norm": 0.5290617346763611,
+      "learning_rate": 8.721131750264737e-05,
+      "loss": 3.176310348510742,
+      "step": 6825
+    },
+    {
+      "epoch": 1.3121998078770414,
+      "grad_norm": 0.5382722020149231,
+      "learning_rate": 8.699501128917321e-05,
+      "loss": 3.1754171371459963,
+      "step": 6830
+    },
+    {
+      "epoch": 1.3131604226705091,
+      "grad_norm": 0.5834631323814392,
+      "learning_rate": 8.677886404374903e-05,
+      "loss": 3.176200103759766,
+      "step": 6835
+    },
+    {
+      "epoch": 1.314121037463977,
+      "grad_norm": 0.44905054569244385,
+      "learning_rate": 8.656287631173656e-05,
+      "loss": 3.178561210632324,
+      "step": 6840
+    },
+    {
+      "epoch": 1.3150816522574447,
+      "grad_norm": 0.4696488380432129,
+      "learning_rate": 8.634704863809501e-05,
+      "loss": 3.1775485992431642,
+      "step": 6845
+    },
+    {
+      "epoch": 1.3160422670509127,
+      "grad_norm": 0.5930847525596619,
+      "learning_rate": 8.61313815673798e-05,
+      "loss": 3.176204299926758,
+      "step": 6850
+    },
+    {
+      "epoch": 1.3170028818443804,
+      "grad_norm": 0.6078386902809143,
+      "learning_rate": 8.591587564374094e-05,
+      "loss": 3.175630569458008,
+      "step": 6855
+    },
+    {
+      "epoch": 1.3179634966378482,
+      "grad_norm": 0.6279536485671997,
+      "learning_rate": 8.570053141092209e-05,
+      "loss": 3.1791570663452147,
+      "step": 6860
+    },
+    {
+      "epoch": 1.3189241114313162,
+      "grad_norm": 0.4797205328941345,
+      "learning_rate": 8.548534941225877e-05,
+      "loss": 3.1787494659423827,
+      "step": 6865
+    },
+    {
+      "epoch": 1.319884726224784,
+      "grad_norm": 0.692002534866333,
+      "learning_rate": 8.527033019067736e-05,
+      "loss": 3.1755470275878905,
+      "step": 6870
+    },
+    {
+      "epoch": 1.3208453410182517,
+      "grad_norm": 0.609710693359375,
+      "learning_rate": 8.505547428869326e-05,
+      "loss": 3.1747051239013673,
+      "step": 6875
+    },
+    {
+      "epoch": 1.3218059558117194,
+      "grad_norm": 0.511199414730072,
+      "learning_rate": 8.484078224841002e-05,
+      "loss": 3.176365280151367,
+      "step": 6880
+    },
+    {
+      "epoch": 1.3227665706051872,
+      "grad_norm": 0.4836551547050476,
+      "learning_rate": 8.462625461151769e-05,
+      "loss": 3.1734235763549803,
+      "step": 6885
+    },
+    {
+      "epoch": 1.3237271853986552,
+      "grad_norm": 0.6896669268608093,
+      "learning_rate": 8.441189191929152e-05,
+      "loss": 3.1753162384033202,
+      "step": 6890
+    },
+    {
+      "epoch": 1.324687800192123,
+      "grad_norm": 0.5728012323379517,
+      "learning_rate": 8.419769471259053e-05,
+      "loss": 3.1740385055541993,
+      "step": 6895
+    },
+    {
+      "epoch": 1.3256484149855907,
+      "grad_norm": 0.6955984234809875,
+      "learning_rate": 8.398366353185633e-05,
+      "loss": 3.175432586669922,
+      "step": 6900
+    },
+    {
+      "epoch": 1.3266090297790587,
+      "grad_norm": 0.48423975706100464,
+      "learning_rate": 8.376979891711145e-05,
+      "loss": 3.1750982284545897,
+      "step": 6905
+    },
+    {
+      "epoch": 1.3275696445725265,
+      "grad_norm": 0.4743121862411499,
+      "learning_rate": 8.355610140795827e-05,
+      "loss": 3.1710472106933594,
+      "step": 6910
+    },
+    {
+      "epoch": 1.3285302593659942,
+      "grad_norm": 0.6560218334197998,
+      "learning_rate": 8.334257154357755e-05,
+      "loss": 3.1815349578857424,
+      "step": 6915
+    },
+    {
+      "epoch": 1.329490874159462,
+      "grad_norm": 0.536353588104248,
+      "learning_rate": 8.31292098627271e-05,
+      "loss": 3.1730472564697267,
+      "step": 6920
+    },
+    {
+      "epoch": 1.3304514889529298,
+      "grad_norm": 0.49340781569480896,
+      "learning_rate": 8.291601690374021e-05,
+      "loss": 3.1722347259521486,
+      "step": 6925
+    },
+    {
+      "epoch": 1.3314121037463977,
+      "grad_norm": 0.5234084129333496,
+      "learning_rate": 8.270299320452467e-05,
+      "loss": 3.172749710083008,
+      "step": 6930
+    },
+    {
+      "epoch": 1.3323727185398655,
+      "grad_norm": 0.6601616740226746,
+      "learning_rate": 8.249013930256108e-05,
+      "loss": 3.1766475677490233,
+      "step": 6935
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.7026968002319336,
+      "learning_rate": 8.227745573490176e-05,
+      "loss": 3.173655319213867,
+      "step": 6940
+    },
+    {
+      "epoch": 1.3342939481268012,
+      "grad_norm": 0.5302594304084778,
+      "learning_rate": 8.206494303816904e-05,
+      "loss": 3.178236389160156,
+      "step": 6945
+    },
+    {
+      "epoch": 1.335254562920269,
+      "grad_norm": 0.46280986070632935,
+      "learning_rate": 8.185260174855433e-05,
+      "loss": 3.173746871948242,
+      "step": 6950
+    },
+    {
+      "epoch": 1.3362151777137368,
+      "grad_norm": 0.5142591595649719,
+      "learning_rate": 8.164043240181646e-05,
+      "loss": 3.173639106750488,
+      "step": 6955
+    },
+    {
+      "epoch": 1.3371757925072045,
+      "grad_norm": 0.682403028011322,
+      "learning_rate": 8.142843553328048e-05,
+      "loss": 3.175322151184082,
+      "step": 6960
+    },
+    {
+      "epoch": 1.3381364073006723,
+      "grad_norm": 0.6334633231163025,
+      "learning_rate": 8.121661167783625e-05,
+      "loss": 3.17366943359375,
+      "step": 6965
+    },
+    {
+      "epoch": 1.3390970220941403,
+      "grad_norm": 0.606118381023407,
+      "learning_rate": 8.100496136993712e-05,
+      "loss": 3.1736032485961916,
+      "step": 6970
+    },
+    {
+      "epoch": 1.340057636887608,
+      "grad_norm": 0.5890917778015137,
+      "learning_rate": 8.07934851435984e-05,
+      "loss": 3.172843360900879,
+      "step": 6975
+    },
+    {
+      "epoch": 1.341018251681076,
+      "grad_norm": 0.47088661789894104,
+      "learning_rate": 8.05821835323964e-05,
+      "loss": 3.1722528457641603,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3419788664745438,
+      "grad_norm": 0.48279380798339844,
+      "learning_rate": 8.037105706946673e-05,
+      "loss": 3.177801513671875,
+      "step": 6985
+    },
+    {
+      "epoch": 1.3429394812680115,
+      "grad_norm": 0.5353664755821228,
+      "learning_rate": 8.01601062875032e-05,
+      "loss": 3.1740434646606444,
+      "step": 6990
+    },
+    {
+      "epoch": 1.3439000960614793,
+      "grad_norm": 0.5050175189971924,
+      "learning_rate": 7.994933171875613e-05,
+      "loss": 3.1727882385253907,
+      "step": 6995
+    },
+    {
+      "epoch": 1.344860710854947,
+      "grad_norm": 0.44954636693000793,
+      "learning_rate": 7.973873389503149e-05,
+      "loss": 3.1749853134155273,
+      "step": 7000
+    },
+    {
+      "epoch": 1.345821325648415,
+      "grad_norm": 0.6471356749534607,
+      "learning_rate": 7.952831334768913e-05,
+      "loss": 3.1730751037597655,
+      "step": 7005
+    },
+    {
+      "epoch": 1.3467819404418828,
+      "grad_norm": 0.5839700102806091,
+      "learning_rate": 7.931807060764176e-05,
+      "loss": 3.172525405883789,
+      "step": 7010
+    },
+    {
+      "epoch": 1.3477425552353506,
+      "grad_norm": 0.5643643140792847,
+      "learning_rate": 7.910800620535329e-05,
+      "loss": 3.172414016723633,
+      "step": 7015
+    },
+    {
+      "epoch": 1.3487031700288186,
+      "grad_norm": 0.5194916129112244,
+      "learning_rate": 7.889812067083777e-05,
+      "loss": 3.1767629623413085,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3496637848222863,
+      "grad_norm": 0.4932091236114502,
+      "learning_rate": 7.868841453365797e-05,
+      "loss": 3.1722953796386717,
+      "step": 7025
+    },
+    {
+      "epoch": 1.350624399615754,
+      "grad_norm": 0.5307772159576416,
+      "learning_rate": 7.847888832292397e-05,
+      "loss": 3.1748939514160157,
+      "step": 7030
+    },
+    {
+      "epoch": 1.3515850144092219,
+      "grad_norm": 0.5100224614143372,
+      "learning_rate": 7.826954256729188e-05,
+      "loss": 3.1737871170043945,
+      "step": 7035
+    },
+    {
+      "epoch": 1.3525456292026896,
+      "grad_norm": 0.5833878517150879,
+      "learning_rate": 7.806037779496264e-05,
+      "loss": 3.171864128112793,
+      "step": 7040
+    },
+    {
+      "epoch": 1.3535062439961576,
+      "grad_norm": 0.4274962842464447,
+      "learning_rate": 7.785139453368025e-05,
+      "loss": 3.174275207519531,
+      "step": 7045
+    },
+    {
+      "epoch": 1.3544668587896254,
+      "grad_norm": 0.43694546818733215,
+      "learning_rate": 7.764259331073096e-05,
+      "loss": 3.1718505859375,
+      "step": 7050
+    },
+    {
+      "epoch": 1.3554274735830931,
+      "grad_norm": 0.5787099003791809,
+      "learning_rate": 7.743397465294174e-05,
+      "loss": 3.171575164794922,
+      "step": 7055
+    },
+    {
+      "epoch": 1.356388088376561,
+      "grad_norm": 0.45274344086647034,
+      "learning_rate": 7.72255390866789e-05,
+      "loss": 3.169318199157715,
+      "step": 7060
+    },
+    {
+      "epoch": 1.3573487031700289,
+      "grad_norm": 0.490500807762146,
+      "learning_rate": 7.701728713784662e-05,
+      "loss": 3.172596740722656,
+      "step": 7065
+    },
+    {
+      "epoch": 1.3583093179634966,
+      "grad_norm": 0.8986951112747192,
+      "learning_rate": 7.6809219331886e-05,
+      "loss": 3.177069664001465,
+      "step": 7070
+    },
+    {
+      "epoch": 1.3592699327569644,
+      "grad_norm": 0.6794254183769226,
+      "learning_rate": 7.660133619377346e-05,
+      "loss": 3.1766523361206054,
+      "step": 7075
+    },
+    {
+      "epoch": 1.3602305475504322,
+      "grad_norm": 0.4294492304325104,
+      "learning_rate": 7.639363824801957e-05,
+      "loss": 3.171103668212891,
+      "step": 7080
+    },
+    {
+      "epoch": 1.3611911623439001,
+      "grad_norm": 0.4653557240962982,
+      "learning_rate": 7.618612601866738e-05,
+      "loss": 3.1739612579345704,
+      "step": 7085
+    },
+    {
+      "epoch": 1.362151777137368,
+      "grad_norm": 0.48339158296585083,
+      "learning_rate": 7.597880002929155e-05,
+      "loss": 3.174298095703125,
+      "step": 7090
+    },
+    {
+      "epoch": 1.3631123919308357,
+      "grad_norm": 0.4108154773712158,
+      "learning_rate": 7.577166080299704e-05,
+      "loss": 3.1711090087890623,
+      "step": 7095
+    },
+    {
+      "epoch": 1.3640730067243036,
+      "grad_norm": 0.5386460423469543,
+      "learning_rate": 7.556470886241715e-05,
+      "loss": 3.172881317138672,
+      "step": 7100
+    },
+    {
+      "epoch": 1.3650336215177714,
+      "grad_norm": 0.521826982498169,
+      "learning_rate": 7.535794472971292e-05,
+      "loss": 3.1787548065185547,
+      "step": 7105
+    },
+    {
+      "epoch": 1.3659942363112392,
+      "grad_norm": 0.5726050138473511,
+      "learning_rate": 7.515136892657152e-05,
+      "loss": 3.1746589660644533,
+      "step": 7110
+    },
+    {
+      "epoch": 1.366954851104707,
+      "grad_norm": 0.49367332458496094,
+      "learning_rate": 7.494498197420486e-05,
+      "loss": 3.171868896484375,
+      "step": 7115
+    },
+    {
+      "epoch": 1.3679154658981747,
+      "grad_norm": 0.6521613597869873,
+      "learning_rate": 7.473878439334833e-05,
+      "loss": 3.1738697052001954,
+      "step": 7120
+    },
+    {
+      "epoch": 1.3688760806916427,
+      "grad_norm": 0.4904996454715729,
+      "learning_rate": 7.453277670425964e-05,
+      "loss": 3.1751945495605467,
+      "step": 7125
+    },
+    {
+      "epoch": 1.3698366954851104,
+      "grad_norm": 0.5470160841941833,
+      "learning_rate": 7.432695942671736e-05,
+      "loss": 3.1728076934814453,
+      "step": 7130
+    },
+    {
+      "epoch": 1.3707973102785782,
+      "grad_norm": 0.6032967567443848,
+      "learning_rate": 7.412133308001952e-05,
+      "loss": 3.166953468322754,
+      "step": 7135
+    },
+    {
+      "epoch": 1.3717579250720462,
+      "grad_norm": 0.4871714413166046,
+      "learning_rate": 7.39158981829825e-05,
+      "loss": 3.1728256225585936,
+      "step": 7140
+    },
+    {
+      "epoch": 1.372718539865514,
+      "grad_norm": 0.42759451270103455,
+      "learning_rate": 7.371065525393965e-05,
+      "loss": 3.1744380950927735,
+      "step": 7145
+    },
+    {
+      "epoch": 1.3736791546589817,
+      "grad_norm": 0.5663604736328125,
+      "learning_rate": 7.350560481073996e-05,
+      "loss": 3.1777429580688477,
+      "step": 7150
+    },
+    {
+      "epoch": 1.3746397694524495,
+      "grad_norm": 0.5459421873092651,
+      "learning_rate": 7.330074737074665e-05,
+      "loss": 3.170821762084961,
+      "step": 7155
+    },
+    {
+      "epoch": 1.3756003842459175,
+      "grad_norm": 0.46091675758361816,
+      "learning_rate": 7.309608345083605e-05,
+      "loss": 3.1742374420166017,
+      "step": 7160
+    },
+    {
+      "epoch": 1.3765609990393852,
+      "grad_norm": 0.38046908378601074,
+      "learning_rate": 7.289161356739638e-05,
+      "loss": 3.1713001251220705,
+      "step": 7165
+    },
+    {
+      "epoch": 1.377521613832853,
+      "grad_norm": 0.468205064535141,
+      "learning_rate": 7.268733823632601e-05,
+      "loss": 3.1741426467895506,
+      "step": 7170
+    },
+    {
+      "epoch": 1.378482228626321,
+      "grad_norm": 0.6682205200195312,
+      "learning_rate": 7.248325797303256e-05,
+      "loss": 3.1712989807128906,
+      "step": 7175
+    },
+    {
+      "epoch": 1.3794428434197887,
+      "grad_norm": 0.5630512833595276,
+      "learning_rate": 7.227937329243149e-05,
+      "loss": 3.171531867980957,
+      "step": 7180
+    },
+    {
+      "epoch": 1.3804034582132565,
+      "grad_norm": 0.54999178647995,
+      "learning_rate": 7.20756847089448e-05,
+      "loss": 3.173867416381836,
+      "step": 7185
+    },
+    {
+      "epoch": 1.3813640730067243,
+      "grad_norm": 0.43921148777008057,
+      "learning_rate": 7.187219273649962e-05,
+      "loss": 3.170378494262695,
+      "step": 7190
+    },
+    {
+      "epoch": 1.382324687800192,
+      "grad_norm": 0.4150092601776123,
+      "learning_rate": 7.166889788852707e-05,
+      "loss": 3.170474624633789,
+      "step": 7195
+    },
+    {
+      "epoch": 1.38328530259366,
+      "grad_norm": 0.38194531202316284,
+      "learning_rate": 7.146580067796102e-05,
+      "loss": 3.1690601348876952,
+      "step": 7200
+    },
+    {
+      "epoch": 1.3842459173871278,
+      "grad_norm": 0.7853124141693115,
+      "learning_rate": 7.126290161723642e-05,
+      "loss": 3.1698192596435546,
+      "step": 7205
+    },
+    {
+      "epoch": 1.3852065321805955,
+      "grad_norm": 0.4423576593399048,
+      "learning_rate": 7.106020121828848e-05,
+      "loss": 3.1732460021972657,
+      "step": 7210
+    },
+    {
+      "epoch": 1.3861671469740635,
+      "grad_norm": 0.4560396671295166,
+      "learning_rate": 7.085769999255108e-05,
+      "loss": 3.1727630615234377,
+      "step": 7215
+    },
+    {
+      "epoch": 1.3871277617675313,
+      "grad_norm": 0.41349244117736816,
+      "learning_rate": 7.065539845095567e-05,
+      "loss": 3.1731325149536134,
+      "step": 7220
+    },
+    {
+      "epoch": 1.388088376560999,
+      "grad_norm": 0.47393569350242615,
+      "learning_rate": 7.045329710392967e-05,
+      "loss": 3.1727859497070314,
+      "step": 7225
+    },
+    {
+      "epoch": 1.3890489913544668,
+      "grad_norm": 0.7396698594093323,
+      "learning_rate": 7.025139646139553e-05,
+      "loss": 3.1715877532958983,
+      "step": 7230
+    },
+    {
+      "epoch": 1.3900096061479346,
+      "grad_norm": 0.7079951763153076,
+      "learning_rate": 7.004969703276941e-05,
+      "loss": 3.169532585144043,
+      "step": 7235
+    },
+    {
+      "epoch": 1.3909702209414025,
+      "grad_norm": 0.46009525656700134,
+      "learning_rate": 6.984819932695956e-05,
+      "loss": 3.169749450683594,
+      "step": 7240
+    },
+    {
+      "epoch": 1.3919308357348703,
+      "grad_norm": 0.4487673044204712,
+      "learning_rate": 6.964690385236534e-05,
+      "loss": 3.1695636749267577,
+      "step": 7245
+    },
+    {
+      "epoch": 1.392891450528338,
+      "grad_norm": 0.4718167781829834,
+      "learning_rate": 6.944581111687593e-05,
+      "loss": 3.168662452697754,
+      "step": 7250
+    },
+    {
+      "epoch": 1.393852065321806,
+      "grad_norm": 0.46017783880233765,
+      "learning_rate": 6.924492162786898e-05,
+      "loss": 3.169230842590332,
+      "step": 7255
+    },
+    {
+      "epoch": 1.3948126801152738,
+      "grad_norm": 0.45555198192596436,
+      "learning_rate": 6.904423589220917e-05,
+      "loss": 3.169890594482422,
+      "step": 7260
+    },
+    {
+      "epoch": 1.3957732949087416,
+      "grad_norm": 0.4948507845401764,
+      "learning_rate": 6.884375441624724e-05,
+      "loss": 3.173093795776367,
+      "step": 7265
+    },
+    {
+      "epoch": 1.3967339097022093,
+      "grad_norm": 0.43780985474586487,
+      "learning_rate": 6.864347770581859e-05,
+      "loss": 3.168751335144043,
+      "step": 7270
+    },
+    {
+      "epoch": 1.397694524495677,
+      "grad_norm": 0.4438318908214569,
+      "learning_rate": 6.844340626624178e-05,
+      "loss": 3.1701900482177736,
+      "step": 7275
+    },
+    {
+      "epoch": 1.398655139289145,
+      "grad_norm": 0.5062385201454163,
+      "learning_rate": 6.824354060231765e-05,
+      "loss": 3.1665103912353514,
+      "step": 7280
+    },
+    {
+      "epoch": 1.3996157540826129,
+      "grad_norm": 0.4086364507675171,
+      "learning_rate": 6.804388121832777e-05,
+      "loss": 3.1716842651367188,
+      "step": 7285
+    },
+    {
+      "epoch": 1.4005763688760806,
+      "grad_norm": 0.5124282836914062,
+      "learning_rate": 6.784442861803331e-05,
+      "loss": 3.1696632385253904,
+      "step": 7290
+    },
+    {
+      "epoch": 1.4015369836695486,
+      "grad_norm": 0.5204159617424011,
+      "learning_rate": 6.764518330467346e-05,
+      "loss": 3.1707353591918945,
+      "step": 7295
+    },
+    {
+      "epoch": 1.4024975984630164,
+      "grad_norm": 0.6575141549110413,
+      "learning_rate": 6.744614578096475e-05,
+      "loss": 3.171962547302246,
+      "step": 7300
+    },
+    {
+      "epoch": 1.4034582132564841,
+      "grad_norm": 0.5063356757164001,
+      "learning_rate": 6.72473165490993e-05,
+      "loss": 3.1709272384643556,
+      "step": 7305
+    },
+    {
+      "epoch": 1.4044188280499519,
+      "grad_norm": 0.4604819416999817,
+      "learning_rate": 6.704869611074351e-05,
+      "loss": 3.173125076293945,
+      "step": 7310
+    },
+    {
+      "epoch": 1.4053794428434199,
+      "grad_norm": 0.4766218960285187,
+      "learning_rate": 6.685028496703719e-05,
+      "loss": 3.1713457107543945,
+      "step": 7315
+    },
+    {
+      "epoch": 1.4063400576368876,
+      "grad_norm": 0.48007673025131226,
+      "learning_rate": 6.665208361859203e-05,
+      "loss": 3.1723331451416015,
+      "step": 7320
+    },
+    {
+      "epoch": 1.4073006724303554,
+      "grad_norm": 0.3352242410182953,
+      "learning_rate": 6.64540925654904e-05,
+      "loss": 3.171708679199219,
+      "step": 7325
+    },
+    {
+      "epoch": 1.4082612872238234,
+      "grad_norm": 0.6097206473350525,
+      "learning_rate": 6.625631230728393e-05,
+      "loss": 3.169724464416504,
+      "step": 7330
+    },
+    {
+      "epoch": 1.4092219020172911,
+      "grad_norm": 0.45327913761138916,
+      "learning_rate": 6.605874334299257e-05,
+      "loss": 3.172129821777344,
+      "step": 7335
+    },
+    {
+      "epoch": 1.410182516810759,
+      "grad_norm": 0.3918154537677765,
+      "learning_rate": 6.586138617110312e-05,
+      "loss": 3.1696929931640625,
+      "step": 7340
+    },
+    {
+      "epoch": 1.4111431316042267,
+      "grad_norm": 0.45734044909477234,
+      "learning_rate": 6.566424128956788e-05,
+      "loss": 3.169087600708008,
+      "step": 7345
+    },
+    {
+      "epoch": 1.4121037463976944,
+      "grad_norm": 0.41583287715911865,
+      "learning_rate": 6.546730919580365e-05,
+      "loss": 3.169902801513672,
+      "step": 7350
+    },
+    {
+      "epoch": 1.4130643611911624,
+      "grad_norm": 0.3783586919307709,
+      "learning_rate": 6.527059038669032e-05,
+      "loss": 3.173455810546875,
+      "step": 7355
+    },
+    {
+      "epoch": 1.4140249759846302,
+      "grad_norm": 0.5969278216362,
+      "learning_rate": 6.507408535856968e-05,
+      "loss": 3.171731185913086,
+      "step": 7360
+    },
+    {
+      "epoch": 1.414985590778098,
+      "grad_norm": 0.5529734492301941,
+      "learning_rate": 6.487779460724388e-05,
+      "loss": 3.169993591308594,
+      "step": 7365
+    },
+    {
+      "epoch": 1.415946205571566,
+      "grad_norm": 0.460746705532074,
+      "learning_rate": 6.468171862797484e-05,
+      "loss": 3.1684539794921873,
+      "step": 7370
+    },
+    {
+      "epoch": 1.4169068203650337,
+      "grad_norm": 0.40012434124946594,
+      "learning_rate": 6.448585791548231e-05,
+      "loss": 3.170420837402344,
+      "step": 7375
+    },
+    {
+      "epoch": 1.4178674351585014,
+      "grad_norm": 0.4528000056743622,
+      "learning_rate": 6.42902129639429e-05,
+      "loss": 3.1724185943603516,
+      "step": 7380
+    },
+    {
+      "epoch": 1.4188280499519692,
+      "grad_norm": 0.5503178834915161,
+      "learning_rate": 6.409478426698893e-05,
+      "loss": 3.170566749572754,
+      "step": 7385
+    },
+    {
+      "epoch": 1.419788664745437,
+      "grad_norm": 0.4782489240169525,
+      "learning_rate": 6.389957231770705e-05,
+      "loss": 3.1753549575805664,
+      "step": 7390
+    },
+    {
+      "epoch": 1.420749279538905,
+      "grad_norm": 0.5017015933990479,
+      "learning_rate": 6.370457760863708e-05,
+      "loss": 3.1712177276611326,
+      "step": 7395
+    },
+    {
+      "epoch": 1.4217098943323727,
+      "grad_norm": 0.4609530568122864,
+      "learning_rate": 6.35098006317706e-05,
+      "loss": 3.1715621948242188,
+      "step": 7400
+    },
+    {
+      "epoch": 1.4226705091258405,
+      "grad_norm": 0.43000540137290955,
+      "learning_rate": 6.331524187854992e-05,
+      "loss": 3.1696197509765627,
+      "step": 7405
+    },
+    {
+      "epoch": 1.4236311239193085,
+      "grad_norm": 0.6466848850250244,
+      "learning_rate": 6.31209018398668e-05,
+      "loss": 3.171028900146484,
+      "step": 7410
+    },
+    {
+      "epoch": 1.4245917387127762,
+      "grad_norm": 0.4881599247455597,
+      "learning_rate": 6.292678100606101e-05,
+      "loss": 3.1710824966430664,
+      "step": 7415
+    },
+    {
+      "epoch": 1.425552353506244,
+      "grad_norm": 0.42242470383644104,
+      "learning_rate": 6.273287986691934e-05,
+      "loss": 3.1717771530151366,
+      "step": 7420
+    },
+    {
+      "epoch": 1.4265129682997117,
+      "grad_norm": 0.40897586941719055,
+      "learning_rate": 6.253919891167427e-05,
+      "loss": 3.168376159667969,
+      "step": 7425
+    },
+    {
+      "epoch": 1.4274735830931795,
+      "grad_norm": 0.31631234288215637,
+      "learning_rate": 6.23457386290028e-05,
+      "loss": 3.167464828491211,
+      "step": 7430
+    },
+    {
+      "epoch": 1.4284341978866475,
+      "grad_norm": 0.423149973154068,
+      "learning_rate": 6.215249950702488e-05,
+      "loss": 3.1704282760620117,
+      "step": 7435
+    },
+    {
+      "epoch": 1.4293948126801153,
+      "grad_norm": 0.663411557674408,
+      "learning_rate": 6.195948203330282e-05,
+      "loss": 3.170808219909668,
+      "step": 7440
+    },
+    {
+      "epoch": 1.430355427473583,
+      "grad_norm": 0.35881131887435913,
+      "learning_rate": 6.176668669483948e-05,
+      "loss": 3.1693355560302736,
+      "step": 7445
+    },
+    {
+      "epoch": 1.431316042267051,
+      "grad_norm": 0.3775772452354431,
+      "learning_rate": 6.15741139780772e-05,
+      "loss": 3.1732666015625,
+      "step": 7450
+    },
+    {
+      "epoch": 1.4322766570605188,
+      "grad_norm": 0.40818941593170166,
+      "learning_rate": 6.138176436889672e-05,
+      "loss": 3.1699262619018556,
+      "step": 7455
+    },
+    {
+      "epoch": 1.4332372718539865,
+      "grad_norm": 0.31840816140174866,
+      "learning_rate": 6.118963835261588e-05,
+      "loss": 3.171256256103516,
+      "step": 7460
+    },
+    {
+      "epoch": 1.4341978866474543,
+      "grad_norm": 0.45461708307266235,
+      "learning_rate": 6.099773641398834e-05,
+      "loss": 3.1705402374267577,
+      "step": 7465
+    },
+    {
+      "epoch": 1.435158501440922,
+      "grad_norm": 0.49488258361816406,
+      "learning_rate": 6.080605903720229e-05,
+      "loss": 3.1715473175048827,
+      "step": 7470
+    },
+    {
+      "epoch": 1.43611911623439,
+      "grad_norm": 0.4354032278060913,
+      "learning_rate": 6.061460670587948e-05,
+      "loss": 3.170640563964844,
+      "step": 7475
+    },
+    {
+      "epoch": 1.4370797310278578,
+      "grad_norm": 0.39679399132728577,
+      "learning_rate": 6.0423379903073824e-05,
+      "loss": 3.170828437805176,
+      "step": 7480
+    },
+    {
+      "epoch": 1.4380403458213258,
+      "grad_norm": 0.37695419788360596,
+      "learning_rate": 6.023237911127005e-05,
+      "loss": 3.167501449584961,
+      "step": 7485
+    },
+    {
+      "epoch": 1.4390009606147935,
+      "grad_norm": 0.37574145197868347,
+      "learning_rate": 6.004160481238281e-05,
+      "loss": 3.172044372558594,
+      "step": 7490
+    },
+    {
+      "epoch": 1.4399615754082613,
+      "grad_norm": 0.5872917771339417,
+      "learning_rate": 5.9851057487755225e-05,
+      "loss": 3.167539596557617,
+      "step": 7495
+    },
+    {
+      "epoch": 1.440922190201729,
+      "grad_norm": 0.43774786591529846,
+      "learning_rate": 5.966073761815776e-05,
+      "loss": 3.170111656188965,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4418828049951968,
+      "grad_norm": 0.319992333650589,
+      "learning_rate": 5.947064568378696e-05,
+      "loss": 3.1693355560302736,
+      "step": 7505
+    },
+    {
+      "epoch": 1.4428434197886648,
+      "grad_norm": 0.4077332019805908,
+      "learning_rate": 5.928078216426427e-05,
+      "loss": 3.1672155380249025,
+      "step": 7510
+    },
+    {
+      "epoch": 1.4438040345821326,
+      "grad_norm": 0.355669766664505,
+      "learning_rate": 5.909114753863488e-05,
+      "loss": 3.1689855575561525,
+      "step": 7515
+    },
+    {
+      "epoch": 1.4447646493756003,
+      "grad_norm": 0.5554761290550232,
+      "learning_rate": 5.8901742285366315e-05,
+      "loss": 3.171320343017578,
+      "step": 7520
+    },
+    {
+      "epoch": 1.4457252641690683,
+      "grad_norm": 0.4790358543395996,
+      "learning_rate": 5.8712566882347504e-05,
+      "loss": 3.1696239471435548,
+      "step": 7525
+    },
+    {
+      "epoch": 1.446685878962536,
+      "grad_norm": 0.4286898076534271,
+      "learning_rate": 5.8523621806887374e-05,
+      "loss": 3.169162368774414,
+      "step": 7530
+    },
+    {
+      "epoch": 1.4476464937560038,
+      "grad_norm": 0.42978528141975403,
+      "learning_rate": 5.8334907535713836e-05,
+      "loss": 3.172647476196289,
+      "step": 7535
+    },
+    {
+      "epoch": 1.4486071085494716,
+      "grad_norm": 0.4055570363998413,
+      "learning_rate": 5.8146424544972205e-05,
+      "loss": 3.1680522918701173,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4495677233429394,
+      "grad_norm": 0.3212641775608063,
+      "learning_rate": 5.7958173310224465e-05,
+      "loss": 3.168552017211914,
+      "step": 7545
+    },
+    {
+      "epoch": 1.4505283381364074,
+      "grad_norm": 0.40570881962776184,
+      "learning_rate": 5.7770154306447866e-05,
+      "loss": 3.172994041442871,
+      "step": 7550
+    },
+    {
+      "epoch": 1.4514889529298751,
+      "grad_norm": 0.4512649476528168,
+      "learning_rate": 5.758236800803351e-05,
+      "loss": 3.171183395385742,
+      "step": 7555
+    },
+    {
+      "epoch": 1.4524495677233429,
+      "grad_norm": 0.5307871103286743,
+      "learning_rate": 5.739481488878558e-05,
+      "loss": 3.172830581665039,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4534101825168109,
+      "grad_norm": 0.38034340739250183,
+      "learning_rate": 5.720749542191979e-05,
+      "loss": 3.1710861206054686,
+      "step": 7565
+    },
+    {
+      "epoch": 1.4543707973102786,
+      "grad_norm": 0.46515336632728577,
+      "learning_rate": 5.7020410080062416e-05,
+      "loss": 3.1718137741088865,
+      "step": 7570
+    },
+    {
+      "epoch": 1.4553314121037464,
+      "grad_norm": 0.4164998531341553,
+      "learning_rate": 5.683355933524894e-05,
+      "loss": 3.1689435958862306,
+      "step": 7575
+    },
+    {
+      "epoch": 1.4562920268972142,
+      "grad_norm": 0.48741379380226135,
+      "learning_rate": 5.6646943658922975e-05,
+      "loss": 3.167072296142578,
+      "step": 7580
+    },
+    {
+      "epoch": 1.457252641690682,
+      "grad_norm": 0.39116013050079346,
+      "learning_rate": 5.646056352193505e-05,
+      "loss": 3.171469497680664,
+      "step": 7585
+    },
+    {
+      "epoch": 1.45821325648415,
+      "grad_norm": 0.4745250642299652,
+      "learning_rate": 5.6274419394541296e-05,
+      "loss": 3.168760871887207,
+      "step": 7590
+    },
+    {
+      "epoch": 1.4591738712776177,
+      "grad_norm": 0.5160530805587769,
+      "learning_rate": 5.608851174640247e-05,
+      "loss": 3.168941307067871,
+      "step": 7595
+    },
+    {
+      "epoch": 1.4601344860710854,
+      "grad_norm": 0.3649737238883972,
+      "learning_rate": 5.590284104658264e-05,
+      "loss": 3.1685501098632813,
+      "step": 7600
+    },
+    {
+      "epoch": 1.4610951008645534,
+      "grad_norm": 0.37525928020477295,
+      "learning_rate": 5.571740776354811e-05,
+      "loss": 3.169935607910156,
+      "step": 7605
+    },
+    {
+      "epoch": 1.4620557156580212,
+      "grad_norm": 0.35180380940437317,
+      "learning_rate": 5.553221236516594e-05,
+      "loss": 3.1680299758911135,
+      "step": 7610
+    },
+    {
+      "epoch": 1.463016330451489,
+      "grad_norm": 0.5641520023345947,
+      "learning_rate": 5.534725531870317e-05,
+      "loss": 3.171617126464844,
+      "step": 7615
+    },
+    {
+      "epoch": 1.4639769452449567,
+      "grad_norm": 0.5281386375427246,
+      "learning_rate": 5.516253709082547e-05,
+      "loss": 3.1680675506591798,
+      "step": 7620
+    },
+    {
+      "epoch": 1.4649375600384245,
+      "grad_norm": 0.37240076065063477,
+      "learning_rate": 5.4978058147595796e-05,
+      "loss": 3.164999008178711,
+      "step": 7625
+    },
+    {
+      "epoch": 1.4658981748318924,
+      "grad_norm": 0.37205061316490173,
+      "learning_rate": 5.479381895447346e-05,
+      "loss": 3.1697301864624023,
+      "step": 7630
+    },
+    {
+      "epoch": 1.4668587896253602,
+      "grad_norm": 0.35913726687431335,
+      "learning_rate": 5.4609819976312854e-05,
+      "loss": 3.1634567260742186,
+      "step": 7635
+    },
+    {
+      "epoch": 1.4678194044188282,
+      "grad_norm": 0.42885205149650574,
+      "learning_rate": 5.4426061677362284e-05,
+      "loss": 3.173727798461914,
+      "step": 7640
+    },
+    {
+      "epoch": 1.468780019212296,
+      "grad_norm": 0.3557131588459015,
+      "learning_rate": 5.424254452126279e-05,
+      "loss": 3.1679935455322266,
+      "step": 7645
+    },
+    {
+      "epoch": 1.4697406340057637,
+      "grad_norm": 0.4141709804534912,
+      "learning_rate": 5.4059268971047e-05,
+      "loss": 3.1674577713012697,
+      "step": 7650
+    },
+    {
+      "epoch": 1.4707012487992315,
+      "grad_norm": 0.40807923674583435,
+      "learning_rate": 5.387623548913795e-05,
+      "loss": 3.171774673461914,
+      "step": 7655
+    },
+    {
+      "epoch": 1.4716618635926992,
+      "grad_norm": 0.3850151598453522,
+      "learning_rate": 5.36934445373478e-05,
+      "loss": 3.164442443847656,
+      "step": 7660
+    },
+    {
+      "epoch": 1.4726224783861672,
+      "grad_norm": 0.43482470512390137,
+      "learning_rate": 5.3510896576876924e-05,
+      "loss": 3.1658775329589846,
+      "step": 7665
+    },
+    {
+      "epoch": 1.473583093179635,
+      "grad_norm": 0.40651193261146545,
+      "learning_rate": 5.3328592068312565e-05,
+      "loss": 3.164666748046875,
+      "step": 7670
+    },
+    {
+      "epoch": 1.4745437079731027,
+      "grad_norm": 0.33272117376327515,
+      "learning_rate": 5.3146531471627737e-05,
+      "loss": 3.169040298461914,
+      "step": 7675
+    },
+    {
+      "epoch": 1.4755043227665707,
+      "grad_norm": 0.3396502435207367,
+      "learning_rate": 5.29647152461799e-05,
+      "loss": 3.1710128784179688,
+      "step": 7680
+    },
+    {
+      "epoch": 1.4764649375600385,
+      "grad_norm": 0.27867111563682556,
+      "learning_rate": 5.278314385071011e-05,
+      "loss": 3.169612693786621,
+      "step": 7685
+    },
+    {
+      "epoch": 1.4774255523535063,
+      "grad_norm": 0.5024138689041138,
+      "learning_rate": 5.260181774334165e-05,
+      "loss": 3.1680776596069338,
+      "step": 7690
+    },
+    {
+      "epoch": 1.478386167146974,
+      "grad_norm": 0.5927129983901978,
+      "learning_rate": 5.2420737381578814e-05,
+      "loss": 3.1725555419921876,
+      "step": 7695
+    },
+    {
+      "epoch": 1.4793467819404418,
+      "grad_norm": 0.2990652322769165,
+      "learning_rate": 5.223990322230596e-05,
+      "loss": 3.1685482025146485,
+      "step": 7700
+    },
+    {
+      "epoch": 1.4803073967339098,
+      "grad_norm": 0.41576090455055237,
+      "learning_rate": 5.205931572178625e-05,
+      "loss": 3.1704280853271483,
+      "step": 7705
+    },
+    {
+      "epoch": 1.4812680115273775,
+      "grad_norm": 0.33696505427360535,
+      "learning_rate": 5.187897533566047e-05,
+      "loss": 3.1678043365478517,
+      "step": 7710
+    },
+    {
+      "epoch": 1.4822286263208453,
+      "grad_norm": 0.33144330978393555,
+      "learning_rate": 5.169888251894587e-05,
+      "loss": 3.1680809020996095,
+      "step": 7715
+    },
+    {
+      "epoch": 1.4831892411143133,
+      "grad_norm": 0.35232898592948914,
+      "learning_rate": 5.151903772603517e-05,
+      "loss": 3.1694143295288084,
+      "step": 7720
+    },
+    {
+      "epoch": 1.484149855907781,
+      "grad_norm": 0.482440710067749,
+      "learning_rate": 5.1339441410695225e-05,
+      "loss": 3.167070007324219,
+      "step": 7725
+    },
+    {
+      "epoch": 1.4851104707012488,
+      "grad_norm": 0.44148990511894226,
+      "learning_rate": 5.116009402606591e-05,
+      "loss": 3.1692365646362304,
+      "step": 7730
+    },
+    {
+      "epoch": 1.4860710854947166,
+      "grad_norm": 0.4089405834674835,
+      "learning_rate": 5.0980996024659075e-05,
+      "loss": 3.165630912780762,
+      "step": 7735
+    },
+    {
+      "epoch": 1.4870317002881843,
+      "grad_norm": 0.3344769775867462,
+      "learning_rate": 5.0802147858357386e-05,
+      "loss": 3.1695476531982423,
+      "step": 7740
+    },
+    {
+      "epoch": 1.4879923150816523,
+      "grad_norm": 0.389249324798584,
+      "learning_rate": 5.0623549978413166e-05,
+      "loss": 3.171805000305176,
+      "step": 7745
+    },
+    {
+      "epoch": 1.48895292987512,
+      "grad_norm": 0.3440057635307312,
+      "learning_rate": 5.0445202835447056e-05,
+      "loss": 3.1670265197753906,
+      "step": 7750
+    },
+    {
+      "epoch": 1.4899135446685878,
+      "grad_norm": 0.3163948059082031,
+      "learning_rate": 5.026710687944728e-05,
+      "loss": 3.1670303344726562,
+      "step": 7755
+    },
+    {
+      "epoch": 1.4908741594620558,
+      "grad_norm": 0.3952416777610779,
+      "learning_rate": 5.0089262559768246e-05,
+      "loss": 3.171236038208008,
+      "step": 7760
+    },
+    {
+      "epoch": 1.4918347742555236,
+      "grad_norm": 0.3361209034919739,
+      "learning_rate": 4.9911670325129304e-05,
+      "loss": 3.1686822891235353,
+      "step": 7765
+    },
+    {
+      "epoch": 1.4927953890489913,
+      "grad_norm": 0.35905590653419495,
+      "learning_rate": 4.9734330623613924e-05,
+      "loss": 3.1694427490234376,
+      "step": 7770
+    },
+    {
+      "epoch": 1.493756003842459,
+      "grad_norm": 0.4923272132873535,
+      "learning_rate": 4.955724390266841e-05,
+      "loss": 3.171158218383789,
+      "step": 7775
+    },
+    {
+      "epoch": 1.4947166186359269,
+      "grad_norm": 0.3285467326641083,
+      "learning_rate": 4.9380410609100674e-05,
+      "loss": 3.1719661712646485,
+      "step": 7780
+    },
+    {
+      "epoch": 1.4956772334293948,
+      "grad_norm": 0.35682541131973267,
+      "learning_rate": 4.920383118907929e-05,
+      "loss": 3.1671146392822265,
+      "step": 7785
+    },
+    {
+      "epoch": 1.4966378482228626,
+      "grad_norm": 0.5638254880905151,
+      "learning_rate": 4.902750608813222e-05,
+      "loss": 3.1699798583984373,
+      "step": 7790
+    },
+    {
+      "epoch": 1.4975984630163304,
+      "grad_norm": 0.2942904531955719,
+      "learning_rate": 4.885143575114587e-05,
+      "loss": 3.1689422607421873,
+      "step": 7795
+    },
+    {
+      "epoch": 1.4985590778097984,
+      "grad_norm": 0.3493422269821167,
+      "learning_rate": 4.8675620622363645e-05,
+      "loss": 3.168619918823242,
+      "step": 7800
+    },
+    {
+      "epoch": 1.4995196926032661,
+      "grad_norm": 0.40712451934814453,
+      "learning_rate": 4.850006114538519e-05,
+      "loss": 3.169388771057129,
+      "step": 7805
+    },
+    {
+      "epoch": 1.5004803073967339,
+      "grad_norm": 0.31108608841896057,
+      "learning_rate": 4.8324757763165075e-05,
+      "loss": 3.1686771392822264,
+      "step": 7810
+    },
+    {
+      "epoch": 1.5014409221902016,
+      "grad_norm": 0.5020172595977783,
+      "learning_rate": 4.814971091801179e-05,
+      "loss": 3.1644512176513673,
+      "step": 7815
+    },
+    {
+      "epoch": 1.5024015369836694,
+      "grad_norm": 0.4801461696624756,
+      "learning_rate": 4.7974921051586385e-05,
+      "loss": 3.1699331283569334,
+      "step": 7820
+    },
+    {
+      "epoch": 1.5033621517771374,
+      "grad_norm": 0.31796136498451233,
+      "learning_rate": 4.780038860490164e-05,
+      "loss": 3.1699861526489257,
+      "step": 7825
+    },
+    {
+      "epoch": 1.5043227665706052,
+      "grad_norm": 0.3169126510620117,
+      "learning_rate": 4.762611401832089e-05,
+      "loss": 3.1705883026123045,
+      "step": 7830
+    },
+    {
+      "epoch": 1.5052833813640731,
+      "grad_norm": 0.3533933758735657,
+      "learning_rate": 4.745209773155671e-05,
+      "loss": 3.165675926208496,
+      "step": 7835
+    },
+    {
+      "epoch": 1.506243996157541,
+      "grad_norm": 0.4423210024833679,
+      "learning_rate": 4.727834018367007e-05,
+      "loss": 3.1672183990478517,
+      "step": 7840
+    },
+    {
+      "epoch": 1.5072046109510087,
+      "grad_norm": 0.33951136469841003,
+      "learning_rate": 4.710484181306912e-05,
+      "loss": 3.169887733459473,
+      "step": 7845
+    },
+    {
+      "epoch": 1.5081652257444764,
+      "grad_norm": 0.298172265291214,
+      "learning_rate": 4.693160305750801e-05,
+      "loss": 3.1656867980957033,
+      "step": 7850
+    },
+    {
+      "epoch": 1.5091258405379442,
+      "grad_norm": 0.3188192546367645,
+      "learning_rate": 4.675862435408591e-05,
+      "loss": 3.1662307739257813,
+      "step": 7855
+    },
+    {
+      "epoch": 1.510086455331412,
+      "grad_norm": 0.3231668174266815,
+      "learning_rate": 4.6585906139245834e-05,
+      "loss": 3.1684810638427736,
+      "step": 7860
+    },
+    {
+      "epoch": 1.51104707012488,
+      "grad_norm": 0.46347707509994507,
+      "learning_rate": 4.641344884877362e-05,
+      "loss": 3.1662145614624024,
+      "step": 7865
+    },
+    {
+      "epoch": 1.5120076849183477,
+      "grad_norm": 0.4303688704967499,
+      "learning_rate": 4.6241252917796576e-05,
+      "loss": 3.166950798034668,
+      "step": 7870
+    },
+    {
+      "epoch": 1.5129682997118157,
+      "grad_norm": 0.3407163918018341,
+      "learning_rate": 4.6069318780782765e-05,
+      "loss": 3.1677284240722656,
+      "step": 7875
+    },
+    {
+      "epoch": 1.5139289145052834,
+      "grad_norm": 0.2888704240322113,
+      "learning_rate": 4.589764687153967e-05,
+      "loss": 3.162090301513672,
+      "step": 7880
+    },
+    {
+      "epoch": 1.5148895292987512,
+      "grad_norm": 0.3005317747592926,
+      "learning_rate": 4.5726237623213155e-05,
+      "loss": 3.1687442779541017,
+      "step": 7885
+    },
+    {
+      "epoch": 1.515850144092219,
+      "grad_norm": 0.2978266477584839,
+      "learning_rate": 4.555509146828624e-05,
+      "loss": 3.1686517715454103,
+      "step": 7890
+    },
+    {
+      "epoch": 1.5168107588856867,
+      "grad_norm": 0.2926294505596161,
+      "learning_rate": 4.53842088385783e-05,
+      "loss": 3.170303726196289,
+      "step": 7895
+    },
+    {
+      "epoch": 1.5177713736791547,
+      "grad_norm": 0.47925063967704773,
+      "learning_rate": 4.521359016524376e-05,
+      "loss": 3.169039726257324,
+      "step": 7900
+    },
+    {
+      "epoch": 1.5187319884726225,
+      "grad_norm": 0.4029441475868225,
+      "learning_rate": 4.5043235878770965e-05,
+      "loss": 3.166218376159668,
+      "step": 7905
+    },
+    {
+      "epoch": 1.5196926032660905,
+      "grad_norm": 0.31125885248184204,
+      "learning_rate": 4.4873146408981295e-05,
+      "loss": 3.1654186248779297,
+      "step": 7910
+    },
+    {
+      "epoch": 1.5206532180595582,
+      "grad_norm": 0.3251868784427643,
+      "learning_rate": 4.4703322185027926e-05,
+      "loss": 3.1693817138671876,
+      "step": 7915
+    },
+    {
+      "epoch": 1.521613832853026,
+      "grad_norm": 0.31480100750923157,
+      "learning_rate": 4.453376363539481e-05,
+      "loss": 3.1616186141967773,
+      "step": 7920
+    },
+    {
+      "epoch": 1.5225744476464937,
+      "grad_norm": 0.3208870589733124,
+      "learning_rate": 4.436447118789555e-05,
+      "loss": 3.168798637390137,
+      "step": 7925
+    },
+    {
+      "epoch": 1.5235350624399615,
+      "grad_norm": 0.3242574632167816,
+      "learning_rate": 4.419544526967238e-05,
+      "loss": 3.171648406982422,
+      "step": 7930
+    },
+    {
+      "epoch": 1.5244956772334293,
+      "grad_norm": 0.3922894597053528,
+      "learning_rate": 4.402668630719504e-05,
+      "loss": 3.1695817947387694,
+      "step": 7935
+    },
+    {
+      "epoch": 1.5254562920268973,
+      "grad_norm": 0.3969297409057617,
+      "learning_rate": 4.385819472625963e-05,
+      "loss": 3.1659820556640623,
+      "step": 7940
+    },
+    {
+      "epoch": 1.526416906820365,
+      "grad_norm": 0.3534747362136841,
+      "learning_rate": 4.368997095198775e-05,
+      "loss": 3.164035415649414,
+      "step": 7945
+    },
+    {
+      "epoch": 1.527377521613833,
+      "grad_norm": 0.387599915266037,
+      "learning_rate": 4.352201540882523e-05,
+      "loss": 3.166557502746582,
+      "step": 7950
+    },
+    {
+      "epoch": 1.5283381364073008,
+      "grad_norm": 0.3567873537540436,
+      "learning_rate": 4.33543285205412e-05,
+      "loss": 3.1643226623535154,
+      "step": 7955
+    },
+    {
+      "epoch": 1.5292987512007685,
+      "grad_norm": 0.3381181061267853,
+      "learning_rate": 4.318691071022676e-05,
+      "loss": 3.1638862609863283,
+      "step": 7960
+    },
+    {
+      "epoch": 1.5302593659942363,
+      "grad_norm": 0.2984130084514618,
+      "learning_rate": 4.301976240029428e-05,
+      "loss": 3.1657569885253904,
+      "step": 7965
+    },
+    {
+      "epoch": 1.531219980787704,
+      "grad_norm": 0.2934906780719757,
+      "learning_rate": 4.285288401247614e-05,
+      "loss": 3.165040969848633,
+      "step": 7970
+    },
+    {
+      "epoch": 1.5321805955811718,
+      "grad_norm": 0.2868797779083252,
+      "learning_rate": 4.268627596782354e-05,
+      "loss": 3.166617202758789,
+      "step": 7975
+    },
+    {
+      "epoch": 1.5331412103746398,
+      "grad_norm": 0.40537068247795105,
+      "learning_rate": 4.251993868670569e-05,
+      "loss": 3.1657033920288087,
+      "step": 7980
+    },
+    {
+      "epoch": 1.5341018251681076,
+      "grad_norm": 0.3606870770454407,
+      "learning_rate": 4.235387258880871e-05,
+      "loss": 3.1681283950805663,
+      "step": 7985
+    },
+    {
+      "epoch": 1.5350624399615755,
+      "grad_norm": 0.3049268126487732,
+      "learning_rate": 4.218807809313428e-05,
+      "loss": 3.165813446044922,
+      "step": 7990
+    },
+    {
+      "epoch": 1.5360230547550433,
+      "grad_norm": 0.3047685921192169,
+      "learning_rate": 4.202255561799897e-05,
+      "loss": 3.1700771331787108,
+      "step": 7995
+    },
+    {
+      "epoch": 1.536983669548511,
+      "grad_norm": 0.32123667001724243,
+      "learning_rate": 4.1857305581032974e-05,
+      "loss": 3.16713924407959,
+      "step": 8000
+    },
+    {
+      "epoch": 1.5379442843419788,
+      "grad_norm": 0.31456029415130615,
+      "learning_rate": 4.1692328399179134e-05,
+      "loss": 3.1665233612060546,
+      "step": 8005
+    },
+    {
+      "epoch": 1.5389048991354466,
+      "grad_norm": 0.2823289632797241,
+      "learning_rate": 4.1527624488691706e-05,
+      "loss": 3.165792465209961,
+      "step": 8010
+    },
+    {
+      "epoch": 1.5398655139289144,
+      "grad_norm": 0.41505372524261475,
+      "learning_rate": 4.1363194265135584e-05,
+      "loss": 3.165696907043457,
+      "step": 8015
+    },
+    {
+      "epoch": 1.5408261287223823,
+      "grad_norm": 0.3873119056224823,
+      "learning_rate": 4.1199038143385114e-05,
+      "loss": 3.1616024017333983,
+      "step": 8020
+    },
+    {
+      "epoch": 1.54178674351585,
+      "grad_norm": 0.29045772552490234,
+      "learning_rate": 4.1035156537623056e-05,
+      "loss": 3.163679504394531,
+      "step": 8025
+    },
+    {
+      "epoch": 1.542747358309318,
+      "grad_norm": 0.31360870599746704,
+      "learning_rate": 4.087154986133944e-05,
+      "loss": 3.1689876556396483,
+      "step": 8030
+    },
+    {
+      "epoch": 1.5437079731027858,
+      "grad_norm": 0.34343641996383667,
+      "learning_rate": 4.070821852733074e-05,
+      "loss": 3.1644660949707033,
+      "step": 8035
+    },
+    {
+      "epoch": 1.5446685878962536,
+      "grad_norm": 0.3275468647480011,
+      "learning_rate": 4.054516294769871e-05,
+      "loss": 3.161859321594238,
+      "step": 8040
+    },
+    {
+      "epoch": 1.5456292026897214,
+      "grad_norm": 0.3492441773414612,
+      "learning_rate": 4.038238353384919e-05,
+      "loss": 3.166157531738281,
+      "step": 8045
+    },
+    {
+      "epoch": 1.5465898174831891,
+      "grad_norm": 0.3080005645751953,
+      "learning_rate": 4.021988069649138e-05,
+      "loss": 3.169676399230957,
+      "step": 8050
+    },
+    {
+      "epoch": 1.547550432276657,
+      "grad_norm": 0.2594451904296875,
+      "learning_rate": 4.0057654845636714e-05,
+      "loss": 3.170206642150879,
+      "step": 8055
+    },
+    {
+      "epoch": 1.5485110470701249,
+      "grad_norm": 0.2814268469810486,
+      "learning_rate": 3.989570639059753e-05,
+      "loss": 3.164537811279297,
+      "step": 8060
+    },
+    {
+      "epoch": 1.5494716618635929,
+      "grad_norm": 0.3419535756111145,
+      "learning_rate": 3.973403573998647e-05,
+      "loss": 3.169506645202637,
+      "step": 8065
+    },
+    {
+      "epoch": 1.5504322766570606,
+      "grad_norm": 0.3430168032646179,
+      "learning_rate": 3.957264330171511e-05,
+      "loss": 3.1708099365234377,
+      "step": 8070
+    },
+    {
+      "epoch": 1.5513928914505284,
+      "grad_norm": 0.3074200749397278,
+      "learning_rate": 3.941152948299321e-05,
+      "loss": 3.1663097381591796,
+      "step": 8075
+    },
+    {
+      "epoch": 1.5523535062439962,
+      "grad_norm": 0.29986321926116943,
+      "learning_rate": 3.9250694690327383e-05,
+      "loss": 3.1678709030151366,
+      "step": 8080
+    },
+    {
+      "epoch": 1.553314121037464,
+      "grad_norm": 0.305226594209671,
+      "learning_rate": 3.9090139329520327e-05,
+      "loss": 3.163295364379883,
+      "step": 8085
+    },
+    {
+      "epoch": 1.5542747358309317,
+      "grad_norm": 0.27455848455429077,
+      "learning_rate": 3.8929863805669706e-05,
+      "loss": 3.163608741760254,
+      "step": 8090
+    },
+    {
+      "epoch": 1.5552353506243997,
+      "grad_norm": 0.30126896500587463,
+      "learning_rate": 3.876986852316715e-05,
+      "loss": 3.1656238555908205,
+      "step": 8095
+    },
+    {
+      "epoch": 1.5561959654178674,
+      "grad_norm": 0.3914654552936554,
+      "learning_rate": 3.861015388569709e-05,
+      "loss": 3.1641334533691405,
+      "step": 8100
+    },
+    {
+      "epoch": 1.5571565802113354,
+      "grad_norm": 0.3612956404685974,
+      "learning_rate": 3.845072029623598e-05,
+      "loss": 3.1632358551025392,
+      "step": 8105
+    },
+    {
+      "epoch": 1.5581171950048032,
+      "grad_norm": 0.2885013222694397,
+      "learning_rate": 3.8291568157051154e-05,
+      "loss": 3.1646907806396483,
+      "step": 8110
+    },
+    {
+      "epoch": 1.559077809798271,
+      "grad_norm": 0.29719772934913635,
+      "learning_rate": 3.8132697869699705e-05,
+      "loss": 3.1659112930297852,
+      "step": 8115
+    },
+    {
+      "epoch": 1.5600384245917387,
+      "grad_norm": 0.308212012052536,
+      "learning_rate": 3.797410983502766e-05,
+      "loss": 3.1631364822387695,
+      "step": 8120
+    },
+    {
+      "epoch": 1.5609990393852065,
+      "grad_norm": 0.30908966064453125,
+      "learning_rate": 3.781580445316906e-05,
+      "loss": 3.1642078399658202,
+      "step": 8125
+    },
+    {
+      "epoch": 1.5619596541786742,
+      "grad_norm": 0.3072703182697296,
+      "learning_rate": 3.765778212354445e-05,
+      "loss": 3.1633129119873047,
+      "step": 8130
+    },
+    {
+      "epoch": 1.5629202689721422,
+      "grad_norm": 0.3363180458545685,
+      "learning_rate": 3.7500043244860475e-05,
+      "loss": 3.16600341796875,
+      "step": 8135
+    },
+    {
+      "epoch": 1.56388088376561,
+      "grad_norm": 0.41240906715393066,
+      "learning_rate": 3.7342588215108446e-05,
+      "loss": 3.168244743347168,
+      "step": 8140
+    },
+    {
+      "epoch": 1.564841498559078,
+      "grad_norm": 0.3254190683364868,
+      "learning_rate": 3.7185417431563644e-05,
+      "loss": 3.1647628784179687,
+      "step": 8145
+    },
+    {
+      "epoch": 1.5658021133525457,
+      "grad_norm": 0.3488558232784271,
+      "learning_rate": 3.702853129078398e-05,
+      "loss": 3.1683582305908202,
+      "step": 8150
+    },
+    {
+      "epoch": 1.5667627281460135,
+      "grad_norm": 0.34358084201812744,
+      "learning_rate": 3.6871930188609325e-05,
+      "loss": 3.1676990509033205,
+      "step": 8155
+    },
+    {
+      "epoch": 1.5677233429394812,
+      "grad_norm": 0.3370071053504944,
+      "learning_rate": 3.671561452016033e-05,
+      "loss": 3.1652057647705076,
+      "step": 8160
+    },
+    {
+      "epoch": 1.568683957732949,
+      "grad_norm": 0.23271813988685608,
+      "learning_rate": 3.655958467983749e-05,
+      "loss": 3.164310073852539,
+      "step": 8165
+    },
+    {
+      "epoch": 1.5696445725264168,
+      "grad_norm": 0.33882594108581543,
+      "learning_rate": 3.6403841061320026e-05,
+      "loss": 3.1641658782958983,
+      "step": 8170
+    },
+    {
+      "epoch": 1.5706051873198847,
+      "grad_norm": 0.25975945591926575,
+      "learning_rate": 3.6248384057565104e-05,
+      "loss": 3.1639142990112306,
+      "step": 8175
+    },
+    {
+      "epoch": 1.5715658021133525,
+      "grad_norm": 0.3010103404521942,
+      "learning_rate": 3.6093214060806686e-05,
+      "loss": 3.166419792175293,
+      "step": 8180
+    },
+    {
+      "epoch": 1.5725264169068205,
+      "grad_norm": 0.26971593499183655,
+      "learning_rate": 3.593833146255461e-05,
+      "loss": 3.1657012939453124,
+      "step": 8185
+    },
+    {
+      "epoch": 1.5734870317002883,
+      "grad_norm": 0.2859204411506653,
+      "learning_rate": 3.5783736653593546e-05,
+      "loss": 3.163465118408203,
+      "step": 8190
+    },
+    {
+      "epoch": 1.574447646493756,
+      "grad_norm": 0.2561207115650177,
+      "learning_rate": 3.56294300239821e-05,
+      "loss": 3.164503288269043,
+      "step": 8195
+    },
+    {
+      "epoch": 1.5754082612872238,
+      "grad_norm": 0.22760829329490662,
+      "learning_rate": 3.547541196305166e-05,
+      "loss": 3.1677932739257812,
+      "step": 8200
+    },
+    {
+      "epoch": 1.5763688760806915,
+      "grad_norm": 0.3261941969394684,
+      "learning_rate": 3.53216828594056e-05,
+      "loss": 3.164171600341797,
+      "step": 8205
+    },
+    {
+      "epoch": 1.5773294908741593,
+      "grad_norm": 0.3237496018409729,
+      "learning_rate": 3.5168243100918254e-05,
+      "loss": 3.1644121170043946,
+      "step": 8210
+    },
+    {
+      "epoch": 1.5782901056676273,
+      "grad_norm": 0.308450311422348,
+      "learning_rate": 3.501509307473391e-05,
+      "loss": 3.167286682128906,
+      "step": 8215
+    },
+    {
+      "epoch": 1.579250720461095,
+      "grad_norm": 0.2737598717212677,
+      "learning_rate": 3.486223316726569e-05,
+      "loss": 3.159839630126953,
+      "step": 8220
+    },
+    {
+      "epoch": 1.580211335254563,
+      "grad_norm": 0.2531875967979431,
+      "learning_rate": 3.470966376419489e-05,
+      "loss": 3.165401268005371,
+      "step": 8225
+    },
+    {
+      "epoch": 1.5811719500480308,
+      "grad_norm": 0.3886658847332001,
+      "learning_rate": 3.455738525046976e-05,
+      "loss": 3.164730453491211,
+      "step": 8230
+    },
+    {
+      "epoch": 1.5821325648414986,
+      "grad_norm": 0.32052338123321533,
+      "learning_rate": 3.440539801030463e-05,
+      "loss": 3.1636079788208007,
+      "step": 8235
+    },
+    {
+      "epoch": 1.5830931796349663,
+      "grad_norm": 0.3415239751338959,
+      "learning_rate": 3.425370242717887e-05,
+      "loss": 3.1651962280273436,
+      "step": 8240
+    },
+    {
+      "epoch": 1.584053794428434,
+      "grad_norm": 0.27457112073898315,
+      "learning_rate": 3.410229888383597e-05,
+      "loss": 3.169073486328125,
+      "step": 8245
+    },
+    {
+      "epoch": 1.585014409221902,
+      "grad_norm": 0.3607783615589142,
+      "learning_rate": 3.395118776228265e-05,
+      "loss": 3.1643707275390627,
+      "step": 8250
+    },
+    {
+      "epoch": 1.5859750240153698,
+      "grad_norm": 0.44193530082702637,
+      "learning_rate": 3.380036944378775e-05,
+      "loss": 3.165296936035156,
+      "step": 8255
+    },
+    {
+      "epoch": 1.5869356388088378,
+      "grad_norm": 0.27032527327537537,
+      "learning_rate": 3.364984430888138e-05,
+      "loss": 3.1653570175170898,
+      "step": 8260
+    },
+    {
+      "epoch": 1.5878962536023056,
+      "grad_norm": 0.3761611580848694,
+      "learning_rate": 3.34996127373539e-05,
+      "loss": 3.1637210845947266,
+      "step": 8265
+    },
+    {
+      "epoch": 1.5888568683957733,
+      "grad_norm": 0.2229749858379364,
+      "learning_rate": 3.3349675108254946e-05,
+      "loss": 3.165303421020508,
+      "step": 8270
+    },
+    {
+      "epoch": 1.589817483189241,
+      "grad_norm": 0.23744769394397736,
+      "learning_rate": 3.320003179989254e-05,
+      "loss": 3.16616268157959,
+      "step": 8275
+    },
+    {
+      "epoch": 1.5907780979827089,
+      "grad_norm": 0.28989988565444946,
+      "learning_rate": 3.305068318983211e-05,
+      "loss": 3.1658939361572265,
+      "step": 8280
+    },
+    {
+      "epoch": 1.5917387127761766,
+      "grad_norm": 0.2983841300010681,
+      "learning_rate": 3.290162965489558e-05,
+      "loss": 3.1631855010986327,
+      "step": 8285
+    },
+    {
+      "epoch": 1.5926993275696446,
+      "grad_norm": 0.2570306658744812,
+      "learning_rate": 3.275287157116021e-05,
+      "loss": 3.1675746917724608,
+      "step": 8290
+    },
+    {
+      "epoch": 1.5936599423631124,
+      "grad_norm": 0.3952568471431732,
+      "learning_rate": 3.260440931395794e-05,
+      "loss": 3.1651771545410154,
+      "step": 8295
+    },
+    {
+      "epoch": 1.5946205571565804,
+      "grad_norm": 0.3825446367263794,
+      "learning_rate": 3.24562432578743e-05,
+      "loss": 3.1654695510864257,
+      "step": 8300
+    },
+    {
+      "epoch": 1.5955811719500481,
+      "grad_norm": 0.2898649275302887,
+      "learning_rate": 3.230837377674746e-05,
+      "loss": 3.1621742248535156,
+      "step": 8305
+    },
+    {
+      "epoch": 1.5965417867435159,
+      "grad_norm": 0.24470216035842896,
+      "learning_rate": 3.216080124366724e-05,
+      "loss": 3.1656517028808593,
+      "step": 8310
+    },
+    {
+      "epoch": 1.5975024015369836,
+      "grad_norm": 0.23829184472560883,
+      "learning_rate": 3.20135260309743e-05,
+      "loss": 3.166254425048828,
+      "step": 8315
+    },
+    {
+      "epoch": 1.5984630163304514,
+      "grad_norm": 0.2703918218612671,
+      "learning_rate": 3.186654851025911e-05,
+      "loss": 3.1674873352050783,
+      "step": 8320
+    },
+    {
+      "epoch": 1.5994236311239192,
+      "grad_norm": 0.2798570394515991,
+      "learning_rate": 3.171986905236104e-05,
+      "loss": 3.16519718170166,
+      "step": 8325
+    },
+    {
+      "epoch": 1.6003842459173871,
+      "grad_norm": 0.29663723707199097,
+      "learning_rate": 3.15734880273674e-05,
+      "loss": 3.1651836395263673,
+      "step": 8330
+    },
+    {
+      "epoch": 1.601344860710855,
+      "grad_norm": 0.2850891351699829,
+      "learning_rate": 3.142740580461261e-05,
+      "loss": 3.1634387969970703,
+      "step": 8335
+    },
+    {
+      "epoch": 1.602305475504323,
+      "grad_norm": 0.3124440610408783,
+      "learning_rate": 3.128162275267697e-05,
+      "loss": 3.1671958923339845,
+      "step": 8340
+    },
+    {
+      "epoch": 1.6032660902977907,
+      "grad_norm": 0.3228466808795929,
+      "learning_rate": 3.113613923938614e-05,
+      "loss": 3.1625482559204103,
+      "step": 8345
+    },
+    {
+      "epoch": 1.6042267050912584,
+      "grad_norm": 0.23934929072856903,
+      "learning_rate": 3.0990955631809965e-05,
+      "loss": 3.1659454345703124,
+      "step": 8350
+    },
+    {
+      "epoch": 1.6051873198847262,
+      "grad_norm": 0.2973260283470154,
+      "learning_rate": 3.08460722962616e-05,
+      "loss": 3.167291259765625,
+      "step": 8355
+    },
+    {
+      "epoch": 1.606147934678194,
+      "grad_norm": 0.26656073331832886,
+      "learning_rate": 3.070148959829649e-05,
+      "loss": 3.165090560913086,
+      "step": 8360
+    },
+    {
+      "epoch": 1.6071085494716617,
+      "grad_norm": 0.25156885385513306,
+      "learning_rate": 3.055720790271164e-05,
+      "loss": 3.1673526763916016,
+      "step": 8365
+    },
+    {
+      "epoch": 1.6080691642651297,
+      "grad_norm": 0.2427086979150772,
+      "learning_rate": 3.0413227573544592e-05,
+      "loss": 3.167298698425293,
+      "step": 8370
+    },
+    {
+      "epoch": 1.6090297790585975,
+      "grad_norm": 0.33012548089027405,
+      "learning_rate": 3.026954897407252e-05,
+      "loss": 3.16772403717041,
+      "step": 8375
+    },
+    {
+      "epoch": 1.6099903938520654,
+      "grad_norm": 0.19765010476112366,
+      "learning_rate": 3.01261724668112e-05,
+      "loss": 3.1641408920288088,
+      "step": 8380
+    },
+    {
+      "epoch": 1.6109510086455332,
+      "grad_norm": 0.25813835859298706,
+      "learning_rate": 2.9983098413514284e-05,
+      "loss": 3.168695068359375,
+      "step": 8385
+    },
+    {
+      "epoch": 1.611911623439001,
+      "grad_norm": 0.2662206292152405,
+      "learning_rate": 2.9840327175172295e-05,
+      "loss": 3.1604537963867188,
+      "step": 8390
+    },
+    {
+      "epoch": 1.6128722382324687,
+      "grad_norm": 0.2579902410507202,
+      "learning_rate": 2.969785911201172e-05,
+      "loss": 3.1639808654785155,
+      "step": 8395
+    },
+    {
+      "epoch": 1.6138328530259365,
+      "grad_norm": 0.27471011877059937,
+      "learning_rate": 2.9555694583494095e-05,
+      "loss": 3.1644439697265625,
+      "step": 8400
+    },
+    {
+      "epoch": 1.6147934678194045,
+      "grad_norm": 0.2601883113384247,
+      "learning_rate": 2.9413833948315163e-05,
+      "loss": 3.1667598724365233,
+      "step": 8405
+    },
+    {
+      "epoch": 1.6157540826128722,
+      "grad_norm": 0.27512237429618835,
+      "learning_rate": 2.9272277564403746e-05,
+      "loss": 3.1641109466552733,
+      "step": 8410
+    },
+    {
+      "epoch": 1.6167146974063402,
+      "grad_norm": 0.25258708000183105,
+      "learning_rate": 2.9131025788921193e-05,
+      "loss": 3.1663232803344727,
+      "step": 8415
+    },
+    {
+      "epoch": 1.617675312199808,
+      "grad_norm": 0.22575309872627258,
+      "learning_rate": 2.8990078978260216e-05,
+      "loss": 3.166961669921875,
+      "step": 8420
+    },
+    {
+      "epoch": 1.6186359269932757,
+      "grad_norm": 0.24490588903427124,
+      "learning_rate": 2.8849437488044118e-05,
+      "loss": 3.1663583755493163,
+      "step": 8425
+    },
+    {
+      "epoch": 1.6195965417867435,
+      "grad_norm": 0.2155570238828659,
+      "learning_rate": 2.8709101673125728e-05,
+      "loss": 3.1656875610351562,
+      "step": 8430
+    },
+    {
+      "epoch": 1.6205571565802113,
+      "grad_norm": 0.2795376479625702,
+      "learning_rate": 2.856907188758674e-05,
+      "loss": 3.162942314147949,
+      "step": 8435
+    },
+    {
+      "epoch": 1.621517771373679,
+      "grad_norm": 0.3358766734600067,
+      "learning_rate": 2.8429348484736658e-05,
+      "loss": 3.1658599853515623,
+      "step": 8440
+    },
+    {
+      "epoch": 1.622478386167147,
+      "grad_norm": 0.2666330933570862,
+      "learning_rate": 2.8289931817112027e-05,
+      "loss": 3.1691190719604494,
+      "step": 8445
+    },
+    {
+      "epoch": 1.6234390009606148,
+      "grad_norm": 0.2384926676750183,
+      "learning_rate": 2.8150822236475283e-05,
+      "loss": 3.1659183502197266,
+      "step": 8450
+    },
+    {
+      "epoch": 1.6243996157540828,
+      "grad_norm": 0.25563499331474304,
+      "learning_rate": 2.8012020093814252e-05,
+      "loss": 3.1584651947021483,
+      "step": 8455
+    },
+    {
+      "epoch": 1.6253602305475505,
+      "grad_norm": 0.2299119532108307,
+      "learning_rate": 2.7873525739340973e-05,
+      "loss": 3.1645713806152345,
+      "step": 8460
+    },
+    {
+      "epoch": 1.6263208453410183,
+      "grad_norm": 0.26834285259246826,
+      "learning_rate": 2.773533952249088e-05,
+      "loss": 3.166657257080078,
+      "step": 8465
+    },
+    {
+      "epoch": 1.627281460134486,
+      "grad_norm": 0.2765657901763916,
+      "learning_rate": 2.7597461791921987e-05,
+      "loss": 3.165719223022461,
+      "step": 8470
+    },
+    {
+      "epoch": 1.6282420749279538,
+      "grad_norm": 0.26235702633857727,
+      "learning_rate": 2.745989289551401e-05,
+      "loss": 3.163796615600586,
+      "step": 8475
+    },
+    {
+      "epoch": 1.6292026897214216,
+      "grad_norm": 0.2585254907608032,
+      "learning_rate": 2.7322633180367286e-05,
+      "loss": 3.1652366638183596,
+      "step": 8480
+    },
+    {
+      "epoch": 1.6301633045148896,
+      "grad_norm": 0.24528291821479797,
+      "learning_rate": 2.7185682992802215e-05,
+      "loss": 3.165533447265625,
+      "step": 8485
+    },
+    {
+      "epoch": 1.6311239193083573,
+      "grad_norm": 0.24676673114299774,
+      "learning_rate": 2.7049042678358157e-05,
+      "loss": 3.162363624572754,
+      "step": 8490
+    },
+    {
+      "epoch": 1.6320845341018253,
+      "grad_norm": 0.28905969858169556,
+      "learning_rate": 2.6912712581792684e-05,
+      "loss": 3.165178680419922,
+      "step": 8495
+    },
+    {
+      "epoch": 1.633045148895293,
+      "grad_norm": 0.25018957257270813,
+      "learning_rate": 2.6776693047080546e-05,
+      "loss": 3.163674736022949,
+      "step": 8500
+    },
+    {
+      "epoch": 1.6340057636887608,
+      "grad_norm": 0.24032709002494812,
+      "learning_rate": 2.6640984417412996e-05,
+      "loss": 3.1674957275390625,
+      "step": 8505
+    },
+    {
+      "epoch": 1.6349663784822286,
+      "grad_norm": 0.2825995981693268,
+      "learning_rate": 2.6505587035196862e-05,
+      "loss": 3.164845085144043,
+      "step": 8510
+    },
+    {
+      "epoch": 1.6359269932756964,
+      "grad_norm": 0.3082216680049896,
+      "learning_rate": 2.6370501242053655e-05,
+      "loss": 3.164422607421875,
+      "step": 8515
+    },
+    {
+      "epoch": 1.6368876080691641,
+      "grad_norm": 0.2581336796283722,
+      "learning_rate": 2.6235727378818617e-05,
+      "loss": 3.1643039703369142,
+      "step": 8520
+    },
+    {
+      "epoch": 1.637848222862632,
+      "grad_norm": 0.2364456057548523,
+      "learning_rate": 2.6101265785540054e-05,
+      "loss": 3.1647056579589843,
+      "step": 8525
+    },
+    {
+      "epoch": 1.6388088376560999,
+      "grad_norm": 0.2648424804210663,
+      "learning_rate": 2.596711680147837e-05,
+      "loss": 3.1666595458984377,
+      "step": 8530
+    },
+    {
+      "epoch": 1.6397694524495678,
+      "grad_norm": 0.21114224195480347,
+      "learning_rate": 2.5833280765105218e-05,
+      "loss": 3.1652767181396486,
+      "step": 8535
+    },
+    {
+      "epoch": 1.6407300672430356,
+      "grad_norm": 0.29811277985572815,
+      "learning_rate": 2.5699758014102627e-05,
+      "loss": 3.1638210296630858,
+      "step": 8540
+    },
+    {
+      "epoch": 1.6416906820365034,
+      "grad_norm": 0.22697743773460388,
+      "learning_rate": 2.5566548885362233e-05,
+      "loss": 3.1660308837890625,
+      "step": 8545
+    },
+    {
+      "epoch": 1.6426512968299711,
+      "grad_norm": 0.2113056480884552,
+      "learning_rate": 2.543365371498434e-05,
+      "loss": 3.1632022857666016,
+      "step": 8550
+    },
+    {
+      "epoch": 1.643611911623439,
+      "grad_norm": 0.24464410543441772,
+      "learning_rate": 2.5301072838277026e-05,
+      "loss": 3.1658231735229494,
+      "step": 8555
+    },
+    {
+      "epoch": 1.6445725264169067,
+      "grad_norm": 0.23001989722251892,
+      "learning_rate": 2.5168806589755497e-05,
+      "loss": 3.163945960998535,
+      "step": 8560
+    },
+    {
+      "epoch": 1.6455331412103746,
+      "grad_norm": 0.20726712048053741,
+      "learning_rate": 2.503685530314109e-05,
+      "loss": 3.162770080566406,
+      "step": 8565
+    },
+    {
+      "epoch": 1.6464937560038426,
+      "grad_norm": 0.23915457725524902,
+      "learning_rate": 2.490521931136036e-05,
+      "loss": 3.1642734527587892,
+      "step": 8570
+    },
+    {
+      "epoch": 1.6474543707973104,
+      "grad_norm": 0.2551625370979309,
+      "learning_rate": 2.4773898946544473e-05,
+      "loss": 3.1642024993896483,
+      "step": 8575
+    },
+    {
+      "epoch": 1.6484149855907781,
+      "grad_norm": 0.2408856302499771,
+      "learning_rate": 2.4642894540028164e-05,
+      "loss": 3.1599807739257812,
+      "step": 8580
+    },
+    {
+      "epoch": 1.649375600384246,
+      "grad_norm": 0.2513860762119293,
+      "learning_rate": 2.4512206422349024e-05,
+      "loss": 3.163848876953125,
+      "step": 8585
+    },
+    {
+      "epoch": 1.6503362151777137,
+      "grad_norm": 0.21447288990020752,
+      "learning_rate": 2.438183492324654e-05,
+      "loss": 3.1619583129882813,
+      "step": 8590
+    },
+    {
+      "epoch": 1.6512968299711814,
+      "grad_norm": 0.1881914734840393,
+      "learning_rate": 2.4251780371661373e-05,
+      "loss": 3.1636112213134764,
+      "step": 8595
+    },
+    {
+      "epoch": 1.6522574447646494,
+      "grad_norm": 0.19982366263866425,
+      "learning_rate": 2.4122043095734518e-05,
+      "loss": 3.1659828186035157,
+      "step": 8600
+    },
+    {
+      "epoch": 1.6532180595581172,
+      "grad_norm": 0.2119777649641037,
+      "learning_rate": 2.3992623422806444e-05,
+      "loss": 3.161235809326172,
+      "step": 8605
+    },
+    {
+      "epoch": 1.6541786743515852,
+      "grad_norm": 0.2393619865179062,
+      "learning_rate": 2.3863521679416237e-05,
+      "loss": 3.1633071899414062,
+      "step": 8610
+    },
+    {
+      "epoch": 1.655139289145053,
+      "grad_norm": 0.24367739260196686,
+      "learning_rate": 2.3734738191300862e-05,
+      "loss": 3.16229133605957,
+      "step": 8615
+    },
+    {
+      "epoch": 1.6560999039385207,
+      "grad_norm": 0.29672330617904663,
+      "learning_rate": 2.36062732833943e-05,
+      "loss": 3.1618356704711914,
+      "step": 8620
+    },
+    {
+      "epoch": 1.6570605187319885,
+      "grad_norm": 0.24469807744026184,
+      "learning_rate": 2.347812727982661e-05,
+      "loss": 3.161136245727539,
+      "step": 8625
+    },
+    {
+      "epoch": 1.6580211335254562,
+      "grad_norm": 0.21228572726249695,
+      "learning_rate": 2.3350300503923352e-05,
+      "loss": 3.1614845275878904,
+      "step": 8630
+    },
+    {
+      "epoch": 1.658981748318924,
+      "grad_norm": 0.20836694538593292,
+      "learning_rate": 2.3222793278204626e-05,
+      "loss": 3.1641204833984373,
+      "step": 8635
+    },
+    {
+      "epoch": 1.659942363112392,
+      "grad_norm": 0.2582319974899292,
+      "learning_rate": 2.309560592438417e-05,
+      "loss": 3.1627979278564453,
+      "step": 8640
+    },
+    {
+      "epoch": 1.6609029779058597,
+      "grad_norm": 0.23718759417533875,
+      "learning_rate": 2.2968738763368765e-05,
+      "loss": 3.1604846954345702,
+      "step": 8645
+    },
+    {
+      "epoch": 1.6618635926993277,
+      "grad_norm": 0.19198426604270935,
+      "learning_rate": 2.2842192115257295e-05,
+      "loss": 3.1641172409057616,
+      "step": 8650
+    },
+    {
+      "epoch": 1.6628242074927955,
+      "grad_norm": 0.258806049823761,
+      "learning_rate": 2.271596629933992e-05,
+      "loss": 3.16113224029541,
+      "step": 8655
+    },
+    {
+      "epoch": 1.6637848222862632,
+      "grad_norm": 0.21029417216777802,
+      "learning_rate": 2.25900616340973e-05,
+      "loss": 3.1632869720458983,
+      "step": 8660
+    },
+    {
+      "epoch": 1.664745437079731,
+      "grad_norm": 0.21884505450725555,
+      "learning_rate": 2.2464478437199808e-05,
+      "loss": 3.163255310058594,
+      "step": 8665
+    },
+    {
+      "epoch": 1.6657060518731988,
+      "grad_norm": 0.22930586338043213,
+      "learning_rate": 2.2339217025506813e-05,
+      "loss": 3.163426399230957,
+      "step": 8670
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.2358139008283615,
+      "learning_rate": 2.2214277715065636e-05,
+      "loss": 3.1608362197875977,
+      "step": 8675
+    },
+    {
+      "epoch": 1.6676272814601345,
+      "grad_norm": 0.22611092031002045,
+      "learning_rate": 2.2089660821110988e-05,
+      "loss": 3.1634220123291015,
+      "step": 8680
+    },
+    {
+      "epoch": 1.6685878962536023,
+      "grad_norm": 0.19509707391262054,
+      "learning_rate": 2.1965366658064086e-05,
+      "loss": 3.158903884887695,
+      "step": 8685
+    },
+    {
+      "epoch": 1.6695485110470702,
+      "grad_norm": 0.19261038303375244,
+      "learning_rate": 2.1841395539531893e-05,
+      "loss": 3.161344528198242,
+      "step": 8690
+    },
+    {
+      "epoch": 1.670509125840538,
+      "grad_norm": 0.1918260157108307,
+      "learning_rate": 2.1717747778306177e-05,
+      "loss": 3.1661489486694334,
+      "step": 8695
+    },
+    {
+      "epoch": 1.6714697406340058,
+      "grad_norm": 0.20543381571769714,
+      "learning_rate": 2.159442368636297e-05,
+      "loss": 3.1635005950927733,
+      "step": 8700
+    },
+    {
+      "epoch": 1.6724303554274735,
+      "grad_norm": 0.2433311641216278,
+      "learning_rate": 2.147142357486164e-05,
+      "loss": 3.163587951660156,
+      "step": 8705
+    },
+    {
+      "epoch": 1.6733909702209413,
+      "grad_norm": 0.21309548616409302,
+      "learning_rate": 2.1348747754144004e-05,
+      "loss": 3.1650224685668946,
+      "step": 8710
+    },
+    {
+      "epoch": 1.674351585014409,
+      "grad_norm": 0.2069951891899109,
+      "learning_rate": 2.1226396533733796e-05,
+      "loss": 3.164658546447754,
+      "step": 8715
+    },
+    {
+      "epoch": 1.675312199807877,
+      "grad_norm": 0.24692699313163757,
+      "learning_rate": 2.1104370222335688e-05,
+      "loss": 3.1658843994140624,
+      "step": 8720
+    },
+    {
+      "epoch": 1.6762728146013448,
+      "grad_norm": 0.23039954900741577,
+      "learning_rate": 2.0982669127834622e-05,
+      "loss": 3.1638277053833006,
+      "step": 8725
+    },
+    {
+      "epoch": 1.6772334293948128,
+      "grad_norm": 0.2396543025970459,
+      "learning_rate": 2.0861293557294862e-05,
+      "loss": 3.162588119506836,
+      "step": 8730
+    },
+    {
+      "epoch": 1.6781940441882806,
+      "grad_norm": 0.2332344651222229,
+      "learning_rate": 2.0740243816959452e-05,
+      "loss": 3.163272476196289,
+      "step": 8735
+    },
+    {
+      "epoch": 1.6791546589817483,
+      "grad_norm": 0.21478639543056488,
+      "learning_rate": 2.061952021224938e-05,
+      "loss": 3.162495803833008,
+      "step": 8740
+    },
+    {
+      "epoch": 1.680115273775216,
+      "grad_norm": 0.19731956720352173,
+      "learning_rate": 2.0499123047762576e-05,
+      "loss": 3.162389945983887,
+      "step": 8745
+    },
+    {
+      "epoch": 1.6810758885686838,
+      "grad_norm": 0.19572487473487854,
+      "learning_rate": 2.03790526272735e-05,
+      "loss": 3.1616092681884767,
+      "step": 8750
+    },
+    {
+      "epoch": 1.6820365033621518,
+      "grad_norm": 0.184941828250885,
+      "learning_rate": 2.025930925373213e-05,
+      "loss": 3.1640493392944338,
+      "step": 8755
+    },
+    {
+      "epoch": 1.6829971181556196,
+      "grad_norm": 0.1868741363286972,
+      "learning_rate": 2.013989322926331e-05,
+      "loss": 3.164847183227539,
+      "step": 8760
+    },
+    {
+      "epoch": 1.6839577329490876,
+      "grad_norm": 0.19907650351524353,
+      "learning_rate": 2.0020804855165857e-05,
+      "loss": 3.164986801147461,
+      "step": 8765
+    },
+    {
+      "epoch": 1.6849183477425553,
+      "grad_norm": 0.19891948997974396,
+      "learning_rate": 1.9902044431912e-05,
+      "loss": 3.16265869140625,
+      "step": 8770
+    },
+    {
+      "epoch": 1.685878962536023,
+      "grad_norm": 0.23641318082809448,
+      "learning_rate": 1.9783612259146485e-05,
+      "loss": 3.164572525024414,
+      "step": 8775
+    },
+    {
+      "epoch": 1.6868395773294909,
+      "grad_norm": 0.2543870210647583,
+      "learning_rate": 1.9665508635685767e-05,
+      "loss": 3.1619945526123048,
+      "step": 8780
+    },
+    {
+      "epoch": 1.6878001921229586,
+      "grad_norm": 0.1898890733718872,
+      "learning_rate": 1.9547733859517443e-05,
+      "loss": 3.161852264404297,
+      "step": 8785
+    },
+    {
+      "epoch": 1.6887608069164264,
+      "grad_norm": 0.18471160531044006,
+      "learning_rate": 1.9430288227799367e-05,
+      "loss": 3.167014312744141,
+      "step": 8790
+    },
+    {
+      "epoch": 1.6897214217098944,
+      "grad_norm": 0.1837644726037979,
+      "learning_rate": 1.931317203685893e-05,
+      "loss": 3.1675960540771486,
+      "step": 8795
+    },
+    {
+      "epoch": 1.6906820365033621,
+      "grad_norm": 0.16458193957805634,
+      "learning_rate": 1.9196385582192247e-05,
+      "loss": 3.161328125,
+      "step": 8800
+    },
+    {
+      "epoch": 1.6916426512968301,
+      "grad_norm": 0.18657121062278748,
+      "learning_rate": 1.9079929158463526e-05,
+      "loss": 3.163207244873047,
+      "step": 8805
+    },
+    {
+      "epoch": 1.6926032660902979,
+      "grad_norm": 0.2095474749803543,
+      "learning_rate": 1.896380305950434e-05,
+      "loss": 3.1646053314208986,
+      "step": 8810
+    },
+    {
+      "epoch": 1.6935638808837656,
+      "grad_norm": 0.1963958740234375,
+      "learning_rate": 1.8848007578312686e-05,
+      "loss": 3.164917755126953,
+      "step": 8815
+    },
+    {
+      "epoch": 1.6945244956772334,
+      "grad_norm": 0.18170610070228577,
+      "learning_rate": 1.8732543007052452e-05,
+      "loss": 3.1576461791992188,
+      "step": 8820
+    },
+    {
+      "epoch": 1.6954851104707012,
+      "grad_norm": 0.2065751552581787,
+      "learning_rate": 1.8617409637052606e-05,
+      "loss": 3.166869354248047,
+      "step": 8825
+    },
+    {
+      "epoch": 1.696445725264169,
+      "grad_norm": 0.18570922315120697,
+      "learning_rate": 1.8502607758806487e-05,
+      "loss": 3.162985992431641,
+      "step": 8830
+    },
+    {
+      "epoch": 1.697406340057637,
+      "grad_norm": 0.22870124876499176,
+      "learning_rate": 1.8388137661970948e-05,
+      "loss": 3.165153694152832,
+      "step": 8835
+    },
+    {
+      "epoch": 1.6983669548511047,
+      "grad_norm": 0.18405385315418243,
+      "learning_rate": 1.8273999635365806e-05,
+      "loss": 3.162175750732422,
+      "step": 8840
+    },
+    {
+      "epoch": 1.6993275696445727,
+      "grad_norm": 0.1959793120622635,
+      "learning_rate": 1.8160193966973047e-05,
+      "loss": 3.161460113525391,
+      "step": 8845
+    },
+    {
+      "epoch": 1.7002881844380404,
+      "grad_norm": 0.20207500457763672,
+      "learning_rate": 1.8046720943936e-05,
+      "loss": 3.1632549285888674,
+      "step": 8850
+    },
+    {
+      "epoch": 1.7012487992315082,
+      "grad_norm": 0.20162688195705414,
+      "learning_rate": 1.7933580852558742e-05,
+      "loss": 3.1618804931640625,
+      "step": 8855
+    },
+    {
+      "epoch": 1.702209414024976,
+      "grad_norm": 0.18540622293949127,
+      "learning_rate": 1.7820773978305365e-05,
+      "loss": 3.161595916748047,
+      "step": 8860
+    },
+    {
+      "epoch": 1.7031700288184437,
+      "grad_norm": 0.19820909202098846,
+      "learning_rate": 1.7708300605799202e-05,
+      "loss": 3.1606048583984374,
+      "step": 8865
+    },
+    {
+      "epoch": 1.7041306436119115,
+      "grad_norm": 0.2218032330274582,
+      "learning_rate": 1.7596161018822007e-05,
+      "loss": 3.1620101928710938,
+      "step": 8870
+    },
+    {
+      "epoch": 1.7050912584053795,
+      "grad_norm": 0.1894785612821579,
+      "learning_rate": 1.7484355500313568e-05,
+      "loss": 3.161252975463867,
+      "step": 8875
+    },
+    {
+      "epoch": 1.7060518731988472,
+      "grad_norm": 0.20882797241210938,
+      "learning_rate": 1.7372884332370677e-05,
+      "loss": 3.162582015991211,
+      "step": 8880
+    },
+    {
+      "epoch": 1.7070124879923152,
+      "grad_norm": 0.2097301483154297,
+      "learning_rate": 1.7261747796246478e-05,
+      "loss": 3.162227249145508,
+      "step": 8885
+    },
+    {
+      "epoch": 1.707973102785783,
+      "grad_norm": 0.18620242178440094,
+      "learning_rate": 1.7150946172349844e-05,
+      "loss": 3.163176345825195,
+      "step": 8890
+    },
+    {
+      "epoch": 1.7089337175792507,
+      "grad_norm": 0.1744890809059143,
+      "learning_rate": 1.7040479740244684e-05,
+      "loss": 3.1600067138671877,
+      "step": 8895
+    },
+    {
+      "epoch": 1.7098943323727185,
+      "grad_norm": 0.20441320538520813,
+      "learning_rate": 1.693034877864913e-05,
+      "loss": 3.161859130859375,
+      "step": 8900
+    },
+    {
+      "epoch": 1.7108549471661862,
+      "grad_norm": 0.18480157852172852,
+      "learning_rate": 1.682055356543487e-05,
+      "loss": 3.1619277954101563,
+      "step": 8905
+    },
+    {
+      "epoch": 1.7118155619596542,
+      "grad_norm": 0.19677984714508057,
+      "learning_rate": 1.6711094377626495e-05,
+      "loss": 3.1647516250610352,
+      "step": 8910
+    },
+    {
+      "epoch": 1.712776176753122,
+      "grad_norm": 0.19380095601081848,
+      "learning_rate": 1.66019714914008e-05,
+      "loss": 3.1602848052978514,
+      "step": 8915
+    },
+    {
+      "epoch": 1.71373679154659,
+      "grad_norm": 0.17474035918712616,
+      "learning_rate": 1.6493185182085967e-05,
+      "loss": 3.164753532409668,
+      "step": 8920
+    },
+    {
+      "epoch": 1.7146974063400577,
+      "grad_norm": 0.18277983367443085,
+      "learning_rate": 1.6384735724161045e-05,
+      "loss": 3.162521743774414,
+      "step": 8925
+    },
+    {
+      "epoch": 1.7156580211335255,
+      "grad_norm": 0.17841650545597076,
+      "learning_rate": 1.6276623391255146e-05,
+      "loss": 3.1669170379638674,
+      "step": 8930
+    },
+    {
+      "epoch": 1.7166186359269933,
+      "grad_norm": 0.18692415952682495,
+      "learning_rate": 1.6168848456146793e-05,
+      "loss": 3.1644006729125977,
+      "step": 8935
+    },
+    {
+      "epoch": 1.717579250720461,
+      "grad_norm": 0.22790934145450592,
+      "learning_rate": 1.606141119076314e-05,
+      "loss": 3.164959716796875,
+      "step": 8940
+    },
+    {
+      "epoch": 1.7185398655139288,
+      "grad_norm": 0.19892485439777374,
+      "learning_rate": 1.595431186617948e-05,
+      "loss": 3.1620708465576173,
+      "step": 8945
+    },
+    {
+      "epoch": 1.7195004803073968,
+      "grad_norm": 0.19590741395950317,
+      "learning_rate": 1.5847550752618427e-05,
+      "loss": 3.158066177368164,
+      "step": 8950
+    },
+    {
+      "epoch": 1.7204610951008645,
+      "grad_norm": 0.2052772343158722,
+      "learning_rate": 1.5741128119449153e-05,
+      "loss": 3.161151885986328,
+      "step": 8955
+    },
+    {
+      "epoch": 1.7214217098943325,
+      "grad_norm": 0.18683621287345886,
+      "learning_rate": 1.5635044235186906e-05,
+      "loss": 3.1654678344726563,
+      "step": 8960
+    },
+    {
+      "epoch": 1.7223823246878003,
+      "grad_norm": 0.19518622756004333,
+      "learning_rate": 1.5529299367492208e-05,
+      "loss": 3.160481262207031,
+      "step": 8965
+    },
+    {
+      "epoch": 1.723342939481268,
+      "grad_norm": 0.1885916143655777,
+      "learning_rate": 1.5423893783170262e-05,
+      "loss": 3.162572479248047,
+      "step": 8970
+    },
+    {
+      "epoch": 1.7243035542747358,
+      "grad_norm": 0.18152864277362823,
+      "learning_rate": 1.531882774817007e-05,
+      "loss": 3.1627574920654298,
+      "step": 8975
+    },
+    {
+      "epoch": 1.7252641690682036,
+      "grad_norm": 0.17760129272937775,
+      "learning_rate": 1.5214101527584071e-05,
+      "loss": 3.1614215850830076,
+      "step": 8980
+    },
+    {
+      "epoch": 1.7262247838616713,
+      "grad_norm": 0.20811888575553894,
+      "learning_rate": 1.5109715385647297e-05,
+      "loss": 3.163728141784668,
+      "step": 8985
+    },
+    {
+      "epoch": 1.7271853986551393,
+      "grad_norm": 0.18515148758888245,
+      "learning_rate": 1.5005669585736618e-05,
+      "loss": 3.161965179443359,
+      "step": 8990
+    },
+    {
+      "epoch": 1.728146013448607,
+      "grad_norm": 0.1933458298444748,
+      "learning_rate": 1.4901964390370308e-05,
+      "loss": 3.163407325744629,
+      "step": 8995
+    },
+    {
+      "epoch": 1.729106628242075,
+      "grad_norm": 0.20849090814590454,
+      "learning_rate": 1.4798600061207195e-05,
+      "loss": 3.163631057739258,
+      "step": 9000
+    },
+    {
+      "epoch": 1.7300672430355428,
+      "grad_norm": 0.18622250854969025,
+      "learning_rate": 1.4695576859046127e-05,
+      "loss": 3.159561538696289,
+      "step": 9005
+    },
+    {
+      "epoch": 1.7310278578290106,
+      "grad_norm": 0.1835029423236847,
+      "learning_rate": 1.4592895043825126e-05,
+      "loss": 3.163999557495117,
+      "step": 9010
+    },
+    {
+      "epoch": 1.7319884726224783,
+      "grad_norm": 0.1808691918849945,
+      "learning_rate": 1.449055487462102e-05,
+      "loss": 3.1594869613647463,
+      "step": 9015
+    },
+    {
+      "epoch": 1.732949087415946,
+      "grad_norm": 0.16524961590766907,
+      "learning_rate": 1.4388556609648572e-05,
+      "loss": 3.163185691833496,
+      "step": 9020
+    },
+    {
+      "epoch": 1.7339097022094139,
+      "grad_norm": 0.16896623373031616,
+      "learning_rate": 1.428690050625979e-05,
+      "loss": 3.1632381439208985,
+      "step": 9025
+    },
+    {
+      "epoch": 1.7348703170028819,
+      "grad_norm": 0.1609111875295639,
+      "learning_rate": 1.4185586820943506e-05,
+      "loss": 3.1643299102783202,
+      "step": 9030
+    },
+    {
+      "epoch": 1.7358309317963496,
+      "grad_norm": 0.16877403855323792,
+      "learning_rate": 1.4084615809324523e-05,
+      "loss": 3.164197540283203,
+      "step": 9035
+    },
+    {
+      "epoch": 1.7367915465898176,
+      "grad_norm": 0.17374147474765778,
+      "learning_rate": 1.3983987726163087e-05,
+      "loss": 3.1597652435302734,
+      "step": 9040
+    },
+    {
+      "epoch": 1.7377521613832854,
+      "grad_norm": 0.1869489699602127,
+      "learning_rate": 1.3883702825354138e-05,
+      "loss": 3.160333251953125,
+      "step": 9045
+    },
+    {
+      "epoch": 1.7387127761767531,
+      "grad_norm": 0.19941848516464233,
+      "learning_rate": 1.3783761359926771e-05,
+      "loss": 3.1596229553222654,
+      "step": 9050
+    },
+    {
+      "epoch": 1.739673390970221,
+      "grad_norm": 0.16694855690002441,
+      "learning_rate": 1.3684163582043595e-05,
+      "loss": 3.159475898742676,
+      "step": 9055
+    },
+    {
+      "epoch": 1.7406340057636887,
+      "grad_norm": 0.1921820491552353,
+      "learning_rate": 1.3584909742999978e-05,
+      "loss": 3.162578010559082,
+      "step": 9060
+    },
+    {
+      "epoch": 1.7415946205571564,
+      "grad_norm": 0.17421427369117737,
+      "learning_rate": 1.3486000093223565e-05,
+      "loss": 3.1633106231689454,
+      "step": 9065
+    },
+    {
+      "epoch": 1.7425552353506244,
+      "grad_norm": 0.16822272539138794,
+      "learning_rate": 1.3387434882273529e-05,
+      "loss": 3.1628774642944335,
+      "step": 9070
+    },
+    {
+      "epoch": 1.7435158501440924,
+      "grad_norm": 0.2147689312696457,
+      "learning_rate": 1.3289214358840022e-05,
+      "loss": 3.164061737060547,
+      "step": 9075
+    },
+    {
+      "epoch": 1.7444764649375601,
+      "grad_norm": 0.19877059757709503,
+      "learning_rate": 1.3191338770743493e-05,
+      "loss": 3.164179039001465,
+      "step": 9080
+    },
+    {
+      "epoch": 1.745437079731028,
+      "grad_norm": 0.17470191419124603,
+      "learning_rate": 1.30938083649341e-05,
+      "loss": 3.162880706787109,
+      "step": 9085
+    },
+    {
+      "epoch": 1.7463976945244957,
+      "grad_norm": 0.17637498676776886,
+      "learning_rate": 1.2996623387491085e-05,
+      "loss": 3.163307952880859,
+      "step": 9090
+    },
+    {
+      "epoch": 1.7473583093179634,
+      "grad_norm": 0.17012238502502441,
+      "learning_rate": 1.2899784083622067e-05,
+      "loss": 3.1613311767578125,
+      "step": 9095
+    },
+    {
+      "epoch": 1.7483189241114312,
+      "grad_norm": 0.15726859867572784,
+      "learning_rate": 1.2803290697662566e-05,
+      "loss": 3.161721038818359,
+      "step": 9100
+    },
+    {
+      "epoch": 1.7492795389048992,
+      "grad_norm": 0.1671910136938095,
+      "learning_rate": 1.2707143473075299e-05,
+      "loss": 3.1644134521484375,
+      "step": 9105
+    },
+    {
+      "epoch": 1.750240153698367,
+      "grad_norm": 0.17212195694446564,
+      "learning_rate": 1.2611342652449597e-05,
+      "loss": 3.1598962783813476,
+      "step": 9110
+    },
+    {
+      "epoch": 1.751200768491835,
+      "grad_norm": 0.17955084145069122,
+      "learning_rate": 1.2515888477500708e-05,
+      "loss": 3.161183166503906,
+      "step": 9115
+    },
+    {
+      "epoch": 1.7521613832853027,
+      "grad_norm": 0.15037564933300018,
+      "learning_rate": 1.2420781189069346e-05,
+      "loss": 3.164091873168945,
+      "step": 9120
+    },
+    {
+      "epoch": 1.7531219980787704,
+      "grad_norm": 0.16244594752788544,
+      "learning_rate": 1.2326021027120958e-05,
+      "loss": 3.162504196166992,
+      "step": 9125
+    },
+    {
+      "epoch": 1.7540826128722382,
+      "grad_norm": 0.1728479117155075,
+      "learning_rate": 1.2231608230745128e-05,
+      "loss": 3.159666633605957,
+      "step": 9130
+    },
+    {
+      "epoch": 1.755043227665706,
+      "grad_norm": 0.17677444219589233,
+      "learning_rate": 1.2137543038155034e-05,
+      "loss": 3.16094970703125,
+      "step": 9135
+    },
+    {
+      "epoch": 1.7560038424591737,
+      "grad_norm": 0.16001811623573303,
+      "learning_rate": 1.2043825686686798e-05,
+      "loss": 3.1602264404296876,
+      "step": 9140
+    },
+    {
+      "epoch": 1.7569644572526417,
+      "grad_norm": 0.1979171186685562,
+      "learning_rate": 1.1950456412798954e-05,
+      "loss": 3.160991096496582,
+      "step": 9145
+    },
+    {
+      "epoch": 1.7579250720461095,
+      "grad_norm": 0.18909943103790283,
+      "learning_rate": 1.1857435452071706e-05,
+      "loss": 3.164293098449707,
+      "step": 9150
+    },
+    {
+      "epoch": 1.7588856868395775,
+      "grad_norm": 0.1578613966703415,
+      "learning_rate": 1.1764763039206516e-05,
+      "loss": 3.1651378631591798,
+      "step": 9155
+    },
+    {
+      "epoch": 1.7598463016330452,
+      "grad_norm": 0.17254026234149933,
+      "learning_rate": 1.1672439408025409e-05,
+      "loss": 3.1591054916381838,
+      "step": 9160
+    },
+    {
+      "epoch": 1.760806916426513,
+      "grad_norm": 0.1548478901386261,
+      "learning_rate": 1.1580464791470317e-05,
+      "loss": 3.162530517578125,
+      "step": 9165
+    },
+    {
+      "epoch": 1.7617675312199808,
+      "grad_norm": 0.16993309557437897,
+      "learning_rate": 1.1488839421602663e-05,
+      "loss": 3.1617578506469726,
+      "step": 9170
+    },
+    {
+      "epoch": 1.7627281460134485,
+      "grad_norm": 0.18223033845424652,
+      "learning_rate": 1.1397563529602654e-05,
+      "loss": 3.1608226776123045,
+      "step": 9175
+    },
+    {
+      "epoch": 1.7636887608069163,
+      "grad_norm": 0.15401305258274078,
+      "learning_rate": 1.130663734576877e-05,
+      "loss": 3.1587234497070313,
+      "step": 9180
+    },
+    {
+      "epoch": 1.7646493756003843,
+      "grad_norm": 0.19235272705554962,
+      "learning_rate": 1.1216061099517016e-05,
+      "loss": 3.158829116821289,
+      "step": 9185
+    },
+    {
+      "epoch": 1.765609990393852,
+      "grad_norm": 0.1655486822128296,
+      "learning_rate": 1.1125835019380614e-05,
+      "loss": 3.160569190979004,
+      "step": 9190
+    },
+    {
+      "epoch": 1.76657060518732,
+      "grad_norm": 0.17558416724205017,
+      "learning_rate": 1.1035959333009231e-05,
+      "loss": 3.1613176345825194,
+      "step": 9195
+    },
+    {
+      "epoch": 1.7675312199807878,
+      "grad_norm": 0.16277319192886353,
+      "learning_rate": 1.0946434267168375e-05,
+      "loss": 3.1619541168212892,
+      "step": 9200
+    },
+    {
+      "epoch": 1.7684918347742555,
+      "grad_norm": 0.15174464881420135,
+      "learning_rate": 1.0857260047739025e-05,
+      "loss": 3.1618457794189454,
+      "step": 9205
+    },
+    {
+      "epoch": 1.7694524495677233,
+      "grad_norm": 0.14863047003746033,
+      "learning_rate": 1.076843689971687e-05,
+      "loss": 3.1616443634033202,
+      "step": 9210
+    },
+    {
+      "epoch": 1.770413064361191,
+      "grad_norm": 0.16424240171909332,
+      "learning_rate": 1.0679965047211841e-05,
+      "loss": 3.1608753204345703,
+      "step": 9215
+    },
+    {
+      "epoch": 1.7713736791546588,
+      "grad_norm": 0.15550029277801514,
+      "learning_rate": 1.0591844713447479e-05,
+      "loss": 3.15771427154541,
+      "step": 9220
+    },
+    {
+      "epoch": 1.7723342939481268,
+      "grad_norm": 0.15241944789886475,
+      "learning_rate": 1.0504076120760413e-05,
+      "loss": 3.1600357055664063,
+      "step": 9225
+    },
+    {
+      "epoch": 1.7732949087415946,
+      "grad_norm": 0.15490137040615082,
+      "learning_rate": 1.0416659490599871e-05,
+      "loss": 3.163216400146484,
+      "step": 9230
+    },
+    {
+      "epoch": 1.7742555235350626,
+      "grad_norm": 0.17097872495651245,
+      "learning_rate": 1.0329595043526905e-05,
+      "loss": 3.164168930053711,
+      "step": 9235
+    },
+    {
+      "epoch": 1.7752161383285303,
+      "grad_norm": 0.19601082801818848,
+      "learning_rate": 1.024288299921408e-05,
+      "loss": 3.1567230224609375,
+      "step": 9240
+    },
+    {
+      "epoch": 1.776176753121998,
+      "grad_norm": 0.1483290195465088,
+      "learning_rate": 1.015652357644477e-05,
+      "loss": 3.1643510818481446,
+      "step": 9245
+    },
+    {
+      "epoch": 1.7771373679154658,
+      "grad_norm": 0.1415608525276184,
+      "learning_rate": 1.0070516993112676e-05,
+      "loss": 3.157514953613281,
+      "step": 9250
+    },
+    {
+      "epoch": 1.7780979827089336,
+      "grad_norm": 0.14610423147678375,
+      "learning_rate": 9.984863466221199e-06,
+      "loss": 3.1627222061157227,
+      "step": 9255
+    },
+    {
+      "epoch": 1.7790585975024016,
+      "grad_norm": 0.17516079545021057,
+      "learning_rate": 9.899563211883e-06,
+      "loss": 3.1590530395507814,
+      "step": 9260
+    },
+    {
+      "epoch": 1.7800192122958693,
+      "grad_norm": 0.1540340632200241,
+      "learning_rate": 9.814616445319384e-06,
+      "loss": 3.1593671798706056,
+      "step": 9265
+    },
+    {
+      "epoch": 1.7809798270893373,
+      "grad_norm": 0.18057873845100403,
+      "learning_rate": 9.730023380859725e-06,
+      "loss": 3.162105751037598,
+      "step": 9270
+    },
+    {
+      "epoch": 1.781940441882805,
+      "grad_norm": 0.17139260470867157,
+      "learning_rate": 9.645784231941005e-06,
+      "loss": 3.1630107879638674,
+      "step": 9275
+    },
+    {
+      "epoch": 1.7829010566762729,
+      "grad_norm": 0.16848276555538177,
+      "learning_rate": 9.561899211107244e-06,
+      "loss": 3.1598880767822264,
+      "step": 9280
+    },
+    {
+      "epoch": 1.7838616714697406,
+      "grad_norm": 0.16549763083457947,
+      "learning_rate": 9.478368530008967e-06,
+      "loss": 3.160906219482422,
+      "step": 9285
+    },
+    {
+      "epoch": 1.7848222862632084,
+      "grad_norm": 0.15802277624607086,
+      "learning_rate": 9.39519239940264e-06,
+      "loss": 3.161147689819336,
+      "step": 9290
+    },
+    {
+      "epoch": 1.7857829010566761,
+      "grad_norm": 0.1761743128299713,
+      "learning_rate": 9.31237102915019e-06,
+      "loss": 3.158751678466797,
+      "step": 9295
+    },
+    {
+      "epoch": 1.7867435158501441,
+      "grad_norm": 0.1671990603208542,
+      "learning_rate": 9.229904628218427e-06,
+      "loss": 3.1624494552612306,
+      "step": 9300
+    },
+    {
+      "epoch": 1.7877041306436119,
+      "grad_norm": 0.19958704710006714,
+      "learning_rate": 9.14779340467851e-06,
+      "loss": 3.1618110656738283,
+      "step": 9305
+    },
+    {
+      "epoch": 1.7886647454370799,
+      "grad_norm": 0.16844980418682098,
+      "learning_rate": 9.06603756570552e-06,
+      "loss": 3.156951141357422,
+      "step": 9310
+    },
+    {
+      "epoch": 1.7896253602305476,
+      "grad_norm": 0.14970090985298157,
+      "learning_rate": 8.984637317577782e-06,
+      "loss": 3.1624423980712892,
+      "step": 9315
+    },
+    {
+      "epoch": 1.7905859750240154,
+      "grad_norm": 0.15226224064826965,
+      "learning_rate": 8.90359286567654e-06,
+      "loss": 3.1624835968017577,
+      "step": 9320
+    },
+    {
+      "epoch": 1.7915465898174832,
+      "grad_norm": 0.1699807345867157,
+      "learning_rate": 8.822904414485194e-06,
+      "loss": 3.1621606826782225,
+      "step": 9325
+    },
+    {
+      "epoch": 1.792507204610951,
+      "grad_norm": 0.15113097429275513,
+      "learning_rate": 8.742572167589008e-06,
+      "loss": 3.16312255859375,
+      "step": 9330
+    },
+    {
+      "epoch": 1.7934678194044187,
+      "grad_norm": 0.16513291001319885,
+      "learning_rate": 8.662596327674499e-06,
+      "loss": 3.1624629974365233,
+      "step": 9335
+    },
+    {
+      "epoch": 1.7944284341978867,
+      "grad_norm": 0.154737189412117,
+      "learning_rate": 8.582977096528887e-06,
+      "loss": 3.1638771057128907,
+      "step": 9340
+    },
+    {
+      "epoch": 1.7953890489913544,
+      "grad_norm": 0.18483127653598785,
+      "learning_rate": 8.503714675039663e-06,
+      "loss": 3.16489315032959,
+      "step": 9345
+    },
+    {
+      "epoch": 1.7963496637848224,
+      "grad_norm": 0.16609624028205872,
+      "learning_rate": 8.424809263194054e-06,
+      "loss": 3.1592342376708986,
+      "step": 9350
+    },
+    {
+      "epoch": 1.7973102785782902,
+      "grad_norm": 0.13648296892642975,
+      "learning_rate": 8.346261060078524e-06,
+      "loss": 3.163716506958008,
+      "step": 9355
+    },
+    {
+      "epoch": 1.798270893371758,
+      "grad_norm": 0.16231390833854675,
+      "learning_rate": 8.268070263878223e-06,
+      "loss": 3.1625925064086915,
+      "step": 9360
+    },
+    {
+      "epoch": 1.7992315081652257,
+      "grad_norm": 0.15581952035427094,
+      "learning_rate": 8.190237071876576e-06,
+      "loss": 3.161655235290527,
+      "step": 9365
+    },
+    {
+      "epoch": 1.8001921229586935,
+      "grad_norm": 0.15568803250789642,
+      "learning_rate": 8.11276168045471e-06,
+      "loss": 3.1620370864868166,
+      "step": 9370
+    },
+    {
+      "epoch": 1.8011527377521612,
+      "grad_norm": 0.1662697196006775,
+      "learning_rate": 8.035644285090958e-06,
+      "loss": 3.1623306274414062,
+      "step": 9375
+    },
+    {
+      "epoch": 1.8021133525456292,
+      "grad_norm": 0.16496489942073822,
+      "learning_rate": 7.958885080360445e-06,
+      "loss": 3.1613933563232424,
+      "step": 9380
+    },
+    {
+      "epoch": 1.803073967339097,
+      "grad_norm": 0.15053115785121918,
+      "learning_rate": 7.882484259934497e-06,
+      "loss": 3.1627952575683596,
+      "step": 9385
+    },
+    {
+      "epoch": 1.804034582132565,
+      "grad_norm": 0.14695799350738525,
+      "learning_rate": 7.806442016580267e-06,
+      "loss": 3.160305404663086,
+      "step": 9390
+    },
+    {
+      "epoch": 1.8049951969260327,
+      "grad_norm": 0.15005071461200714,
+      "learning_rate": 7.730758542160064e-06,
+      "loss": 3.1611255645751952,
+      "step": 9395
+    },
+    {
+      "epoch": 1.8059558117195005,
+      "grad_norm": 0.15226683020591736,
+      "learning_rate": 7.655434027631124e-06,
+      "loss": 3.1629497528076174,
+      "step": 9400
+    },
+    {
+      "epoch": 1.8069164265129682,
+      "grad_norm": 0.15546758472919464,
+      "learning_rate": 7.580468663044903e-06,
+      "loss": 3.1586400985717775,
+      "step": 9405
+    },
+    {
+      "epoch": 1.807877041306436,
+      "grad_norm": 0.1452452838420868,
+      "learning_rate": 7.505862637546684e-06,
+      "loss": 3.160877227783203,
+      "step": 9410
+    },
+    {
+      "epoch": 1.808837656099904,
+      "grad_norm": 0.13006171584129333,
+      "learning_rate": 7.43161613937514e-06,
+      "loss": 3.1579397201538084,
+      "step": 9415
+    },
+    {
+      "epoch": 1.8097982708933718,
+      "grad_norm": 0.15625065565109253,
+      "learning_rate": 7.3577293558618075e-06,
+      "loss": 3.1598045349121096,
+      "step": 9420
+    },
+    {
+      "epoch": 1.8107588856868397,
+      "grad_norm": 0.16364313662052155,
+      "learning_rate": 7.284202473430612e-06,
+      "loss": 3.1616336822509767,
+      "step": 9425
+    },
+    {
+      "epoch": 1.8117195004803075,
+      "grad_norm": 0.15080556273460388,
+      "learning_rate": 7.211035677597421e-06,
+      "loss": 3.1628395080566407,
+      "step": 9430
+    },
+    {
+      "epoch": 1.8126801152737753,
+      "grad_norm": 0.1453072428703308,
+      "learning_rate": 7.138229152969582e-06,
+      "loss": 3.161026382446289,
+      "step": 9435
+    },
+    {
+      "epoch": 1.813640730067243,
+      "grad_norm": 0.1453263759613037,
+      "learning_rate": 7.065783083245435e-06,
+      "loss": 3.16085262298584,
+      "step": 9440
+    },
+    {
+      "epoch": 1.8146013448607108,
+      "grad_norm": 0.13254213333129883,
+      "learning_rate": 6.993697651213798e-06,
+      "loss": 3.1604400634765626,
+      "step": 9445
+    },
+    {
+      "epoch": 1.8155619596541785,
+      "grad_norm": 0.1310938000679016,
+      "learning_rate": 6.921973038753647e-06,
+      "loss": 3.158857727050781,
+      "step": 9450
+    },
+    {
+      "epoch": 1.8165225744476465,
+      "grad_norm": 0.13702644407749176,
+      "learning_rate": 6.8506094268335255e-06,
+      "loss": 3.16072998046875,
+      "step": 9455
+    },
+    {
+      "epoch": 1.8174831892411143,
+      "grad_norm": 0.15652552247047424,
+      "learning_rate": 6.779606995511183e-06,
+      "loss": 3.161703109741211,
+      "step": 9460
+    },
+    {
+      "epoch": 1.8184438040345823,
+      "grad_norm": 0.1748208999633789,
+      "learning_rate": 6.7089659239329685e-06,
+      "loss": 3.162480926513672,
+      "step": 9465
+    },
+    {
+      "epoch": 1.81940441882805,
+      "grad_norm": 0.15371812880039215,
+      "learning_rate": 6.638686390333592e-06,
+      "loss": 3.1619606018066406,
+      "step": 9470
+    },
+    {
+      "epoch": 1.8203650336215178,
+      "grad_norm": 0.15862195193767548,
+      "learning_rate": 6.56876857203556e-06,
+      "loss": 3.1612228393554687,
+      "step": 9475
+    },
+    {
+      "epoch": 1.8213256484149856,
+      "grad_norm": 0.14958879351615906,
+      "learning_rate": 6.499212645448626e-06,
+      "loss": 3.161460113525391,
+      "step": 9480
+    },
+    {
+      "epoch": 1.8222862632084533,
+      "grad_norm": 0.14320014417171478,
+      "learning_rate": 6.430018786069574e-06,
+      "loss": 3.1611602783203123,
+      "step": 9485
+    },
+    {
+      "epoch": 1.823246878001921,
+      "grad_norm": 0.14477957785129547,
+      "learning_rate": 6.361187168481618e-06,
+      "loss": 3.162501907348633,
+      "step": 9490
+    },
+    {
+      "epoch": 1.824207492795389,
+      "grad_norm": 0.14906539022922516,
+      "learning_rate": 6.292717966353955e-06,
+      "loss": 3.1564708709716798,
+      "step": 9495
+    },
+    {
+      "epoch": 1.8251681075888568,
+      "grad_norm": 0.1561676263809204,
+      "learning_rate": 6.224611352441444e-06,
+      "loss": 3.164084053039551,
+      "step": 9500
+    },
+    {
+      "epoch": 1.8261287223823248,
+      "grad_norm": 0.15524768829345703,
+      "learning_rate": 6.156867498584028e-06,
+      "loss": 3.164693832397461,
+      "step": 9505
+    },
+    {
+      "epoch": 1.8270893371757926,
+      "grad_norm": 0.13482025265693665,
+      "learning_rate": 6.089486575706431e-06,
+      "loss": 3.161751556396484,
+      "step": 9510
+    },
+    {
+      "epoch": 1.8280499519692603,
+      "grad_norm": 0.1538052260875702,
+      "learning_rate": 6.022468753817611e-06,
+      "loss": 3.162434196472168,
+      "step": 9515
+    },
+    {
+      "epoch": 1.829010566762728,
+      "grad_norm": 0.13271762430667877,
+      "learning_rate": 5.955814202010406e-06,
+      "loss": 3.1618423461914062,
+      "step": 9520
+    },
+    {
+      "epoch": 1.8299711815561959,
+      "grad_norm": 0.14370056986808777,
+      "learning_rate": 5.88952308846109e-06,
+      "loss": 3.1616491317749023,
+      "step": 9525
+    },
+    {
+      "epoch": 1.8309317963496636,
+      "grad_norm": 0.13553740084171295,
+      "learning_rate": 5.823595580428969e-06,
+      "loss": 3.1615522384643553,
+      "step": 9530
+    },
+    {
+      "epoch": 1.8318924111431316,
+      "grad_norm": 0.14558495581150055,
+      "learning_rate": 5.758031844255884e-06,
+      "loss": 3.163438415527344,
+      "step": 9535
+    },
+    {
+      "epoch": 1.8328530259365994,
+      "grad_norm": 0.1292106956243515,
+      "learning_rate": 5.692832045365858e-06,
+      "loss": 3.1632017135620116,
+      "step": 9540
+    },
+    {
+      "epoch": 1.8338136407300674,
+      "grad_norm": 0.13839663565158844,
+      "learning_rate": 5.627996348264685e-06,
+      "loss": 3.1595331192016602,
+      "step": 9545
+    },
+    {
+      "epoch": 1.8347742555235351,
+      "grad_norm": 0.13135622441768646,
+      "learning_rate": 5.563524916539508e-06,
+      "loss": 3.161990165710449,
+      "step": 9550
+    },
+    {
+      "epoch": 1.8357348703170029,
+      "grad_norm": 0.13140968978405,
+      "learning_rate": 5.499417912858289e-06,
+      "loss": 3.159449005126953,
+      "step": 9555
+    },
+    {
+      "epoch": 1.8366954851104706,
+      "grad_norm": 0.14151158928871155,
+      "learning_rate": 5.435675498969661e-06,
+      "loss": 3.1604162216186524,
+      "step": 9560
+    },
+    {
+      "epoch": 1.8376560999039384,
+      "grad_norm": 0.1417761594057083,
+      "learning_rate": 5.372297835702222e-06,
+      "loss": 3.1592178344726562,
+      "step": 9565
+    },
+    {
+      "epoch": 1.8386167146974062,
+      "grad_norm": 0.13075558841228485,
+      "learning_rate": 5.309285082964343e-06,
+      "loss": 3.159617233276367,
+      "step": 9570
+    },
+    {
+      "epoch": 1.8395773294908742,
+      "grad_norm": 0.13188965618610382,
+      "learning_rate": 5.24663739974363e-06,
+      "loss": 3.1601367950439454,
+      "step": 9575
+    },
+    {
+      "epoch": 1.8405379442843421,
+      "grad_norm": 0.13191638886928558,
+      "learning_rate": 5.184354944106661e-06,
+      "loss": 3.16494197845459,
+      "step": 9580
+    },
+    {
+      "epoch": 1.84149855907781,
+      "grad_norm": 0.16160184144973755,
+      "learning_rate": 5.122437873198415e-06,
+      "loss": 3.159535217285156,
+      "step": 9585
+    },
+    {
+      "epoch": 1.8424591738712777,
+      "grad_norm": 0.13549911975860596,
+      "learning_rate": 5.060886343242043e-06,
+      "loss": 3.1587257385253906,
+      "step": 9590
+    },
+    {
+      "epoch": 1.8434197886647454,
+      "grad_norm": 0.13970741629600525,
+      "learning_rate": 4.999700509538368e-06,
+      "loss": 3.1649681091308595,
+      "step": 9595
+    },
+    {
+      "epoch": 1.8443804034582132,
+      "grad_norm": 0.134404718875885,
+      "learning_rate": 4.938880526465516e-06,
+      "loss": 3.1567209243774412,
+      "step": 9600
+    },
+    {
+      "epoch": 1.845341018251681,
+      "grad_norm": 0.14060495793819427,
+      "learning_rate": 4.878426547478537e-06,
+      "loss": 3.161220741271973,
+      "step": 9605
+    },
+    {
+      "epoch": 1.846301633045149,
+      "grad_norm": 0.14048533141613007,
+      "learning_rate": 4.818338725109033e-06,
+      "loss": 3.1618812561035154,
+      "step": 9610
+    },
+    {
+      "epoch": 1.8472622478386167,
+      "grad_norm": 0.13674895465373993,
+      "learning_rate": 4.758617210964749e-06,
+      "loss": 3.1565229415893556,
+      "step": 9615
+    },
+    {
+      "epoch": 1.8482228626320847,
+      "grad_norm": 0.13577382266521454,
+      "learning_rate": 4.6992621557292e-06,
+      "loss": 3.158913230895996,
+      "step": 9620
+    },
+    {
+      "epoch": 1.8491834774255524,
+      "grad_norm": 0.13337448239326477,
+      "learning_rate": 4.640273709161224e-06,
+      "loss": 3.161669158935547,
+      "step": 9625
+    },
+    {
+      "epoch": 1.8501440922190202,
+      "grad_norm": 0.134219229221344,
+      "learning_rate": 4.5816520200948005e-06,
+      "loss": 3.160819435119629,
+      "step": 9630
+    },
+    {
+      "epoch": 1.851104707012488,
+      "grad_norm": 0.13151714205741882,
+      "learning_rate": 4.5233972364383975e-06,
+      "loss": 3.1615375518798827,
+      "step": 9635
+    },
+    {
+      "epoch": 1.8520653218059557,
+      "grad_norm": 0.11864073574542999,
+      "learning_rate": 4.465509505174858e-06,
+      "loss": 3.1609018325805662,
+      "step": 9640
+    },
+    {
+      "epoch": 1.8530259365994235,
+      "grad_norm": 0.14346528053283691,
+      "learning_rate": 4.407988972360849e-06,
+      "loss": 3.161561393737793,
+      "step": 9645
+    },
+    {
+      "epoch": 1.8539865513928915,
+      "grad_norm": 0.13390202820301056,
+      "learning_rate": 4.350835783126594e-06,
+      "loss": 3.1598655700683596,
+      "step": 9650
+    },
+    {
+      "epoch": 1.8549471661863592,
+      "grad_norm": 0.12858052551746368,
+      "learning_rate": 4.2940500816754094e-06,
+      "loss": 3.160798263549805,
+      "step": 9655
+    },
+    {
+      "epoch": 1.8559077809798272,
+      "grad_norm": 0.13176590204238892,
+      "learning_rate": 4.237632011283471e-06,
+      "loss": 3.158690643310547,
+      "step": 9660
+    },
+    {
+      "epoch": 1.856868395773295,
+      "grad_norm": 0.12532593309879303,
+      "learning_rate": 4.181581714299359e-06,
+      "loss": 3.159891128540039,
+      "step": 9665
+    },
+    {
+      "epoch": 1.8578290105667628,
+      "grad_norm": 0.11781672388315201,
+      "learning_rate": 4.12589933214375e-06,
+      "loss": 3.161539077758789,
+      "step": 9670
+    },
+    {
+      "epoch": 1.8587896253602305,
+      "grad_norm": 0.1212988868355751,
+      "learning_rate": 4.070585005308946e-06,
+      "loss": 3.16253662109375,
+      "step": 9675
+    },
+    {
+      "epoch": 1.8597502401536983,
+      "grad_norm": 0.11885207146406174,
+      "learning_rate": 4.015638873358707e-06,
+      "loss": 3.162197303771973,
+      "step": 9680
+    },
+    {
+      "epoch": 1.860710854947166,
+      "grad_norm": 0.12261178344488144,
+      "learning_rate": 3.961061074927752e-06,
+      "loss": 3.1611923217773437,
+      "step": 9685
+    },
+    {
+      "epoch": 1.861671469740634,
+      "grad_norm": 0.13136336207389832,
+      "learning_rate": 3.906851747721495e-06,
+      "loss": 3.160177993774414,
+      "step": 9690
+    },
+    {
+      "epoch": 1.8626320845341018,
+      "grad_norm": 0.13114266097545624,
+      "learning_rate": 3.853011028515563e-06,
+      "loss": 3.15836124420166,
+      "step": 9695
+    },
+    {
+      "epoch": 1.8635926993275698,
+      "grad_norm": 0.13357418775558472,
+      "learning_rate": 3.79953905315567e-06,
+      "loss": 3.161663627624512,
+      "step": 9700
+    },
+    {
+      "epoch": 1.8645533141210375,
+      "grad_norm": 0.11809264123439789,
+      "learning_rate": 3.746435956557081e-06,
+      "loss": 3.1624309539794924,
+      "step": 9705
+    },
+    {
+      "epoch": 1.8655139289145053,
+      "grad_norm": 0.1218031793832779,
+      "learning_rate": 3.6937018727043523e-06,
+      "loss": 3.1634284973144533,
+      "step": 9710
+    },
+    {
+      "epoch": 1.866474543707973,
+      "grad_norm": 0.12914207577705383,
+      "learning_rate": 3.6413369346509848e-06,
+      "loss": 3.1623727798461916,
+      "step": 9715
+    },
+    {
+      "epoch": 1.8674351585014408,
+      "grad_norm": 0.12263166159391403,
+      "learning_rate": 3.5893412745191085e-06,
+      "loss": 3.1578353881835937,
+      "step": 9720
+    },
+    {
+      "epoch": 1.8683957732949086,
+      "grad_norm": 0.11959100514650345,
+      "learning_rate": 3.5377150234990824e-06,
+      "loss": 3.1602123260498045,
+      "step": 9725
+    },
+    {
+      "epoch": 1.8693563880883766,
+      "grad_norm": 0.132080078125,
+      "learning_rate": 3.4864583118492438e-06,
+      "loss": 3.1630222320556642,
+      "step": 9730
+    },
+    {
+      "epoch": 1.8703170028818443,
+      "grad_norm": 0.13228577375411987,
+      "learning_rate": 3.435571268895526e-06,
+      "loss": 3.155730438232422,
+      "step": 9735
+    },
+    {
+      "epoch": 1.8712776176753123,
+      "grad_norm": 0.12898947298526764,
+      "learning_rate": 3.3850540230311918e-06,
+      "loss": 3.1614667892456056,
+      "step": 9740
+    },
+    {
+      "epoch": 1.87223823246878,
+      "grad_norm": 0.1316165179014206,
+      "learning_rate": 3.3349067017163833e-06,
+      "loss": 3.162592315673828,
+      "step": 9745
+    },
+    {
+      "epoch": 1.8731988472622478,
+      "grad_norm": 0.12914599478244781,
+      "learning_rate": 3.28512943147799e-06,
+      "loss": 3.160664749145508,
+      "step": 9750
+    },
+    {
+      "epoch": 1.8741594620557156,
+      "grad_norm": 0.11813253164291382,
+      "learning_rate": 3.23572233790913e-06,
+      "loss": 3.1575859069824217,
+      "step": 9755
+    },
+    {
+      "epoch": 1.8751200768491834,
+      "grad_norm": 0.1261204481124878,
+      "learning_rate": 3.1866855456689878e-06,
+      "loss": 3.1581016540527345,
+      "step": 9760
+    },
+    {
+      "epoch": 1.8760806916426513,
+      "grad_norm": 0.120733842253685,
+      "learning_rate": 3.138019178482409e-06,
+      "loss": 3.1588390350341795,
+      "step": 9765
+    },
+    {
+      "epoch": 1.877041306436119,
+      "grad_norm": 0.13560213148593903,
+      "learning_rate": 3.0897233591396562e-06,
+      "loss": 3.1578319549560545,
+      "step": 9770
+    },
+    {
+      "epoch": 1.878001921229587,
+      "grad_norm": 0.11367765814065933,
+      "learning_rate": 3.0417982094960212e-06,
+      "loss": 3.1566787719726563,
+      "step": 9775
+    },
+    {
+      "epoch": 1.8789625360230549,
+      "grad_norm": 0.11928807944059372,
+      "learning_rate": 2.994243850471545e-06,
+      "loss": 3.162850570678711,
+      "step": 9780
+    },
+    {
+      "epoch": 1.8799231508165226,
+      "grad_norm": 0.11847756057977676,
+      "learning_rate": 2.9470604020507994e-06,
+      "loss": 3.1612308502197264,
+      "step": 9785
+    },
+    {
+      "epoch": 1.8808837656099904,
+      "grad_norm": 0.1250849962234497,
+      "learning_rate": 2.9002479832824553e-06,
+      "loss": 3.1637001037597656,
+      "step": 9790
+    },
+    {
+      "epoch": 1.8818443804034581,
+      "grad_norm": 0.12311087548732758,
+      "learning_rate": 2.8538067122790164e-06,
+      "loss": 3.1589778900146483,
+      "step": 9795
+    },
+    {
+      "epoch": 1.882804995196926,
+      "grad_norm": 0.11985550820827484,
+      "learning_rate": 2.8077367062166013e-06,
+      "loss": 3.1609506607055664,
+      "step": 9800
+    },
+    {
+      "epoch": 1.8837656099903939,
+      "grad_norm": 0.12335570156574249,
+      "learning_rate": 2.762038081334561e-06,
+      "loss": 3.1584909439086912,
+      "step": 9805
+    },
+    {
+      "epoch": 1.8847262247838616,
+      "grad_norm": 0.12690205872058868,
+      "learning_rate": 2.716710952935214e-06,
+      "loss": 3.160411071777344,
+      "step": 9810
+    },
+    {
+      "epoch": 1.8856868395773296,
+      "grad_norm": 0.12958942353725433,
+      "learning_rate": 2.6717554353835435e-06,
+      "loss": 3.16156005859375,
+      "step": 9815
+    },
+    {
+      "epoch": 1.8866474543707974,
+      "grad_norm": 0.1530323475599289,
+      "learning_rate": 2.6271716421069344e-06,
+      "loss": 3.1630611419677734,
+      "step": 9820
+    },
+    {
+      "epoch": 1.8876080691642652,
+      "grad_norm": 0.11445850133895874,
+      "learning_rate": 2.5829596855948876e-06,
+      "loss": 3.160757064819336,
+      "step": 9825
+    },
+    {
+      "epoch": 1.888568683957733,
+      "grad_norm": 0.12335135042667389,
+      "learning_rate": 2.539119677398671e-06,
+      "loss": 3.159043884277344,
+      "step": 9830
+    },
+    {
+      "epoch": 1.8895292987512007,
+      "grad_norm": 0.12193870544433594,
+      "learning_rate": 2.495651728131154e-06,
+      "loss": 3.1591602325439454,
+      "step": 9835
+    },
+    {
+      "epoch": 1.8904899135446684,
+      "grad_norm": 0.11826537549495697,
+      "learning_rate": 2.452555947466439e-06,
+      "loss": 3.161988067626953,
+      "step": 9840
+    },
+    {
+      "epoch": 1.8914505283381364,
+      "grad_norm": 0.12225425243377686,
+      "learning_rate": 2.4098324441395645e-06,
+      "loss": 3.164390754699707,
+      "step": 9845
+    },
+    {
+      "epoch": 1.8924111431316042,
+      "grad_norm": 0.11919926851987839,
+      "learning_rate": 2.367481325946352e-06,
+      "loss": 3.160402297973633,
+      "step": 9850
+    },
+    {
+      "epoch": 1.8933717579250722,
+      "grad_norm": 0.11494944989681244,
+      "learning_rate": 2.325502699742976e-06,
+      "loss": 3.159711456298828,
+      "step": 9855
+    },
+    {
+      "epoch": 1.89433237271854,
+      "grad_norm": 0.11843105405569077,
+      "learning_rate": 2.283896671445862e-06,
+      "loss": 3.159711456298828,
+      "step": 9860
+    },
+    {
+      "epoch": 1.8952929875120077,
+      "grad_norm": 0.12161135673522949,
+      "learning_rate": 2.2426633460312373e-06,
+      "loss": 3.1607955932617187,
+      "step": 9865
+    },
+    {
+      "epoch": 1.8962536023054755,
+      "grad_norm": 0.114869125187397,
+      "learning_rate": 2.2018028275350652e-06,
+      "loss": 3.161262321472168,
+      "step": 9870
+    },
+    {
+      "epoch": 1.8972142170989432,
+      "grad_norm": 0.129729226231575,
+      "learning_rate": 2.1613152190525785e-06,
+      "loss": 3.161954879760742,
+      "step": 9875
+    },
+    {
+      "epoch": 1.898174831892411,
+      "grad_norm": 0.12527115643024445,
+      "learning_rate": 2.1212006227382117e-06,
+      "loss": 3.159313201904297,
+      "step": 9880
+    },
+    {
+      "epoch": 1.899135446685879,
+      "grad_norm": 0.1106216311454773,
+      "learning_rate": 2.0814591398051527e-06,
+      "loss": 3.160965347290039,
+      "step": 9885
+    },
+    {
+      "epoch": 1.9000960614793467,
+      "grad_norm": 0.11833595484495163,
+      "learning_rate": 2.0420908705252926e-06,
+      "loss": 3.160472106933594,
+      "step": 9890
+    },
+    {
+      "epoch": 1.9010566762728147,
+      "grad_norm": 0.10949314385652542,
+      "learning_rate": 2.003095914228775e-06,
+      "loss": 3.163528060913086,
+      "step": 9895
+    },
+    {
+      "epoch": 1.9020172910662825,
+      "grad_norm": 0.1126236766576767,
+      "learning_rate": 1.9644743693038977e-06,
+      "loss": 3.1650867462158203,
+      "step": 9900
+    },
+    {
+      "epoch": 1.9029779058597502,
+      "grad_norm": 0.1123654693365097,
+      "learning_rate": 1.9262263331967785e-06,
+      "loss": 3.1654218673706054,
+      "step": 9905
+    },
+    {
+      "epoch": 1.903938520653218,
+      "grad_norm": 0.12411278486251831,
+      "learning_rate": 1.8883519024111227e-06,
+      "loss": 3.156511688232422,
+      "step": 9910
+    },
+    {
+      "epoch": 1.9048991354466858,
+      "grad_norm": 0.1198534443974495,
+      "learning_rate": 1.8508511725080388e-06,
+      "loss": 3.160737991333008,
+      "step": 9915
+    },
+    {
+      "epoch": 1.9058597502401537,
+      "grad_norm": 0.11347641795873642,
+      "learning_rate": 1.8137242381056571e-06,
+      "loss": 3.1631191253662108,
+      "step": 9920
+    },
+    {
+      "epoch": 1.9068203650336215,
+      "grad_norm": 0.11204273998737335,
+      "learning_rate": 1.7769711928790953e-06,
+      "loss": 3.158831787109375,
+      "step": 9925
+    },
+    {
+      "epoch": 1.9077809798270895,
+      "grad_norm": 0.12282492965459824,
+      "learning_rate": 1.7405921295600421e-06,
+      "loss": 3.1593994140625,
+      "step": 9930
+    },
+    {
+      "epoch": 1.9087415946205573,
+      "grad_norm": 0.11055224388837814,
+      "learning_rate": 1.7045871399365918e-06,
+      "loss": 3.159510040283203,
+      "step": 9935
+    },
+    {
+      "epoch": 1.909702209414025,
+      "grad_norm": 0.11742359399795532,
+      "learning_rate": 1.668956314853026e-06,
+      "loss": 3.161764144897461,
+      "step": 9940
+    },
+    {
+      "epoch": 1.9106628242074928,
+      "grad_norm": 0.10971947759389877,
+      "learning_rate": 1.6336997442095823e-06,
+      "loss": 3.1589191436767576,
+      "step": 9945
+    },
+    {
+      "epoch": 1.9116234390009605,
+      "grad_norm": 0.11144320666790009,
+      "learning_rate": 1.5988175169622197e-06,
+      "loss": 3.1609643936157226,
+      "step": 9950
+    },
+    {
+      "epoch": 1.9125840537944283,
+      "grad_norm": 0.11729393154382706,
+      "learning_rate": 1.5643097211223533e-06,
+      "loss": 3.1646373748779295,
+      "step": 9955
+    },
+    {
+      "epoch": 1.9135446685878963,
+      "grad_norm": 0.11545123904943466,
+      "learning_rate": 1.5301764437567032e-06,
+      "loss": 3.1612579345703127,
+      "step": 9960
+    },
+    {
+      "epoch": 1.914505283381364,
+      "grad_norm": 0.1158137395977974,
+      "learning_rate": 1.4964177709870128e-06,
+      "loss": 3.1609870910644533,
+      "step": 9965
+    },
+    {
+      "epoch": 1.915465898174832,
+      "grad_norm": 0.1098775789141655,
+      "learning_rate": 1.4630337879899146e-06,
+      "loss": 3.1606449127197265,
+      "step": 9970
+    },
+    {
+      "epoch": 1.9164265129682998,
+      "grad_norm": 0.11874253302812576,
+      "learning_rate": 1.4300245789965803e-06,
+      "loss": 3.1594825744628907,
+      "step": 9975
+    },
+    {
+      "epoch": 1.9173871277617676,
+      "grad_norm": 0.11549612134695053,
+      "learning_rate": 1.3973902272926384e-06,
+      "loss": 3.160371780395508,
+      "step": 9980
+    },
+    {
+      "epoch": 1.9183477425552353,
+      "grad_norm": 0.10717111080884933,
+      "learning_rate": 1.3651308152179407e-06,
+      "loss": 3.1619895935058593,
+      "step": 9985
+    },
+    {
+      "epoch": 1.919308357348703,
+      "grad_norm": 0.11006828397512436,
+      "learning_rate": 1.3332464241662622e-06,
+      "loss": 3.1590812683105467,
+      "step": 9990
+    },
+    {
+      "epoch": 1.9202689721421708,
+      "grad_norm": 0.11138733476400375,
+      "learning_rate": 1.3017371345851846e-06,
+      "loss": 3.158272933959961,
+      "step": 9995
+    },
+    {
+      "epoch": 1.9212295869356388,
+      "grad_norm": 0.1066170483827591,
+      "learning_rate": 1.2706030259759137e-06,
+      "loss": 3.158770942687988,
+      "step": 10000
+    },
+    {
+      "epoch": 1.9221902017291066,
+      "grad_norm": 0.11191676557064056,
+      "learning_rate": 1.2398441768929624e-06,
+      "loss": 3.1578929901123045,
+      "step": 10005
+    },
+    {
+      "epoch": 1.9231508165225746,
+      "grad_norm": 0.11436916887760162,
+      "learning_rate": 1.2094606649440841e-06,
+      "loss": 3.1585205078125,
+      "step": 10010
+    },
+    {
+      "epoch": 1.9241114313160423,
+      "grad_norm": 0.11085876077413559,
+      "learning_rate": 1.1794525667899901e-06,
+      "loss": 3.1594661712646483,
+      "step": 10015
+    },
+    {
+      "epoch": 1.92507204610951,
+      "grad_norm": 0.11293391138315201,
+      "learning_rate": 1.149819958144199e-06,
+      "loss": 3.1591583251953126,
+      "step": 10020
+    },
+    {
+      "epoch": 1.9260326609029779,
+      "grad_norm": 0.10681528598070145,
+      "learning_rate": 1.1205629137728044e-06,
+      "loss": 3.160277557373047,
+      "step": 10025
+    },
+    {
+      "epoch": 1.9269932756964456,
+      "grad_norm": 0.1138753667473793,
+      "learning_rate": 1.0916815074943242e-06,
+      "loss": 3.15789909362793,
+      "step": 10030
+    },
+    {
+      "epoch": 1.9279538904899134,
+      "grad_norm": 0.10544762760400772,
+      "learning_rate": 1.0631758121795508e-06,
+      "loss": 3.1582752227783204,
+      "step": 10035
+    },
+    {
+      "epoch": 1.9289145052833814,
+      "grad_norm": 0.11651872098445892,
+      "learning_rate": 1.0350458997512356e-06,
+      "loss": 3.1609378814697267,
+      "step": 10040
+    },
+    {
+      "epoch": 1.9298751200768491,
+      "grad_norm": 0.11160679161548615,
+      "learning_rate": 1.0072918411840548e-06,
+      "loss": 3.161101722717285,
+      "step": 10045
+    },
+    {
+      "epoch": 1.9308357348703171,
+      "grad_norm": 0.11564800143241882,
+      "learning_rate": 9.799137065043428e-07,
+      "loss": 3.1580835342407227,
+      "step": 10050
+    },
+    {
+      "epoch": 1.9317963496637849,
+      "grad_norm": 0.12115105986595154,
+      "learning_rate": 9.529115647899599e-07,
+      "loss": 3.160675811767578,
+      "step": 10055
+    },
+    {
+      "epoch": 1.9327569644572526,
+      "grad_norm": 0.11305255442857742,
+      "learning_rate": 9.262854841700418e-07,
+      "loss": 3.1640155792236326,
+      "step": 10060
+    },
+    {
+      "epoch": 1.9337175792507204,
+      "grad_norm": 0.11431337147951126,
+      "learning_rate": 9.00035531824933e-07,
+      "loss": 3.162336730957031,
+      "step": 10065
+    },
+    {
+      "epoch": 1.9346781940441882,
+      "grad_norm": 0.10593244433403015,
+      "learning_rate": 8.741617739859708e-07,
+      "loss": 3.1616371154785154,
+      "step": 10070
+    },
+    {
+      "epoch": 1.9356388088376562,
+      "grad_norm": 0.10147248208522797,
+      "learning_rate": 8.486642759353013e-07,
+      "loss": 3.1602996826171874,
+      "step": 10075
+    },
+    {
+      "epoch": 1.936599423631124,
+      "grad_norm": 0.11662466824054718,
+      "learning_rate": 8.23543102005697e-07,
+      "loss": 3.1570274353027346,
+      "step": 10080
+    },
+    {
+      "epoch": 1.937560038424592,
+      "grad_norm": 0.10574644058942795,
+      "learning_rate": 7.9879831558049e-07,
+      "loss": 3.156863975524902,
+      "step": 10085
+    },
+    {
+      "epoch": 1.9385206532180597,
+      "grad_norm": 0.10977693647146225,
+      "learning_rate": 7.744299790933217e-07,
+      "loss": 3.1590835571289064,
+      "step": 10090
+    },
+    {
+      "epoch": 1.9394812680115274,
+      "grad_norm": 0.10733279585838318,
+      "learning_rate": 7.504381540279603e-07,
+      "loss": 3.1599714279174806,
+      "step": 10095
+    },
+    {
+      "epoch": 1.9404418828049952,
+      "grad_norm": 0.1001046821475029,
+      "learning_rate": 7.268229009182836e-07,
+      "loss": 3.158583068847656,
+      "step": 10100
+    },
+    {
+      "epoch": 1.941402497598463,
+      "grad_norm": 0.10647624731063843,
+      "learning_rate": 7.035842793479795e-07,
+      "loss": 3.158983612060547,
+      "step": 10105
+    },
+    {
+      "epoch": 1.9423631123919307,
+      "grad_norm": 0.10598830133676529,
+      "learning_rate": 6.807223479504631e-07,
+      "loss": 3.1631649017333983,
+      "step": 10110
+    },
+    {
+      "epoch": 1.9433237271853987,
+      "grad_norm": 0.10469937324523926,
+      "learning_rate": 6.582371644087592e-07,
+      "loss": 3.1588485717773436,
+      "step": 10115
+    },
+    {
+      "epoch": 1.9442843419788665,
+      "grad_norm": 0.10287413746118546,
+      "learning_rate": 6.361287854552865e-07,
+      "loss": 3.161878967285156,
+      "step": 10120
+    },
+    {
+      "epoch": 1.9452449567723344,
+      "grad_norm": 0.10505318641662598,
+      "learning_rate": 6.143972668717411e-07,
+      "loss": 3.1627128601074217,
+      "step": 10125
+    },
+    {
+      "epoch": 1.9462055715658022,
+      "grad_norm": 0.1040089949965477,
+      "learning_rate": 5.930426634889796e-07,
+      "loss": 3.1591156005859373,
+      "step": 10130
+    },
+    {
+      "epoch": 1.94716618635927,
+      "grad_norm": 0.10172892361879349,
+      "learning_rate": 5.720650291868523e-07,
+      "loss": 3.1622541427612303,
+      "step": 10135
+    },
+    {
+      "epoch": 1.9481268011527377,
+      "grad_norm": 0.10863189399242401,
+      "learning_rate": 5.514644168941041e-07,
+      "loss": 3.161451721191406,
+      "step": 10140
+    },
+    {
+      "epoch": 1.9490874159462055,
+      "grad_norm": 0.10447549819946289,
+      "learning_rate": 5.312408785881573e-07,
+      "loss": 3.1583093643188476,
+      "step": 10145
+    },
+    {
+      "epoch": 1.9500480307396733,
+      "grad_norm": 0.10685736685991287,
+      "learning_rate": 5.113944652951118e-07,
+      "loss": 3.1573591232299805,
+      "step": 10150
+    },
+    {
+      "epoch": 1.9510086455331412,
+      "grad_norm": 0.10184746235609055,
+      "learning_rate": 4.919252270894625e-07,
+      "loss": 3.155690383911133,
+      "step": 10155
+    },
+    {
+      "epoch": 1.951969260326609,
+      "grad_norm": 0.10571934282779694,
+      "learning_rate": 4.7283321309413145e-07,
+      "loss": 3.159967613220215,
+      "step": 10160
+    },
+    {
+      "epoch": 1.952929875120077,
+      "grad_norm": 0.1015496551990509,
+      "learning_rate": 4.541184714802193e-07,
+      "loss": 3.1597368240356447,
+      "step": 10165
+    },
+    {
+      "epoch": 1.9538904899135447,
+      "grad_norm": 0.1008259505033493,
+      "learning_rate": 4.3578104946692114e-07,
+      "loss": 3.156237030029297,
+      "step": 10170
+    },
+    {
+      "epoch": 1.9548511047070125,
+      "grad_norm": 0.10270283371210098,
+      "learning_rate": 4.1782099332146046e-07,
+      "loss": 3.162929916381836,
+      "step": 10175
+    },
+    {
+      "epoch": 1.9558117195004803,
+      "grad_norm": 0.10830524563789368,
+      "learning_rate": 4.002383483588723e-07,
+      "loss": 3.157078170776367,
+      "step": 10180
+    },
+    {
+      "epoch": 1.956772334293948,
+      "grad_norm": 0.09829416126012802,
+      "learning_rate": 3.8303315894200326e-07,
+      "loss": 3.1587419509887695,
+      "step": 10185
+    },
+    {
+      "epoch": 1.9577329490874158,
+      "grad_norm": 0.10378840565681458,
+      "learning_rate": 3.662054684812954e-07,
+      "loss": 3.1602014541625976,
+      "step": 10190
+    },
+    {
+      "epoch": 1.9586935638808838,
+      "grad_norm": 0.10607574880123138,
+      "learning_rate": 3.497553194347358e-07,
+      "loss": 3.1587203979492187,
+      "step": 10195
+    },
+    {
+      "epoch": 1.9596541786743515,
+      "grad_norm": 0.10446880012750626,
+      "learning_rate": 3.3368275330775686e-07,
+      "loss": 3.160129928588867,
+      "step": 10200
+    },
+    {
+      "epoch": 1.9606147934678195,
+      "grad_norm": 0.10566939413547516,
+      "learning_rate": 3.1798781065305314e-07,
+      "loss": 3.162166213989258,
+      "step": 10205
+    },
+    {
+      "epoch": 1.9615754082612873,
+      "grad_norm": 0.10455495119094849,
+      "learning_rate": 3.0267053107061456e-07,
+      "loss": 3.1622539520263673,
+      "step": 10210
+    },
+    {
+      "epoch": 1.962536023054755,
+      "grad_norm": 0.10329185426235199,
+      "learning_rate": 2.877309532074934e-07,
+      "loss": 3.155780029296875,
+      "step": 10215
+    },
+    {
+      "epoch": 1.9634966378482228,
+      "grad_norm": 0.09786950796842575,
+      "learning_rate": 2.7316911475777083e-07,
+      "loss": 3.160215377807617,
+      "step": 10220
+    },
+    {
+      "epoch": 1.9644572526416906,
+      "grad_norm": 0.10430426150560379,
+      "learning_rate": 2.589850524624737e-07,
+      "loss": 3.158669662475586,
+      "step": 10225
+    },
+    {
+      "epoch": 1.9654178674351583,
+      "grad_norm": 0.10500769317150116,
+      "learning_rate": 2.451788021094414e-07,
+      "loss": 3.1637641906738283,
+      "step": 10230
+    },
+    {
+      "epoch": 1.9663784822286263,
+      "grad_norm": 0.09861162304878235,
+      "learning_rate": 2.317503985332425e-07,
+      "loss": 3.1611236572265624,
+      "step": 10235
+    },
+    {
+      "epoch": 1.967339097022094,
+      "grad_norm": 0.09918565303087234,
+      "learning_rate": 2.1869987561514147e-07,
+      "loss": 3.15871639251709,
+      "step": 10240
+    },
+    {
+      "epoch": 1.968299711815562,
+      "grad_norm": 0.1011766865849495,
+      "learning_rate": 2.060272662829321e-07,
+      "loss": 3.16119499206543,
+      "step": 10245
+    },
+    {
+      "epoch": 1.9692603266090298,
+      "grad_norm": 0.11001937091350555,
+      "learning_rate": 1.9373260251092092e-07,
+      "loss": 3.1602630615234375,
+      "step": 10250
+    },
+    {
+      "epoch": 1.9702209414024976,
+      "grad_norm": 0.10449342429637909,
+      "learning_rate": 1.8181591531977736e-07,
+      "loss": 3.160749816894531,
+      "step": 10255
+    },
+    {
+      "epoch": 1.9711815561959654,
+      "grad_norm": 0.09981942176818848,
+      "learning_rate": 1.7027723477656686e-07,
+      "loss": 3.159575653076172,
+      "step": 10260
+    },
+    {
+      "epoch": 1.9721421709894331,
+      "grad_norm": 0.10080744326114655,
+      "learning_rate": 1.5911658999453458e-07,
+      "loss": 3.16530704498291,
+      "step": 10265
+    },
+    {
+      "epoch": 1.973102785782901,
+      "grad_norm": 0.09806544333696365,
+      "learning_rate": 1.4833400913313864e-07,
+      "loss": 3.1608566284179687,
+      "step": 10270
+    },
+    {
+      "epoch": 1.9740634005763689,
+      "grad_norm": 0.10161542147397995,
+      "learning_rate": 1.379295193979335e-07,
+      "loss": 3.1610599517822267,
+      "step": 10275
+    },
+    {
+      "epoch": 1.9750240153698368,
+      "grad_norm": 0.10026436299085617,
+      "learning_rate": 1.2790314704052008e-07,
+      "loss": 3.1599550247192383,
+      "step": 10280
+    },
+    {
+      "epoch": 1.9759846301633046,
+      "grad_norm": 0.09924504905939102,
+      "learning_rate": 1.1825491735846237e-07,
+      "loss": 3.1612056732177733,
+      "step": 10285
+    },
+    {
+      "epoch": 1.9769452449567724,
+      "grad_norm": 0.10245411098003387,
+      "learning_rate": 1.0898485469523766e-07,
+      "loss": 3.1604587554931642,
+      "step": 10290
+    },
+    {
+      "epoch": 1.9779058597502401,
+      "grad_norm": 0.09821955114603043,
+      "learning_rate": 1.0009298244016972e-07,
+      "loss": 3.1620269775390626,
+      "step": 10295
+    },
+    {
+      "epoch": 1.978866474543708,
+      "grad_norm": 0.09943889081478119,
+      "learning_rate": 9.15793230283457e-08,
+      "loss": 3.160098648071289,
+      "step": 10300
+    },
+    {
+      "epoch": 1.9798270893371757,
+      "grad_norm": 0.09800975024700165,
+      "learning_rate": 8.3443897940616e-08,
+      "loss": 3.1579994201660155,
+      "step": 10305
+    },
+    {
+      "epoch": 1.9807877041306436,
+      "grad_norm": 0.0975160002708435,
+      "learning_rate": 7.568672770349449e-08,
+      "loss": 3.160677909851074,
+      "step": 10310
+    },
+    {
+      "epoch": 1.9817483189241114,
+      "grad_norm": 0.09866613149642944,
+      "learning_rate": 6.830783188910838e-08,
+      "loss": 3.15645809173584,
+      "step": 10315
+    },
+    {
+      "epoch": 1.9827089337175794,
+      "grad_norm": 0.09947178512811661,
+      "learning_rate": 6.130722911516505e-08,
+      "loss": 3.1583110809326174,
+      "step": 10320
+    },
+    {
+      "epoch": 1.9836695485110472,
+      "grad_norm": 0.09944622218608856,
+      "learning_rate": 5.468493704491872e-08,
+      "loss": 3.1639808654785155,
+      "step": 10325
+    },
+    {
+      "epoch": 1.984630163304515,
+      "grad_norm": 0.09704981744289398,
+      "learning_rate": 4.844097238708711e-08,
+      "loss": 3.1581130981445313,
+      "step": 10330
+    },
+    {
+      "epoch": 1.9855907780979827,
+      "grad_norm": 0.09650828689336777,
+      "learning_rate": 4.257535089581821e-08,
+      "loss": 3.156038284301758,
+      "step": 10335
+    },
+    {
+      "epoch": 1.9865513928914504,
+      "grad_norm": 0.09726572781801224,
+      "learning_rate": 3.70880873707069e-08,
+      "loss": 3.1570310592651367,
+      "step": 10340
+    },
+    {
+      "epoch": 1.9875120076849182,
+      "grad_norm": 0.09816308319568634,
+      "learning_rate": 3.1979195656645084e-08,
+      "loss": 3.160786819458008,
+      "step": 10345
+    },
+    {
+      "epoch": 1.9884726224783862,
+      "grad_norm": 0.09846552461385727,
+      "learning_rate": 2.72486886439216e-08,
+      "loss": 3.1611724853515626,
+      "step": 10350
+    },
+    {
+      "epoch": 1.989433237271854,
+      "grad_norm": 0.09757398068904877,
+      "learning_rate": 2.289657826807234e-08,
+      "loss": 3.1581047058105467,
+      "step": 10355
+    },
+    {
+      "epoch": 1.990393852065322,
+      "grad_norm": 0.10126172751188278,
+      "learning_rate": 1.8922875509930212e-08,
+      "loss": 3.160317611694336,
+      "step": 10360
+    },
+    {
+      "epoch": 1.9913544668587897,
+      "grad_norm": 0.09841560572385788,
+      "learning_rate": 1.532759039554188e-08,
+      "loss": 3.158650207519531,
+      "step": 10365
+    },
+    {
+      "epoch": 1.9923150816522575,
+      "grad_norm": 0.09537240862846375,
+      "learning_rate": 1.2110731996201051e-08,
+      "loss": 3.1590465545654296,
+      "step": 10370
+    },
+    {
+      "epoch": 1.9932756964457252,
+      "grad_norm": 0.09531023353338242,
+      "learning_rate": 9.272308428348586e-09,
+      "loss": 3.159567642211914,
+      "step": 10375
+    },
+    {
+      "epoch": 1.994236311239193,
+      "grad_norm": 0.10250640660524368,
+      "learning_rate": 6.812326853639083e-09,
+      "loss": 3.1602352142333983,
+      "step": 10380
+    },
+    {
+      "epoch": 1.9951969260326607,
+      "grad_norm": 0.09563779830932617,
+      "learning_rate": 4.730793478840977e-09,
+      "loss": 3.157382583618164,
+      "step": 10385
+    },
+    {
+      "epoch": 1.9961575408261287,
+      "grad_norm": 0.09357914328575134,
+      "learning_rate": 3.0277135558864906e-09,
+      "loss": 3.1614322662353516,
+      "step": 10390
+    },
+    {
+      "epoch": 1.9971181556195965,
+      "grad_norm": 0.09247033298015594,
+      "learning_rate": 1.7030913818050217e-09,
+      "loss": 3.161212921142578,
+      "step": 10395
+    },
+    {
+      "epoch": 1.9980787704130645,
+      "grad_norm": 0.09770449995994568,
+      "learning_rate": 7.569302987897597e-10,
+      "loss": 3.1566755294799806,
+      "step": 10400
+    },
+    {
+      "epoch": 1.9990393852065322,
+      "grad_norm": 0.0940788984298706,
+      "learning_rate": 1.8923269406445618e-10,
+      "loss": 3.157200050354004,
+      "step": 10405
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.10163447260856628,
+      "learning_rate": 0.0,
+      "loss": 3.162412643432617,
+      "step": 10410
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 10410,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.9541356371946714e+20,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}