diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12832 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999452264884702,
+  "eval_steps": 500,
+  "global_step": 9128,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00010954702305964835,
+      "grad_norm": 11.6875,
+      "learning_rate": 2.190580503833516e-08,
+      "loss": 1.6695,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005477351152982418,
+      "grad_norm": 10.5,
+      "learning_rate": 1.095290251916758e-07,
+      "loss": 1.7344,
+      "step": 5
+    },
+    {
+      "epoch": 0.0010954702305964836,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.190580503833516e-07,
+      "loss": 1.6283,
+      "step": 10
+    },
+    {
+      "epoch": 0.0016432053458947254,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.285870755750274e-07,
+      "loss": 1.6588,
+      "step": 15
+    },
+    {
+      "epoch": 0.0021909404611929672,
+      "grad_norm": 9.4375,
+      "learning_rate": 4.381161007667032e-07,
+      "loss": 1.6831,
+      "step": 20
+    },
+    {
+      "epoch": 0.002738675576491209,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.47645125958379e-07,
+      "loss": 1.6779,
+      "step": 25
+    },
+    {
+      "epoch": 0.003286410691789451,
+      "grad_norm": 11.25,
+      "learning_rate": 6.571741511500548e-07,
+      "loss": 1.7032,
+      "step": 30
+    },
+    {
+      "epoch": 0.0038341458070876924,
+      "grad_norm": 9.125,
+      "learning_rate": 7.667031763417306e-07,
+      "loss": 1.6961,
+      "step": 35
+    },
+    {
+      "epoch": 0.0043818809223859344,
+      "grad_norm": 9.5625,
+      "learning_rate": 8.762322015334064e-07,
+      "loss": 1.6005,
+      "step": 40
+    },
+    {
+      "epoch": 0.004929616037684176,
+      "grad_norm": 11.4375,
+      "learning_rate": 9.857612267250823e-07,
+      "loss": 1.6734,
+      "step": 45
+    },
+    {
+      "epoch": 0.005477351152982418,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.095290251916758e-06,
+      "loss": 1.6523,
+      "step": 50
+    },
+    {
+      "epoch": 0.00602508626828066,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2048192771084338e-06,
+      "loss": 1.6274,
+      "step": 55
+    },
+    {
+      "epoch": 0.006572821383578902,
+      "grad_norm": 9.25,
+      "learning_rate": 1.3143483023001096e-06,
+      "loss": 1.6391,
+      "step": 60
+    },
+    {
+      "epoch": 0.007120556498877143,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.4238773274917855e-06,
+      "loss": 1.6723,
+      "step": 65
+    },
+    {
+      "epoch": 0.007668291614175385,
+      "grad_norm": 8.75,
+      "learning_rate": 1.5334063526834611e-06,
+      "loss": 1.616,
+      "step": 70
+    },
+    {
+      "epoch": 0.008216026729473627,
+      "grad_norm": 9.5,
+      "learning_rate": 1.642935377875137e-06,
+      "loss": 1.6565,
+      "step": 75
+    },
+    {
+      "epoch": 0.008763761844771869,
+      "grad_norm": 8.875,
+      "learning_rate": 1.7524644030668128e-06,
+      "loss": 1.5556,
+      "step": 80
+    },
+    {
+      "epoch": 0.009311496960070111,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8619934282584886e-06,
+      "loss": 1.5898,
+      "step": 85
+    },
+    {
+      "epoch": 0.009859232075368351,
+      "grad_norm": 8.75,
+      "learning_rate": 1.9715224534501647e-06,
+      "loss": 1.5372,
+      "step": 90
+    },
+    {
+      "epoch": 0.010406967190666593,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.0810514786418403e-06,
+      "loss": 1.5206,
+      "step": 95
+    },
+    {
+      "epoch": 0.010954702305964835,
+      "grad_norm": 5.65625,
+      "learning_rate": 2.190580503833516e-06,
+      "loss": 1.5475,
+      "step": 100
+    },
+    {
+      "epoch": 0.011502437421263077,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.300109529025192e-06,
+      "loss": 1.5011,
+      "step": 105
+    },
+    {
+      "epoch": 0.01205017253656132,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.4096385542168676e-06,
+      "loss": 1.4805,
+      "step": 110
+    },
+    {
+      "epoch": 0.012597907651859561,
+      "grad_norm": 4.5625,
+      "learning_rate": 2.5191675794085437e-06,
+      "loss": 1.4625,
+      "step": 115
+    },
+    {
+      "epoch": 0.013145642767157803,
+      "grad_norm": 4.25,
+      "learning_rate": 2.6286966046002193e-06,
+      "loss": 1.4329,
+      "step": 120
+    },
+    {
+      "epoch": 0.013693377882456044,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.7382256297918953e-06,
+      "loss": 1.4614,
+      "step": 125
+    },
+    {
+      "epoch": 0.014241112997754286,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.847754654983571e-06,
+      "loss": 1.43,
+      "step": 130
+    },
+    {
+      "epoch": 0.014788848113052528,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.957283680175247e-06,
+      "loss": 1.4254,
+      "step": 135
+    },
+    {
+      "epoch": 0.01533658322835077,
+      "grad_norm": 2.375,
+      "learning_rate": 3.0668127053669222e-06,
+      "loss": 1.3846,
+      "step": 140
+    },
+    {
+      "epoch": 0.01588431834364901,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1763417305585983e-06,
+      "loss": 1.3968,
+      "step": 145
+    },
+    {
+      "epoch": 0.016432053458947254,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.285870755750274e-06,
+      "loss": 1.3592,
+      "step": 150
+    },
+    {
+      "epoch": 0.016979788574245494,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.39539978094195e-06,
+      "loss": 1.3269,
+      "step": 155
+    },
+    {
+      "epoch": 0.017527523689543738,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.5049288061336256e-06,
+      "loss": 1.3384,
+      "step": 160
+    },
+    {
+      "epoch": 0.018075258804841978,
+      "grad_norm": 1.734375,
+      "learning_rate": 3.6144578313253016e-06,
+      "loss": 1.3149,
+      "step": 165
+    },
+    {
+      "epoch": 0.018622993920140222,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.7239868565169773e-06,
+      "loss": 1.331,
+      "step": 170
+    },
+    {
+      "epoch": 0.019170729035438462,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.833515881708653e-06,
+      "loss": 1.3452,
+      "step": 175
+    },
+    {
+      "epoch": 0.019718464150736702,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.943044906900329e-06,
+      "loss": 1.2923,
+      "step": 180
+    },
+    {
+      "epoch": 0.020266199266034946,
+      "grad_norm": 1.734375,
+      "learning_rate": 4.0525739320920046e-06,
+      "loss": 1.351,
+      "step": 185
+    },
+    {
+      "epoch": 0.020813934381333186,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.162102957283681e-06,
+      "loss": 1.2837,
+      "step": 190
+    },
+    {
+      "epoch": 0.02136166949663143,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.271631982475356e-06,
+      "loss": 1.2891,
+      "step": 195
+    },
+    {
+      "epoch": 0.02190940461192967,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.381161007667032e-06,
+      "loss": 1.3041,
+      "step": 200
+    },
+    {
+      "epoch": 0.022457139727227914,
+      "grad_norm": 1.625,
+      "learning_rate": 4.490690032858708e-06,
+      "loss": 1.2525,
+      "step": 205
+    },
+    {
+      "epoch": 0.023004874842526155,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.600219058050384e-06,
+      "loss": 1.3449,
+      "step": 210
+    },
+    {
+      "epoch": 0.023552609957824395,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.709748083242059e-06,
+      "loss": 1.2768,
+      "step": 215
+    },
+    {
+      "epoch": 0.02410034507312264,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.819277108433735e-06,
+      "loss": 1.28,
+      "step": 220
+    },
+    {
+      "epoch": 0.02464808018842088,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.928806133625411e-06,
+      "loss": 1.2782,
+      "step": 225
+    },
+    {
+      "epoch": 0.025195815303719123,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.038335158817087e-06,
+      "loss": 1.2688,
+      "step": 230
+    },
+    {
+      "epoch": 0.025743550419017363,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.1478641840087625e-06,
+      "loss": 1.2547,
+      "step": 235
+    },
+    {
+      "epoch": 0.026291285534315607,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.257393209200439e-06,
+      "loss": 1.2791,
+      "step": 240
+    },
+    {
+      "epoch": 0.026839020649613847,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.366922234392115e-06,
+      "loss": 1.2681,
+      "step": 245
+    },
+    {
+      "epoch": 0.027386755764912087,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.476451259583791e-06,
+      "loss": 1.2936,
+      "step": 250
+    },
+    {
+      "epoch": 0.02793449088021033,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.585980284775466e-06,
+      "loss": 1.2555,
+      "step": 255
+    },
+    {
+      "epoch": 0.02848222599550857,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.695509309967142e-06,
+      "loss": 1.2712,
+      "step": 260
+    },
+    {
+      "epoch": 0.029029961110806815,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.805038335158818e-06,
+      "loss": 1.2957,
+      "step": 265
+    },
+    {
+      "epoch": 0.029577696226105055,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.914567360350494e-06,
+      "loss": 1.2894,
+      "step": 270
+    },
+    {
+      "epoch": 0.030125431341403296,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.02409638554217e-06,
+      "loss": 1.2609,
+      "step": 275
+    },
+    {
+      "epoch": 0.03067316645670154,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.1336254107338444e-06,
+      "loss": 1.2343,
+      "step": 280
+    },
+    {
+      "epoch": 0.03122090157199978,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.2431544359255205e-06,
+      "loss": 1.2609,
+      "step": 285
+    },
+    {
+      "epoch": 0.03176863668729802,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.3526834611171965e-06,
+      "loss": 1.228,
+      "step": 290
+    },
+    {
+      "epoch": 0.032316371802596264,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.462212486308872e-06,
+      "loss": 1.2499,
+      "step": 295
+    },
+    {
+      "epoch": 0.03286410691789451,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.571741511500548e-06,
+      "loss": 1.2826,
+      "step": 300
+    },
+    {
+      "epoch": 0.03341184203319275,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.681270536692224e-06,
+      "loss": 1.2581,
+      "step": 305
+    },
+    {
+      "epoch": 0.03395957714849099,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.7907995618839e-06,
+      "loss": 1.2385,
+      "step": 310
+    },
+    {
+      "epoch": 0.03450731226378923,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.900328587075575e-06,
+      "loss": 1.2487,
+      "step": 315
+    },
+    {
+      "epoch": 0.035055047379087476,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.009857612267251e-06,
+      "loss": 1.2359,
+      "step": 320
+    },
+    {
+      "epoch": 0.03560278249438571,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.119386637458927e-06,
+      "loss": 1.2917,
+      "step": 325
+    },
+    {
+      "epoch": 0.036150517609683956,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.228915662650603e-06,
+      "loss": 1.2511,
+      "step": 330
+    },
+    {
+      "epoch": 0.0366982527249822,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.3384446878422785e-06,
+      "loss": 1.2733,
+      "step": 335
+    },
+    {
+      "epoch": 0.037245987840280444,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.4479737130339545e-06,
+      "loss": 1.25,
+      "step": 340
+    },
+    {
+      "epoch": 0.03779372295557868,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.5575027382256306e-06,
+      "loss": 1.2793,
+      "step": 345
+    },
+    {
+      "epoch": 0.038341458070876924,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.667031763417307e-06,
+      "loss": 1.2786,
+      "step": 350
+    },
+    {
+      "epoch": 0.03888919318617517,
+      "grad_norm": 1.5,
+      "learning_rate": 7.776560788608982e-06,
+      "loss": 1.2847,
+      "step": 355
+    },
+    {
+      "epoch": 0.039436928301473405,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.886089813800659e-06,
+      "loss": 1.3118,
+      "step": 360
+    },
+    {
+      "epoch": 0.03998466341677165,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.995618838992334e-06,
+      "loss": 1.2672,
+      "step": 365
+    },
+    {
+      "epoch": 0.04053239853206989,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.105147864184009e-06,
+      "loss": 1.2752,
+      "step": 370
+    },
+    {
+      "epoch": 0.041080133647368136,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.214676889375686e-06,
+      "loss": 1.2129,
+      "step": 375
+    },
+    {
+      "epoch": 0.04162786876266637,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.324205914567361e-06,
+      "loss": 1.2755,
+      "step": 380
+    },
+    {
+      "epoch": 0.04217560387796462,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.433734939759038e-06,
+      "loss": 1.1983,
+      "step": 385
+    },
+    {
+      "epoch": 0.04272333899326286,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.543263964950712e-06,
+      "loss": 1.2275,
+      "step": 390
+    },
+    {
+      "epoch": 0.0432710741085611,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.652792990142389e-06,
+      "loss": 1.1843,
+      "step": 395
+    },
+    {
+      "epoch": 0.04381880922385934,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.762322015334064e-06,
+      "loss": 1.2471,
+      "step": 400
+    },
+    {
+      "epoch": 0.044366544339157585,
+      "grad_norm": 1.5,
+      "learning_rate": 8.871851040525739e-06,
+      "loss": 1.2367,
+      "step": 405
+    },
+    {
+      "epoch": 0.04491427945445583,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.981380065717416e-06,
+      "loss": 1.2504,
+      "step": 410
+    },
+    {
+      "epoch": 0.045462014569754065,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.3088,
+      "step": 415
+    },
+    {
+      "epoch": 0.04600974968505231,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.200438116100768e-06,
+      "loss": 1.2398,
+      "step": 420
+    },
+    {
+      "epoch": 0.04655748480035055,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.309967141292443e-06,
+      "loss": 1.2922,
+      "step": 425
+    },
+    {
+      "epoch": 0.04710521991564879,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.419496166484118e-06,
+      "loss": 1.2491,
+      "step": 430
+    },
+    {
+      "epoch": 0.04765295503094703,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.529025191675795e-06,
+      "loss": 1.1853,
+      "step": 435
+    },
+    {
+      "epoch": 0.04820069014624528,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.63855421686747e-06,
+      "loss": 1.2793,
+      "step": 440
+    },
+    {
+      "epoch": 0.04874842526154352,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.748083242059146e-06,
+      "loss": 1.2385,
+      "step": 445
+    },
+    {
+      "epoch": 0.04929616037684176,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.857612267250823e-06,
+      "loss": 1.2048,
+      "step": 450
+    },
+    {
+      "epoch": 0.04984389549214,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.967141292442498e-06,
+      "loss": 1.208,
+      "step": 455
+    },
+    {
+      "epoch": 0.050391630607438245,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.0076670317634175e-05,
+      "loss": 1.2326,
+      "step": 460
+    },
+    {
+      "epoch": 0.05093936572273648,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.0186199342825848e-05,
+      "loss": 1.195,
+      "step": 465
+    },
+    {
+      "epoch": 0.051487100838034726,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.0295728368017525e-05,
+      "loss": 1.1941,
+      "step": 470
+    },
+    {
+      "epoch": 0.05203483595333297,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.04052573932092e-05,
+      "loss": 1.2478,
+      "step": 475
+    },
+    {
+      "epoch": 0.05258257106863121,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0514786418400877e-05,
+      "loss": 1.2514,
+      "step": 480
+    },
+    {
+      "epoch": 0.05313030618392945,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0624315443592552e-05,
+      "loss": 1.2535,
+      "step": 485
+    },
+    {
+      "epoch": 0.053678041299227694,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.073384446878423e-05,
+      "loss": 1.232,
+      "step": 490
+    },
+    {
+      "epoch": 0.05422577641452594,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0843373493975904e-05,
+      "loss": 1.3235,
+      "step": 495
+    },
+    {
+      "epoch": 0.054773511529824175,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.0952902519167581e-05,
+      "loss": 1.2356,
+      "step": 500
+    },
+    {
+      "epoch": 0.05532124664512242,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.1062431544359255e-05,
+      "loss": 1.1983,
+      "step": 505
+    },
+    {
+      "epoch": 0.05586898176042066,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.1171960569550932e-05,
+      "loss": 1.2143,
+      "step": 510
+    },
+    {
+      "epoch": 0.0564167168757189,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.1281489594742607e-05,
+      "loss": 1.2818,
+      "step": 515
+    },
+    {
+      "epoch": 0.05696445199101714,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1391018619934284e-05,
+      "loss": 1.1754,
+      "step": 520
+    },
+    {
+      "epoch": 0.057512187106315386,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.1500547645125959e-05,
+      "loss": 1.2255,
+      "step": 525
+    },
+    {
+      "epoch": 0.05805992222161363,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.1610076670317636e-05,
+      "loss": 1.162,
+      "step": 530
+    },
+    {
+      "epoch": 0.05860765733691187,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1719605695509311e-05,
+      "loss": 1.2283,
+      "step": 535
+    },
+    {
+      "epoch": 0.05915539245221011,
+      "grad_norm": 1.5,
+      "learning_rate": 1.1829134720700988e-05,
+      "loss": 1.2758,
+      "step": 540
+    },
+    {
+      "epoch": 0.059703127567508354,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.1938663745892662e-05,
+      "loss": 1.2252,
+      "step": 545
+    },
+    {
+      "epoch": 0.06025086268280659,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.204819277108434e-05,
+      "loss": 1.2171,
+      "step": 550
+    },
+    {
+      "epoch": 0.060798597798104835,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.2157721796276014e-05,
+      "loss": 1.2346,
+      "step": 555
+    },
+    {
+      "epoch": 0.06134633291340308,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.2267250821467689e-05,
+      "loss": 1.2295,
+      "step": 560
+    },
+    {
+      "epoch": 0.06189406802870132,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.2376779846659366e-05,
+      "loss": 1.2068,
+      "step": 565
+    },
+    {
+      "epoch": 0.06244180314399956,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.2486308871851041e-05,
+      "loss": 1.1862,
+      "step": 570
+    },
+    {
+      "epoch": 0.0629895382592978,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.2595837897042718e-05,
+      "loss": 1.219,
+      "step": 575
+    },
+    {
+      "epoch": 0.06353727337459604,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.2705366922234393e-05,
+      "loss": 1.2089,
+      "step": 580
+    },
+    {
+      "epoch": 0.06408500848989429,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.281489594742607e-05,
+      "loss": 1.2738,
+      "step": 585
+    },
+    {
+      "epoch": 0.06463274360519253,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.2924424972617743e-05,
+      "loss": 1.2507,
+      "step": 590
+    },
+    {
+      "epoch": 0.06518047872049076,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.303395399780942e-05,
+      "loss": 1.2554,
+      "step": 595
+    },
+    {
+      "epoch": 0.06572821383578901,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.3143483023001096e-05,
+      "loss": 1.2269,
+      "step": 600
+    },
+    {
+      "epoch": 0.06627594895108725,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.3253012048192772e-05,
+      "loss": 1.2278,
+      "step": 605
+    },
+    {
+      "epoch": 0.0668236840663855,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.3362541073384448e-05,
+      "loss": 1.197,
+      "step": 610
+    },
+    {
+      "epoch": 0.06737141918168374,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.3472070098576125e-05,
+      "loss": 1.2462,
+      "step": 615
+    },
+    {
+      "epoch": 0.06791915429698198,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.35815991237678e-05,
+      "loss": 1.1692,
+      "step": 620
+    },
+    {
+      "epoch": 0.06846688941228023,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.3691128148959477e-05,
+      "loss": 1.2331,
+      "step": 625
+    },
+    {
+      "epoch": 0.06901462452757846,
+      "grad_norm": 1.375,
+      "learning_rate": 1.380065717415115e-05,
+      "loss": 1.2679,
+      "step": 630
+    },
+    {
+      "epoch": 0.0695623596428767,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.3910186199342827e-05,
+      "loss": 1.212,
+      "step": 635
+    },
+    {
+      "epoch": 0.07011009475817495,
+      "grad_norm": 1.375,
+      "learning_rate": 1.4019715224534502e-05,
+      "loss": 1.2085,
+      "step": 640
+    },
+    {
+      "epoch": 0.07065782987347319,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.412924424972618e-05,
+      "loss": 1.2214,
+      "step": 645
+    },
+    {
+      "epoch": 0.07120556498877142,
+      "grad_norm": 1.375,
+      "learning_rate": 1.4238773274917854e-05,
+      "loss": 1.2937,
+      "step": 650
+    },
+    {
+      "epoch": 0.07175330010406968,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.4348302300109531e-05,
+      "loss": 1.2745,
+      "step": 655
+    },
+    {
+      "epoch": 0.07230103521936791,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.4457831325301207e-05,
+      "loss": 1.1891,
+      "step": 660
+    },
+    {
+      "epoch": 0.07284877033466615,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4567360350492883e-05,
+      "loss": 1.2059,
+      "step": 665
+    },
+    {
+      "epoch": 0.0733965054499644,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.4676889375684557e-05,
+      "loss": 1.2321,
+      "step": 670
+    },
+    {
+      "epoch": 0.07394424056526264,
+      "grad_norm": 1.375,
+      "learning_rate": 1.4786418400876232e-05,
+      "loss": 1.2003,
+      "step": 675
+    },
+    {
+      "epoch": 0.07449197568056089,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.4895947426067909e-05,
+      "loss": 1.2113,
+      "step": 680
+    },
+    {
+      "epoch": 0.07503971079585912,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5005476451259584e-05,
+      "loss": 1.2612,
+      "step": 685
+    },
+    {
+      "epoch": 0.07558744591115736,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5115005476451261e-05,
+      "loss": 1.2022,
+      "step": 690
+    },
+    {
+      "epoch": 0.07613518102645561,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.5224534501642936e-05,
+      "loss": 1.1903,
+      "step": 695
+    },
+    {
+      "epoch": 0.07668291614175385,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.5334063526834613e-05,
+      "loss": 1.1864,
+      "step": 700
+    },
+    {
+      "epoch": 0.07723065125705209,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.5443592552026287e-05,
+      "loss": 1.1988,
+      "step": 705
+    },
+    {
+      "epoch": 0.07777838637235034,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5553121577217964e-05,
+      "loss": 1.2798,
+      "step": 710
+    },
+    {
+      "epoch": 0.07832612148764857,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.566265060240964e-05,
+      "loss": 1.2141,
+      "step": 715
+    },
+    {
+      "epoch": 0.07887385660294681,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5772179627601317e-05,
+      "loss": 1.265,
+      "step": 720
+    },
+    {
+      "epoch": 0.07942159171824506,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.588170865279299e-05,
+      "loss": 1.2063,
+      "step": 725
+    },
+    {
+      "epoch": 0.0799693268335433,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.5991237677984668e-05,
+      "loss": 1.192,
+      "step": 730
+    },
+    {
+      "epoch": 0.08051706194884153,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.610076670317634e-05,
+      "loss": 1.2135,
+      "step": 735
+    },
+    {
+      "epoch": 0.08106479706413978,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.6210295728368018e-05,
+      "loss": 1.1986,
+      "step": 740
+    },
+    {
+      "epoch": 0.08161253217943802,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6319824753559695e-05,
+      "loss": 1.2055,
+      "step": 745
+    },
+    {
+      "epoch": 0.08216026729473627,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.6429353778751372e-05,
+      "loss": 1.2332,
+      "step": 750
+    },
+    {
+      "epoch": 0.08270800241003451,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6538882803943046e-05,
+      "loss": 1.2425,
+      "step": 755
+    },
+    {
+      "epoch": 0.08325573752533275,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6648411829134722e-05,
+      "loss": 1.2201,
+      "step": 760
+    },
+    {
+      "epoch": 0.083803472640631,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.67579408543264e-05,
+      "loss": 1.2385,
+      "step": 765
+    },
+    {
+      "epoch": 0.08435120775592923,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6867469879518076e-05,
+      "loss": 1.1299,
+      "step": 770
+    },
+    {
+      "epoch": 0.08489894287122747,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.697699890470975e-05,
+      "loss": 1.212,
+      "step": 775
+    },
+    {
+      "epoch": 0.08544667798652572,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.7086527929901423e-05,
+      "loss": 1.2242,
+      "step": 780
+    },
+    {
+      "epoch": 0.08599441310182396,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.71960569550931e-05,
+      "loss": 1.2346,
+      "step": 785
+    },
+    {
+      "epoch": 0.0865421482171222,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7305585980284777e-05,
+      "loss": 1.2316,
+      "step": 790
+    },
+    {
+      "epoch": 0.08708988333242045,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7415115005476454e-05,
+      "loss": 1.1931,
+      "step": 795
+    },
+    {
+      "epoch": 0.08763761844771868,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7524644030668127e-05,
+      "loss": 1.2037,
+      "step": 800
+    },
+    {
+      "epoch": 0.08818535356301692,
+      "grad_norm": 1.375,
+      "learning_rate": 1.7634173055859804e-05,
+      "loss": 1.2317,
+      "step": 805
+    },
+    {
+      "epoch": 0.08873308867831517,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7743702081051478e-05,
+      "loss": 1.2425,
+      "step": 810
+    },
+    {
+      "epoch": 0.0892808237936134,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7853231106243155e-05,
+      "loss": 1.2056,
+      "step": 815
+    },
+    {
+      "epoch": 0.08982855890891166,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.796276013143483e-05,
+      "loss": 1.1954,
+      "step": 820
+    },
+    {
+      "epoch": 0.0903762940242099,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.807228915662651e-05,
+      "loss": 1.2034,
+      "step": 825
+    },
+    {
+      "epoch": 0.09092402913950813,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.1815,
+      "step": 830
+    },
+    {
+      "epoch": 0.09147176425480638,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.829134720700986e-05,
+      "loss": 1.2272,
+      "step": 835
+    },
+    {
+      "epoch": 0.09201949937010462,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.8400876232201536e-05,
+      "loss": 1.2151,
+      "step": 840
+    },
+    {
+      "epoch": 0.09256723448540286,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8510405257393213e-05,
+      "loss": 1.2056,
+      "step": 845
+    },
+    {
+      "epoch": 0.0931149696007011,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.8619934282584886e-05,
+      "loss": 1.2219,
+      "step": 850
+    },
+    {
+      "epoch": 0.09366270471599934,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8729463307776563e-05,
+      "loss": 1.1675,
+      "step": 855
+    },
+    {
+      "epoch": 0.09421043983129758,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.8838992332968237e-05,
+      "loss": 1.2177,
+      "step": 860
+    },
+    {
+      "epoch": 0.09475817494659583,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.8948521358159914e-05,
+      "loss": 1.2518,
+      "step": 865
+    },
+    {
+      "epoch": 0.09530591006189407,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.905805038335159e-05,
+      "loss": 1.2463,
+      "step": 870
+    },
+    {
+      "epoch": 0.0958536451771923,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9167579408543267e-05,
+      "loss": 1.1985,
+      "step": 875
+    },
+    {
+      "epoch": 0.09640138029249055,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.927710843373494e-05,
+      "loss": 1.2317,
+      "step": 880
+    },
+    {
+      "epoch": 0.09694911540778879,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9386637458926618e-05,
+      "loss": 1.2383,
+      "step": 885
+    },
+    {
+      "epoch": 0.09749685052308704,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.949616648411829e-05,
+      "loss": 1.1178,
+      "step": 890
+    },
+    {
+      "epoch": 0.09804458563838528,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9605695509309968e-05,
+      "loss": 1.2037,
+      "step": 895
+    },
+    {
+      "epoch": 0.09859232075368352,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.9715224534501645e-05,
+      "loss": 1.1711,
+      "step": 900
+    },
+    {
+      "epoch": 0.09914005586898177,
+      "grad_norm": 1.25,
+      "learning_rate": 1.982475355969332e-05,
+      "loss": 1.223,
+      "step": 905
+    },
+    {
+      "epoch": 0.09968779098428,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9934282584884995e-05,
+      "loss": 1.2393,
+      "step": 910
+    },
+    {
+      "epoch": 0.10023552609957824,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.9999997075076013e-05,
+      "loss": 1.1597,
+      "step": 915
+    },
+    {
+      "epoch": 0.10078326121487649,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.999996416970079e-05,
+      "loss": 1.2222,
+      "step": 920
+    },
+    {
+      "epoch": 0.10133099633017473,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9999894702916073e-05,
+      "loss": 1.1993,
+      "step": 925
+    },
+    {
+      "epoch": 0.10187873144547296,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9999788674975834e-05,
+      "loss": 1.2389,
+      "step": 930
+    },
+    {
+      "epoch": 0.10242646656077121,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.9999646086267734e-05,
+      "loss": 1.1934,
+      "step": 935
+    },
+    {
+      "epoch": 0.10297420167606945,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9999466937313098e-05,
+      "loss": 1.2261,
+      "step": 940
+    },
+    {
+      "epoch": 0.10352193679136769,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.9999251228766922e-05,
+      "loss": 1.1829,
+      "step": 945
+    },
+    {
+      "epoch": 0.10406967190666594,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.999899896141787e-05,
+      "loss": 1.2095,
+      "step": 950
+    },
+    {
+      "epoch": 0.10461740702196418,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9998710136188267e-05,
+      "loss": 1.1984,
+      "step": 955
+    },
+    {
+      "epoch": 0.10516514213726243,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.999838475413411e-05,
+      "loss": 1.1844,
+      "step": 960
+    },
+    {
+      "epoch": 0.10571287725256066,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.9998022816445037e-05,
+      "loss": 1.2157,
+      "step": 965
+    },
+    {
+      "epoch": 0.1062606123678589,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.999762432444435e-05,
+      "loss": 1.1304,
+      "step": 970
+    },
+    {
+      "epoch": 0.10680834748315715,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9997189279589003e-05,
+      "loss": 1.163,
+      "step": 975
+    },
+    {
+      "epoch": 0.10735608259845539,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9996717683469582e-05,
+      "loss": 1.1851,
+      "step": 980
+    },
+    {
+      "epoch": 0.10790381771375362,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9996209537810317e-05,
+      "loss": 1.211,
+      "step": 985
+    },
+    {
+      "epoch": 0.10845155282905188,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9995664844469064e-05,
+      "loss": 1.1801,
+      "step": 990
+    },
+    {
+      "epoch": 0.10899928794435011,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9995083605437312e-05,
+      "loss": 1.1569,
+      "step": 995
+    },
+    {
+      "epoch": 0.10954702305964835,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9994465822840152e-05,
+      "loss": 1.2466,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1100947581749466,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.99938114989363e-05,
+      "loss": 1.1697,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11064249329024484,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9993120636118055e-05,
+      "loss": 1.2242,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11119022840554307,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.999239323691133e-05,
+      "loss": 1.1978,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11173796352084132,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.99916293039756e-05,
+      "loss": 1.2025,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11228569863613956,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.999082884010393e-05,
+      "loss": 1.2639,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1128334337514378,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.998999184822293e-05,
+      "loss": 1.1972,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11338116886673605,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.9989118331392775e-05,
+      "loss": 1.2215,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11392890398203429,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.998820829280718e-05,
+      "loss": 1.2178,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11447663909733254,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.998726173579338e-05,
+      "loss": 1.1748,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11502437421263077,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9986278663812137e-05,
+      "loss": 1.212,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11557210932792901,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.998525908045771e-05,
+      "loss": 1.1583,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11611984444322726,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.998420298945786e-05,
+      "loss": 1.1714,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1166675795585255,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.998311039467382e-05,
+      "loss": 1.2036,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11721531467382373,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.9981981300100267e-05,
+      "loss": 1.1606,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11776304978912198,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9980815709865365e-05,
+      "loss": 1.1495,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11831078490442022,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9979613628230683e-05,
+      "loss": 1.1872,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11885852001971846,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9978375059591214e-05,
+      "loss": 1.2167,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11940625513501671,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.997710000847536e-05,
+      "loss": 1.1988,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11995399025031495,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9975788479544908e-05,
+      "loss": 1.1961,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12050172536561318,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9974440477595e-05,
+      "loss": 1.2294,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12104946048091143,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9973056007554145e-05,
+      "loss": 1.1643,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12159719559620967,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.9971635074484174e-05,
+      "loss": 1.2341,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12214493071150792,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9970177683580245e-05,
+      "loss": 1.2422,
+      "step": 1115
+    },
+    {
+      "epoch": 0.12269266582680616,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.99686838401708e-05,
+      "loss": 1.1809,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1232404009421044,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.996715354971755e-05,
+      "loss": 1.2402,
+      "step": 1125
+    },
+    {
+      "epoch": 0.12378813605740265,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9965586817815494e-05,
+      "loss": 1.1718,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12433587117270088,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.996398365019283e-05,
+      "loss": 1.1678,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12488360628799912,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.996234405271099e-05,
+      "loss": 1.2031,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12543134140329737,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9960668031364593e-05,
+      "loss": 1.2042,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1259790765185956,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9958955592281436e-05,
+      "loss": 1.2292,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12652681163389384,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9957206741722455e-05,
+      "loss": 1.2361,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12707454674919208,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9955421486081718e-05,
+      "loss": 1.1373,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12762228186449034,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9953599831886398e-05,
+      "loss": 1.1941,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12817001697978858,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9951741785796737e-05,
+      "loss": 1.1622,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12871775209508682,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9949847354606046e-05,
+      "loss": 1.1535,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12926548721038505,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.994791654524065e-05,
+      "loss": 1.1981,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1298132223256833,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9945949364759887e-05,
+      "loss": 1.1849,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13036095744098153,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9943945820356075e-05,
+      "loss": 1.1899,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1309086925562798,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.994190591935448e-05,
+      "loss": 1.2066,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13145642767157803,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9939829669213296e-05,
+      "loss": 1.1682,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13200416278687627,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9937717077523607e-05,
+      "loss": 1.1963,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1325518979021745,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9935568152009378e-05,
+      "loss": 1.177,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13309963301747274,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9933382900527413e-05,
+      "loss": 1.1594,
+      "step": 1215
+    },
+    {
+      "epoch": 0.133647368132771,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.9931161331067327e-05,
+      "loss": 1.183,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13419510324806924,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9928903451751517e-05,
+      "loss": 1.21,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13474283836336748,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.992660927083514e-05,
+      "loss": 1.2326,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13529057347866572,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.992427879670608e-05,
+      "loss": 1.1703,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13583830859396395,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.99219120378849e-05,
+      "loss": 1.2061,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1363860437092622,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.991950900302484e-05,
+      "loss": 1.2041,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13693377882456045,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9917069700911766e-05,
+      "loss": 1.1703,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1374815139398587,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.991459414046414e-05,
+      "loss": 1.1598,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13802924905515693,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9912082330732992e-05,
+      "loss": 1.1528,
+      "step": 1260
+    },
+    {
+      "epoch": 0.13857698417045516,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.990953428090189e-05,
+      "loss": 1.1381,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1391247192857534,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9906950000286894e-05,
+      "loss": 1.1351,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13967245440105164,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.990432949833653e-05,
+      "loss": 1.1824,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1402201895163499,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.990167278463176e-05,
+      "loss": 1.1525,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14076792463164814,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9898979868885933e-05,
+      "loss": 1.1933,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14131565974694638,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.989625076094477e-05,
+      "loss": 1.1801,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1418633948622446,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9893485470786307e-05,
+      "loss": 1.2175,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14241112997754285,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9890684008520872e-05,
+      "loss": 1.1855,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14295886509284111,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9887846384391048e-05,
+      "loss": 1.1998,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14350660020813935,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9884972608771612e-05,
+      "loss": 1.1501,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1440543353234376,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9882062692169544e-05,
+      "loss": 1.2048,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14460207043873582,
+      "grad_norm": 1.25,
+      "learning_rate": 1.987911664522394e-05,
+      "loss": 1.1473,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14514980555403406,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9876134478706004e-05,
+      "loss": 1.1571,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1456975406693323,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9873116203518997e-05,
+      "loss": 1.2003,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14624527578463056,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9870061830698196e-05,
+      "loss": 1.1797,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1467930108999288,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9866971371410858e-05,
+      "loss": 1.1441,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14734074601522704,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9863844836956177e-05,
+      "loss": 1.1912,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14788848113052527,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.986068223876525e-05,
+      "loss": 1.171,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1484362162458235,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.9857483588401023e-05,
+      "loss": 1.1628,
+      "step": 1355
+    },
+    {
+      "epoch": 0.14898395136112177,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9854248897558247e-05,
+      "loss": 1.1522,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14953168647642,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.985097817806346e-05,
+      "loss": 1.1907,
+      "step": 1365
+    },
+    {
+      "epoch": 0.15007942159171825,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9847671441874907e-05,
+      "loss": 1.1937,
+      "step": 1370
+    },
+    {
+      "epoch": 0.15062715670701649,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9844328701082532e-05,
+      "loss": 1.1997,
+      "step": 1375
+    },
+    {
+      "epoch": 0.15117489182231472,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9840949967907906e-05,
+      "loss": 1.1507,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15172262693761296,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.98375352547042e-05,
+      "loss": 1.1952,
+      "step": 1385
+    },
+    {
+      "epoch": 0.15227036205291122,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.983408457395613e-05,
+      "loss": 1.1519,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15281809716820946,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9830597938279915e-05,
+      "loss": 1.2213,
+      "step": 1395
+    },
+    {
+      "epoch": 0.1533658322835077,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9827075360423236e-05,
+      "loss": 1.168,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15391356739880593,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.982351685326518e-05,
+      "loss": 1.1586,
+      "step": 1405
+    },
+    {
+      "epoch": 0.15446130251410417,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9819922429816193e-05,
+      "loss": 1.1292,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1550090376294024,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.981629210321805e-05,
+      "loss": 1.2092,
+      "step": 1415
+    },
+    {
+      "epoch": 0.15555677274470067,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9812625886743775e-05,
+      "loss": 1.1997,
+      "step": 1420
+    },
+    {
+      "epoch": 0.1561045078599989,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.980892379379762e-05,
+      "loss": 1.1885,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15665224297529715,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9805185837915014e-05,
+      "loss": 1.1733,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15719997809059538,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.9801412032762495e-05,
+      "loss": 1.1742,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15774771320589362,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9797602392137678e-05,
+      "loss": 1.1946,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15829544832119188,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9793756929969195e-05,
+      "loss": 1.1446,
+      "step": 1445
+    },
+    {
+      "epoch": 0.15884318343649012,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.978987566031665e-05,
+      "loss": 1.2024,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15939091855178836,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9785958597370557e-05,
+      "loss": 1.194,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1599386536670866,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9782005755452306e-05,
+      "loss": 1.1704,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16048638878238483,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9778017149014098e-05,
+      "loss": 1.2247,
+      "step": 1465
+    },
+    {
+      "epoch": 0.16103412389768307,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.977399279263889e-05,
+      "loss": 1.1891,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16158185901298133,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9769932701040343e-05,
+      "loss": 1.1585,
+      "step": 1475
+    },
+    {
+      "epoch": 0.16212959412827957,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9765836889062788e-05,
+      "loss": 1.1668,
+      "step": 1480
+    },
+    {
+      "epoch": 0.1626773292435778,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9761705371681138e-05,
+      "loss": 1.2052,
+      "step": 1485
+    },
+    {
+      "epoch": 0.16322506435887604,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.975753816400086e-05,
+      "loss": 1.1191,
+      "step": 1490
+    },
+    {
+      "epoch": 0.16377279947417428,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.975333528125791e-05,
+      "loss": 1.1966,
+      "step": 1495
+    },
+    {
+      "epoch": 0.16432053458947254,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9749096738818663e-05,
+      "loss": 1.1594,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16486826970477078,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9744822552179895e-05,
+      "loss": 1.2079,
+      "step": 1505
+    },
+    {
+      "epoch": 0.16541600482006902,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9740512736968688e-05,
+      "loss": 1.1655,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16596373993536725,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9736167308942385e-05,
+      "loss": 1.1866,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1665114750506655,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.973178628398855e-05,
+      "loss": 1.1686,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16705921016596373,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9727369678124876e-05,
+      "loss": 1.1725,
+      "step": 1525
+    },
+    {
+      "epoch": 0.167606945281262,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.9722917507499154e-05,
+      "loss": 1.2279,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16815468039656023,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9718429788389214e-05,
+      "loss": 1.2022,
+      "step": 1535
+    },
+    {
+      "epoch": 0.16870241551185847,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.971390653720284e-05,
+      "loss": 1.2493,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1692501506271567,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9709347770477743e-05,
+      "loss": 1.1685,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16979788574245494,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9704753504881476e-05,
+      "loss": 1.2464,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17034562085775318,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9700123757211372e-05,
+      "loss": 1.1843,
+      "step": 1555
+    },
+    {
+      "epoch": 0.17089335597305144,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.969545854439451e-05,
+      "loss": 1.1613,
+      "step": 1560
+    },
+    {
+      "epoch": 0.17144109108834968,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9690757883487626e-05,
+      "loss": 1.174,
+      "step": 1565
+    },
+    {
+      "epoch": 0.17198882620364792,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9686021791677055e-05,
+      "loss": 1.2428,
+      "step": 1570
+    },
+    {
+      "epoch": 0.17253656131894615,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9681250286278685e-05,
+      "loss": 1.1463,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1730842964342444,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9676443384737873e-05,
+      "loss": 1.2007,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17363203154954265,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9671601104629388e-05,
+      "loss": 1.1912,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1741797666648409,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9666723463657357e-05,
+      "loss": 1.2221,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17472750178013913,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9661810479655184e-05,
+      "loss": 1.2012,
+      "step": 1595
+    },
+    {
+      "epoch": 0.17527523689543736,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9656862170585494e-05,
+      "loss": 1.1746,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1758229720107356,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.965187855454007e-05,
+      "loss": 1.1739,
+      "step": 1605
+    },
+    {
+      "epoch": 0.17637070712603384,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.964685964973978e-05,
+      "loss": 1.175,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1769184422413321,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9641805474534514e-05,
+      "loss": 1.1344,
+      "step": 1615
+    },
+    {
+      "epoch": 0.17746617735663034,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.963671604740311e-05,
+      "loss": 1.1242,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17801391247192858,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9631591386953303e-05,
+      "loss": 1.1899,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1785616475872268,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9626431511921638e-05,
+      "loss": 1.2024,
+      "step": 1630
+    },
+    {
+      "epoch": 0.17910938270252505,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.9621236441173414e-05,
+      "loss": 1.1714,
+      "step": 1635
+    },
+    {
+      "epoch": 0.17965711781782331,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9616006193702608e-05,
+      "loss": 1.1386,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18020485293312155,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9610740788631816e-05,
+      "loss": 1.1612,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1807525880484198,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9605440245212165e-05,
+      "loss": 1.1568,
+      "step": 1650
+    },
+    {
+      "epoch": 0.18130032316371802,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.960010458282326e-05,
+      "loss": 1.145,
+      "step": 1655
+    },
+    {
+      "epoch": 0.18184805827901626,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9594733820973105e-05,
+      "loss": 1.1805,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1823957933943145,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.958932797929803e-05,
+      "loss": 1.1542,
+      "step": 1665
+    },
+    {
+      "epoch": 0.18294352850961276,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.958388707756263e-05,
+      "loss": 1.1288,
+      "step": 1670
+    },
+    {
+      "epoch": 0.183491263624911,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.957841113565967e-05,
+      "loss": 1.2251,
+      "step": 1675
+    },
+    {
+      "epoch": 0.18403899874020924,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.957290017361005e-05,
+      "loss": 1.166,
+      "step": 1680
+    },
+    {
+      "epoch": 0.18458673385550747,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9567354211562693e-05,
+      "loss": 1.179,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1851344689708057,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9561773269794484e-05,
+      "loss": 1.1975,
+      "step": 1690
+    },
+    {
+      "epoch": 0.18568220408610395,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.955615736871021e-05,
+      "loss": 1.188,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1862299392014022,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.955050652884247e-05,
+      "loss": 1.2116,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18677767431670045,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9544820770851608e-05,
+      "loss": 1.1793,
+      "step": 1705
+    },
+    {
+      "epoch": 0.18732540943199869,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9539100115525625e-05,
+      "loss": 1.1777,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18787314454729692,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9533344583780124e-05,
+      "loss": 1.1605,
+      "step": 1715
+    },
+    {
+      "epoch": 0.18842087966259516,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.952755419665821e-05,
+      "loss": 1.2486,
+      "step": 1720
+    },
+    {
+      "epoch": 0.18896861477789342,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9521728975330436e-05,
+      "loss": 1.216,
+      "step": 1725
+    },
+    {
+      "epoch": 0.18951634989319166,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.951586894109471e-05,
+      "loss": 1.1769,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1900640850084899,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9509974115376208e-05,
+      "loss": 1.1962,
+      "step": 1735
+    },
+    {
+      "epoch": 0.19061182012378813,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9504044519727333e-05,
+      "loss": 1.1603,
+      "step": 1740
+    },
+    {
+      "epoch": 0.19115955523908637,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9498080175827598e-05,
+      "loss": 1.1867,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1917072903543846,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.949208110548356e-05,
+      "loss": 1.1531,
+      "step": 1750
+    },
+    {
+      "epoch": 0.19225502546968287,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9486047330628745e-05,
+      "loss": 1.1891,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1928027605849811,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9479978873323565e-05,
+      "loss": 1.1366,
+      "step": 1760
+    },
+    {
+      "epoch": 0.19335049570027935,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9473875755755235e-05,
+      "loss": 1.1604,
+      "step": 1765
+    },
+    {
+      "epoch": 0.19389823081557758,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9467738000237685e-05,
+      "loss": 1.1663,
+      "step": 1770
+    },
+    {
+      "epoch": 0.19444596593087582,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.946156562921151e-05,
+      "loss": 1.1747,
+      "step": 1775
+    },
+    {
+      "epoch": 0.19499370104617408,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9455358665243836e-05,
+      "loss": 1.1973,
+      "step": 1780
+    },
+    {
+      "epoch": 0.19554143616147232,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9449117131028282e-05,
+      "loss": 1.1809,
+      "step": 1785
+    },
+    {
+      "epoch": 0.19608917127677056,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9442841049384865e-05,
+      "loss": 1.2348,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1966369063920688,
+      "grad_norm": 1.25,
+      "learning_rate": 1.94365304432599e-05,
+      "loss": 1.1703,
+      "step": 1795
+    },
+    {
+      "epoch": 0.19718464150736703,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9430185335725942e-05,
+      "loss": 1.1491,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19773237662266527,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9423805749981673e-05,
+      "loss": 1.143,
+      "step": 1805
+    },
+    {
+      "epoch": 0.19828011173796353,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9417391709351845e-05,
+      "loss": 1.1962,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19882784685326177,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9410943237287178e-05,
+      "loss": 1.2093,
+      "step": 1815
+    },
+    {
+      "epoch": 0.19937558196856,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9404460357364282e-05,
+      "loss": 1.165,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19992331708385824,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.939794309328556e-05,
+      "loss": 1.1203,
+      "step": 1825
+    },
+    {
+      "epoch": 0.20047105219915648,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.939139146887914e-05,
+      "loss": 1.1476,
+      "step": 1830
+    },
+    {
+      "epoch": 0.20101878731445472,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9384805508098763e-05,
+      "loss": 1.1903,
+      "step": 1835
+    },
+    {
+      "epoch": 0.20156652242975298,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.937818523502372e-05,
+      "loss": 1.21,
+      "step": 1840
+    },
+    {
+      "epoch": 0.20211425754505122,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9371530673858753e-05,
+      "loss": 1.1682,
+      "step": 1845
+    },
+    {
+      "epoch": 0.20266199266034945,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.936484184893395e-05,
+      "loss": 1.1534,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2032097277756477,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.93581187847047e-05,
+      "loss": 1.136,
+      "step": 1855
+    },
+    {
+      "epoch": 0.20375746289094593,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9351361505751554e-05,
+      "loss": 1.1976,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2043051980062442,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9344570036780165e-05,
+      "loss": 1.1293,
+      "step": 1865
+    },
+    {
+      "epoch": 0.20485293312154243,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9337744402621194e-05,
+      "loss": 1.1543,
+      "step": 1870
+    },
+    {
+      "epoch": 0.20540066823684067,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9330884628230203e-05,
+      "loss": 1.2295,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2059484033521389,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.932399073868759e-05,
+      "loss": 1.1531,
+      "step": 1880
+    },
+    {
+      "epoch": 0.20649613846743714,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9317062759198478e-05,
+      "loss": 1.1634,
+      "step": 1885
+    },
+    {
+      "epoch": 0.20704387358273538,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9310100715092624e-05,
+      "loss": 1.1359,
+      "step": 1890
+    },
+    {
+      "epoch": 0.20759160869803364,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9303104631824338e-05,
+      "loss": 1.1439,
+      "step": 1895
+    },
+    {
+      "epoch": 0.20813934381333188,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.929607453497237e-05,
+      "loss": 1.16,
+      "step": 1900
+    },
+    {
+      "epoch": 0.20868707892863012,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9289010450239843e-05,
+      "loss": 1.1588,
+      "step": 1905
+    },
+    {
+      "epoch": 0.20923481404392835,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.928191240345414e-05,
+      "loss": 1.1687,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2097825491592266,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9274780420566813e-05,
+      "loss": 1.1872,
+      "step": 1915
+    },
+    {
+      "epoch": 0.21033028427452485,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.926761452765349e-05,
+      "loss": 1.1762,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2108780193898231,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.926041475091377e-05,
+      "loss": 1.1131,
+      "step": 1925
+    },
+    {
+      "epoch": 0.21142575450512133,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.925318111667116e-05,
+      "loss": 1.1835,
+      "step": 1930
+    },
+    {
+      "epoch": 0.21197348962041956,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9245913651372935e-05,
+      "loss": 1.2117,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2125212247357178,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.923861238159007e-05,
+      "loss": 1.1396,
+      "step": 1940
+    },
+    {
+      "epoch": 0.21306895985101604,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9231277334017126e-05,
+      "loss": 1.1932,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2136166949663143,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9223908535472172e-05,
+      "loss": 1.1839,
+      "step": 1950
+    },
+    {
+      "epoch": 0.21416443008161254,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.921650601289667e-05,
+      "loss": 1.1628,
+      "step": 1955
+    },
+    {
+      "epoch": 0.21471216519691078,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9209069793355382e-05,
+      "loss": 1.1491,
+      "step": 1960
+    },
+    {
+      "epoch": 0.215259900312209,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9201599904036276e-05,
+      "loss": 1.186,
+      "step": 1965
+    },
+    {
+      "epoch": 0.21580763542750725,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.919409637225041e-05,
+      "loss": 1.1556,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2163553705428055,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9186559225431857e-05,
+      "loss": 1.1598,
+      "step": 1975
+    },
+    {
+      "epoch": 0.21690310565810375,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.917898849113759e-05,
+      "loss": 1.2076,
+      "step": 1980
+    },
+    {
+      "epoch": 0.217450840773402,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9171384197047376e-05,
+      "loss": 1.1219,
+      "step": 1985
+    },
+    {
+      "epoch": 0.21799857588870022,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9163746370963687e-05,
+      "loss": 1.1347,
+      "step": 1990
+    },
+    {
+      "epoch": 0.21854631100399846,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9156075040811585e-05,
+      "loss": 1.1717,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2190940461192967,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9148370234638646e-05,
+      "loss": 1.1705,
+      "step": 2000
+    },
+    {
+      "epoch": 0.21964178123459496,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.914063198061482e-05,
+      "loss": 1.1666,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2201895163498932,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9132860307032362e-05,
+      "loss": 1.1711,
+      "step": 2010
+    },
+    {
+      "epoch": 0.22073725146519144,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9125055242305702e-05,
+      "loss": 1.1176,
+      "step": 2015
+    },
+    {
+      "epoch": 0.22128498658048967,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9117216814971365e-05,
+      "loss": 1.1716,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2218327216957879,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9109345053687845e-05,
+      "loss": 1.1684,
+      "step": 2025
+    },
+    {
+      "epoch": 0.22238045681108615,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.9101439987235514e-05,
+      "loss": 1.1766,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2229281919263844,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.909350164451651e-05,
+      "loss": 1.21,
+      "step": 2035
+    },
+    {
+      "epoch": 0.22347592704168265,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9085530054554642e-05,
+      "loss": 1.2172,
+      "step": 2040
+    },
+    {
+      "epoch": 0.22402366215698089,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.907752524649527e-05,
+      "loss": 1.1885,
+      "step": 2045
+    },
+    {
+      "epoch": 0.22457139727227912,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9069487249605202e-05,
+      "loss": 1.1708,
+      "step": 2050
+    },
+    {
+      "epoch": 0.22511913238757736,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.90614160932726e-05,
+      "loss": 1.1645,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2256668675028756,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9053311807006845e-05,
+      "loss": 1.1632,
+      "step": 2060
+    },
+    {
+      "epoch": 0.22621460261817386,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9045174420438465e-05,
+      "loss": 1.1619,
+      "step": 2065
+    },
+    {
+      "epoch": 0.2267623377334721,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9037003963318993e-05,
+      "loss": 1.1658,
+      "step": 2070
+    },
+    {
+      "epoch": 0.22731007284877033,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9028800465520878e-05,
+      "loss": 1.1941,
+      "step": 2075
+    },
+    {
+      "epoch": 0.22785780796406857,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9020563957037378e-05,
+      "loss": 1.1463,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2284055430793668,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.901229446798243e-05,
+      "loss": 1.16,
+      "step": 2085
+    },
+    {
+      "epoch": 0.22895327819466507,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9003992028590568e-05,
+      "loss": 1.1567,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2295010133099633,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.8995656669216786e-05,
+      "loss": 1.2039,
+      "step": 2095
+    },
+    {
+      "epoch": 0.23004874842526155,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.898728842033644e-05,
+      "loss": 1.1733,
+      "step": 2100
+    },
+    {
+      "epoch": 0.23059648354055978,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8978887312545135e-05,
+      "loss": 1.1702,
+      "step": 2105
+    },
+    {
+      "epoch": 0.23114421865585802,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.897045337655862e-05,
+      "loss": 1.2218,
+      "step": 2110
+    },
+    {
+      "epoch": 0.23169195377115626,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8961986643212664e-05,
+      "loss": 1.2084,
+      "step": 2115
+    },
+    {
+      "epoch": 0.23223968888645452,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.895348714346294e-05,
+      "loss": 1.1865,
+      "step": 2120
+    },
+    {
+      "epoch": 0.23278742400175276,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.894495490838494e-05,
+      "loss": 1.2198,
+      "step": 2125
+    },
+    {
+      "epoch": 0.233335159117051,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8936389969173825e-05,
+      "loss": 1.134,
+      "step": 2130
+    },
+    {
+      "epoch": 0.23388289423234923,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.892779235714433e-05,
+      "loss": 1.1399,
+      "step": 2135
+    },
+    {
+      "epoch": 0.23443062934764747,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8919162103730656e-05,
+      "loss": 1.1123,
+      "step": 2140
+    },
+    {
+      "epoch": 0.23497836446294573,
+      "grad_norm": 1.25,
+      "learning_rate": 1.891049924048633e-05,
+      "loss": 1.1695,
+      "step": 2145
+    },
+    {
+      "epoch": 0.23552609957824397,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.8901803799084117e-05,
+      "loss": 1.1359,
+      "step": 2150
+    },
+    {
+      "epoch": 0.2360738346935422,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.8893075811315894e-05,
+      "loss": 1.2111,
+      "step": 2155
+    },
+    {
+      "epoch": 0.23662156980884044,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.888431530909253e-05,
+      "loss": 1.1088,
+      "step": 2160
+    },
+    {
+      "epoch": 0.23716930492413868,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8875522324443762e-05,
+      "loss": 1.1838,
+      "step": 2165
+    },
+    {
+      "epoch": 0.23771704003943692,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8866696889518107e-05,
+      "loss": 1.1323,
+      "step": 2170
+    },
+    {
+      "epoch": 0.23826477515473518,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.8857839036582707e-05,
+      "loss": 1.1296,
+      "step": 2175
+    },
+    {
+      "epoch": 0.23881251027003342,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8848948798023238e-05,
+      "loss": 1.1617,
+      "step": 2180
+    },
+    {
+      "epoch": 0.23936024538533165,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8840026206343786e-05,
+      "loss": 1.1627,
+      "step": 2185
+    },
+    {
+      "epoch": 0.2399079805006299,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8831071294166717e-05,
+      "loss": 1.1173,
+      "step": 2190
+    },
+    {
+      "epoch": 0.24045571561592813,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.882208409423257e-05,
+      "loss": 1.212,
+      "step": 2195
+    },
+    {
+      "epoch": 0.24100345073122637,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8813064639399933e-05,
+      "loss": 1.1543,
+      "step": 2200
+    },
+    {
+      "epoch": 0.24155118584652463,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.8804012962645322e-05,
+      "loss": 1.1847,
+      "step": 2205
+    },
+    {
+      "epoch": 0.24209892096182287,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8794929097063062e-05,
+      "loss": 1.2248,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2426466560771211,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8785813075865164e-05,
+      "loss": 1.2115,
+      "step": 2215
+    },
+    {
+      "epoch": 0.24319439119241934,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8776664932381208e-05,
+      "loss": 1.1039,
+      "step": 2220
+    },
+    {
+      "epoch": 0.24374212630771758,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8767484700058212e-05,
+      "loss": 1.1608,
+      "step": 2225
+    },
+    {
+      "epoch": 0.24428986142301584,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.875827241246052e-05,
+      "loss": 1.1998,
+      "step": 2230
+    },
+    {
+      "epoch": 0.24483759653831408,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.874902810326968e-05,
+      "loss": 1.2347,
+      "step": 2235
+    },
+    {
+      "epoch": 0.24538533165361232,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.873975180628431e-05,
+      "loss": 1.149,
+      "step": 2240
+    },
+    {
+      "epoch": 0.24593306676891055,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.873044355541997e-05,
+      "loss": 1.1524,
+      "step": 2245
+    },
+    {
+      "epoch": 0.2464808018842088,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.872110338470907e-05,
+      "loss": 1.1628,
+      "step": 2250
+    },
+    {
+      "epoch": 0.24702853699950703,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8711731328300715e-05,
+      "loss": 1.201,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2475762721148053,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8702327420460575e-05,
+      "loss": 1.1418,
+      "step": 2260
+    },
+    {
+      "epoch": 0.24812400723010353,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.8692891695570795e-05,
+      "loss": 1.1357,
+      "step": 2265
+    },
+    {
+      "epoch": 0.24867174234540176,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.8683424188129836e-05,
+      "loss": 1.1994,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2492194774607,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8673924932752366e-05,
+      "loss": 1.1592,
+      "step": 2275
+    },
+    {
+      "epoch": 0.24976721257599824,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.8664393964169125e-05,
+      "loss": 1.1777,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2503149476912965,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.86548313172268e-05,
+      "loss": 1.1569,
+      "step": 2285
+    },
+    {
+      "epoch": 0.25086268280659474,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8645237026887896e-05,
+      "loss": 1.162,
+      "step": 2290
+    },
+    {
+      "epoch": 0.25141041792189295,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8635611128230632e-05,
+      "loss": 1.1158,
+      "step": 2295
+    },
+    {
+      "epoch": 0.2519581530371912,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8625953656448764e-05,
+      "loss": 1.1439,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2525058881524895,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.86162646468515e-05,
+      "loss": 1.1504,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2530536232677877,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8606544134863355e-05,
+      "loss": 1.1322,
+      "step": 2310
+    },
+    {
+      "epoch": 0.25360135838308595,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8596792156024014e-05,
+      "loss": 1.2144,
+      "step": 2315
+    },
+    {
+      "epoch": 0.25414909349838416,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.858700874598821e-05,
+      "loss": 1.1363,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2546968286136824,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8577193940525608e-05,
+      "loss": 1.1982,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2552445637289807,
+      "grad_norm": 1.25,
+      "learning_rate": 1.8567347775520642e-05,
+      "loss": 1.1562,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2557922988442789,
+      "grad_norm": 1.25,
+      "learning_rate": 1.8557470286972405e-05,
+      "loss": 1.1552,
+      "step": 2335
+    },
+    {
+      "epoch": 0.25634003395957716,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.8547561510994525e-05,
+      "loss": 1.1051,
+      "step": 2340
+    },
+    {
+      "epoch": 0.25688776907487537,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.853762148381501e-05,
+      "loss": 1.1806,
+      "step": 2345
+    },
+    {
+      "epoch": 0.25743550419017364,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8527650241776128e-05,
+      "loss": 1.1662,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2579832393054719,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8517647821334278e-05,
+      "loss": 1.2243,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2585309744207701,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.8507614259059847e-05,
+      "loss": 1.1888,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2590787095360684,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.849754959163709e-05,
+      "loss": 1.1542,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2596264446513666,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.848745385586398e-05,
+      "loss": 1.1973,
+      "step": 2370
+    },
+    {
+      "epoch": 0.26017417976666485,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8477327088652078e-05,
+      "loss": 1.1869,
+      "step": 2375
+    },
+    {
+      "epoch": 0.26072191488196306,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.846716932702641e-05,
+      "loss": 1.1914,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2612696499972613,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8456980608125317e-05,
+      "loss": 1.1582,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2618173851125596,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8446760969200318e-05,
+      "loss": 1.1917,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2623651202278578,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8436510447615997e-05,
+      "loss": 1.1456,
+      "step": 2395
+    },
+    {
+      "epoch": 0.26291285534315606,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8426229080849833e-05,
+      "loss": 1.1333,
+      "step": 2400
+    },
+    {
+      "epoch": 0.26346059045845427,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8415916906492093e-05,
+      "loss": 1.1286,
+      "step": 2405
+    },
+    {
+      "epoch": 0.26400832557375253,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8405573962245666e-05,
+      "loss": 1.1542,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2645560606890508,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.839520028592596e-05,
+      "loss": 1.1858,
+      "step": 2415
+    },
+    {
+      "epoch": 0.265103795804349,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.838479591546072e-05,
+      "loss": 1.1588,
+      "step": 2420
+    },
+    {
+      "epoch": 0.26565153091964727,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8374360888889943e-05,
+      "loss": 1.174,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2661992660349455,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8363895244365687e-05,
+      "loss": 1.2234,
+      "step": 2430
+    },
+    {
+      "epoch": 0.26674700115024375,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8353399020151954e-05,
+      "loss": 1.1582,
+      "step": 2435
+    },
+    {
+      "epoch": 0.267294736265542,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8342872254624565e-05,
+      "loss": 1.1537,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2678424713808402,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8332314986270994e-05,
+      "loss": 1.154,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2683902064961385,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.832172725369024e-05,
+      "loss": 1.1304,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2689379416114367,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.831110909559269e-05,
+      "loss": 1.1849,
+      "step": 2455
+    },
+    {
+      "epoch": 0.26948567672673496,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.830046055079995e-05,
+      "loss": 1.1697,
+      "step": 2460
+    },
+    {
+      "epoch": 0.27003341184203317,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.8289781658244757e-05,
+      "loss": 1.1658,
+      "step": 2465
+    },
+    {
+      "epoch": 0.27058114695733143,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.827907245697078e-05,
+      "loss": 1.1697,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2711288820726297,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.826833298613251e-05,
+      "loss": 1.1042,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2716766171879279,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.825756328499511e-05,
+      "loss": 1.2129,
+      "step": 2480
+    },
+    {
+      "epoch": 0.27222435230322617,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8246763392934256e-05,
+      "loss": 1.2177,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2727720874185244,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8235933349436025e-05,
+      "loss": 1.134,
+      "step": 2490
+    },
+    {
+      "epoch": 0.27331982253382264,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8225073194096728e-05,
+      "loss": 1.1721,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2738675576491209,
+      "grad_norm": 1.25,
+      "learning_rate": 1.8214182966622758e-05,
+      "loss": 1.2035,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2744152927644191,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8203262706830467e-05,
+      "loss": 1.1489,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2749630278797174,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8192312454646007e-05,
+      "loss": 1.1163,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2755107629950156,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.818133225010519e-05,
+      "loss": 1.1847,
+      "step": 2515
+    },
+    {
+      "epoch": 0.27605849811031385,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8170322133353328e-05,
+      "loss": 1.2177,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2766062332256121,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.815928214464511e-05,
+      "loss": 1.1929,
+      "step": 2525
+    },
+    {
+      "epoch": 0.27715396834091033,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.814821232434444e-05,
+      "loss": 1.1931,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2777017034562086,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8137112712924273e-05,
+      "loss": 1.1286,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2782494385715068,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.812598335096651e-05,
+      "loss": 1.1489,
+      "step": 2540
+    },
+    {
+      "epoch": 0.27879717368680507,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8114824279161806e-05,
+      "loss": 1.1457,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2793449088021033,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.810363553830945e-05,
+      "loss": 1.1604,
+      "step": 2550
+    },
+    {
+      "epoch": 0.27989264391740154,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.8092417169317203e-05,
+      "loss": 1.2167,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2804403790326998,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8081169213201145e-05,
+      "loss": 1.1854,
+      "step": 2560
+    },
+    {
+      "epoch": 0.280988114147998,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.806989171108554e-05,
+      "loss": 1.1755,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2815358492632963,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8058584704202674e-05,
+      "loss": 1.1673,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2820835843785945,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8047248233892706e-05,
+      "loss": 1.2167,
+      "step": 2575
+    },
+    {
+      "epoch": 0.28263131949389275,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8035882341603518e-05,
+      "loss": 1.1441,
+      "step": 2580
+    },
+    {
+      "epoch": 0.283179054609191,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8024487068890556e-05,
+      "loss": 1.1872,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2837267897244892,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.8013062457416702e-05,
+      "loss": 1.1503,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2842745248397875,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.800160854895209e-05,
+      "loss": 1.1635,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2848222599550857,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7990125385373978e-05,
+      "loss": 1.1482,
+      "step": 2600
+    },
+    {
+      "epoch": 0.28536999507038396,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.7978613008666577e-05,
+      "loss": 1.1801,
+      "step": 2605
+    },
+    {
+      "epoch": 0.28591773018568223,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.796707146092091e-05,
+      "loss": 1.1781,
+      "step": 2610
+    },
+    {
+      "epoch": 0.28646546530098044,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7955500784334655e-05,
+      "loss": 1.1593,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2870132004162787,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.794390102121199e-05,
+      "loss": 1.1746,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2875609355315769,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.7932272213963434e-05,
+      "loss": 1.1758,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2881086706468752,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.7920614405105695e-05,
+      "loss": 1.2296,
+      "step": 2630
+    },
+    {
+      "epoch": 0.28865640576217344,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7908927637261522e-05,
+      "loss": 1.1928,
+      "step": 2635
+    },
+    {
+      "epoch": 0.28920414087747165,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.789721195315954e-05,
+      "loss": 1.1237,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2897518759927699,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7885467395634087e-05,
+      "loss": 1.1734,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2902996111080681,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.7873694007625084e-05,
+      "loss": 1.1656,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2908473462233664,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.786189183217784e-05,
+      "loss": 1.1558,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2913950813386646,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.785006091244294e-05,
+      "loss": 1.1529,
+      "step": 2660
+    },
+    {
+      "epoch": 0.29194281645396286,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7838201291676037e-05,
+      "loss": 1.1561,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2924905515692611,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.7826313013237744e-05,
+      "loss": 1.1463,
+      "step": 2670
+    },
+    {
+      "epoch": 0.29303828668455933,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7814396120593428e-05,
+      "loss": 1.1611,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2935860217998576,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7802450657313086e-05,
+      "loss": 1.1687,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2941337569151558,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7790476667071175e-05,
+      "loss": 1.1818,
+      "step": 2685
+    },
+    {
+      "epoch": 0.2946814920304541,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7778474193646448e-05,
+      "loss": 1.1589,
+      "step": 2690
+    },
+    {
+      "epoch": 0.29522922714575234,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.776644328092179e-05,
+      "loss": 1.1448,
+      "step": 2695
+    },
+    {
+      "epoch": 0.29577696226105055,
+      "grad_norm": 1.25,
+      "learning_rate": 1.7754383972884084e-05,
+      "loss": 1.1354,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2963246973763488,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.7742296313624005e-05,
+      "loss": 1.1309,
+      "step": 2705
+    },
+    {
+      "epoch": 0.296872432491647,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.773018034733591e-05,
+      "loss": 1.1921,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2974201676069453,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.771803611831762e-05,
+      "loss": 1.1484,
+      "step": 2715
+    },
+    {
+      "epoch": 0.29796790272224355,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7705863670970327e-05,
+      "loss": 1.2024,
+      "step": 2720
+    },
+    {
+      "epoch": 0.29851563783754176,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.7693663049798363e-05,
+      "loss": 1.1696,
+      "step": 2725
+    },
+    {
+      "epoch": 0.29906337295284,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.7681434299409077e-05,
+      "loss": 1.1191,
+      "step": 2730
+    },
+    {
+      "epoch": 0.29961110806813823,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.766917746451267e-05,
+      "loss": 1.1662,
+      "step": 2735
+    },
+    {
+      "epoch": 0.3001588431834365,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.7656892589922017e-05,
+      "loss": 1.1653,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3007065782987347,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.7644579720552513e-05,
+      "loss": 1.1145,
+      "step": 2745
+    },
+    {
+      "epoch": 0.30125431341403297,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.76322389014219e-05,
+      "loss": 1.1201,
+      "step": 2750
+    },
+    {
+      "epoch": 0.30180204852933123,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.761987017765012e-05,
+      "loss": 1.1349,
+      "step": 2755
+    },
+    {
+      "epoch": 0.30234978364462944,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.7607473594459127e-05,
+      "loss": 1.1937,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3028975187599277,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.7595049197172744e-05,
+      "loss": 1.1489,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3034452538752259,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7582597031216476e-05,
+      "loss": 1.1559,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3039929889905242,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.7570117142117366e-05,
+      "loss": 1.1413,
+      "step": 2775
+    },
+    {
+      "epoch": 0.30454072410582245,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7557609575503808e-05,
+      "loss": 1.1424,
+      "step": 2780
+    },
+    {
+      "epoch": 0.30508845922112066,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.754507437710539e-05,
+      "loss": 1.1737,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3056361943364189,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7532511592752734e-05,
+      "loss": 1.1835,
+      "step": 2790
+    },
+    {
+      "epoch": 0.30618392945171713,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.751992126837731e-05,
+      "loss": 1.2163,
+      "step": 2795
+    },
+    {
+      "epoch": 0.3067316645670154,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7507303450011287e-05,
+      "loss": 1.1411,
+      "step": 2800
+    },
+    {
+      "epoch": 0.30727939968231366,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7494658183787344e-05,
+      "loss": 1.2053,
+      "step": 2805
+    },
+    {
+      "epoch": 0.30782713479761187,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.7481985515938538e-05,
+      "loss": 1.1699,
+      "step": 2810
+    },
+    {
+      "epoch": 0.30837486991291013,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.746928549279808e-05,
+      "loss": 1.1804,
+      "step": 2815
+    },
+    {
+      "epoch": 0.30892260502820834,
+      "grad_norm": 1.25,
+      "learning_rate": 1.745655816079922e-05,
+      "loss": 1.1437,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3094703401435066,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.744380356647504e-05,
+      "loss": 1.13,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3100180752588048,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.743102175645831e-05,
+      "loss": 1.1917,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3105658103741031,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.741821277748128e-05,
+      "loss": 1.1613,
+      "step": 2835
+    },
+    {
+      "epoch": 0.31111354548940134,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.7405376676375567e-05,
+      "loss": 1.1408,
+      "step": 2840
+    },
+    {
+      "epoch": 0.31166128060469955,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.7392513500071927e-05,
+      "loss": 1.1696,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3122090157199978,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.737962329560011e-05,
+      "loss": 1.1046,
+      "step": 2850
+    },
+    {
+      "epoch": 0.312756750835296,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.7366706110088697e-05,
+      "loss": 1.1929,
+      "step": 2855
+    },
+    {
+      "epoch": 0.3133044859505943,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.73537619907649e-05,
+      "loss": 1.1473,
+      "step": 2860
+    },
+    {
+      "epoch": 0.31385222106589256,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7340790984954425e-05,
+      "loss": 1.098,
+      "step": 2865
+    },
+    {
+      "epoch": 0.31439995618119077,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.7327793140081256e-05,
+      "loss": 1.1105,
+      "step": 2870
+    },
+    {
+      "epoch": 0.31494769129648903,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.731476850366752e-05,
+      "loss": 1.186,
+      "step": 2875
+    },
+    {
+      "epoch": 0.31549542641178724,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7301717123333297e-05,
+      "loss": 1.2037,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3160431615270855,
+      "grad_norm": 1.25,
+      "learning_rate": 1.7288639046796442e-05,
+      "loss": 1.2007,
+      "step": 2885
+    },
+    {
+      "epoch": 0.31659089664238377,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7275534321872415e-05,
+      "loss": 1.1487,
+      "step": 2890
+    },
+    {
+      "epoch": 0.317138631757682,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.726240299647411e-05,
+      "loss": 1.1364,
+      "step": 2895
+    },
+    {
+      "epoch": 0.31768636687298024,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.7249245118611684e-05,
+      "loss": 1.1387,
+      "step": 2900
+    },
+    {
+      "epoch": 0.31823410198827845,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.723606073639235e-05,
+      "loss": 1.1701,
+      "step": 2905
+    },
+    {
+      "epoch": 0.3187818371035767,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7222849898020256e-05,
+      "loss": 1.1947,
+      "step": 2910
+    },
+    {
+      "epoch": 0.319329572218875,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.720961265179625e-05,
+      "loss": 1.1591,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3198773073341732,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7196349046117753e-05,
+      "loss": 1.1816,
+      "step": 2920
+    },
+    {
+      "epoch": 0.32042504244947145,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.7183059129478543e-05,
+      "loss": 1.1463,
+      "step": 2925
+    },
+    {
+      "epoch": 0.32097277756476966,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7169742950468607e-05,
+      "loss": 1.0909,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3215205126800679,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7156400557773944e-05,
+      "loss": 1.1657,
+      "step": 2935
+    },
+    {
+      "epoch": 0.32206824779536614,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.71430320001764e-05,
+      "loss": 1.1314,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3226159829106644,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.712963732655348e-05,
+      "loss": 1.1076,
+      "step": 2945
+    },
+    {
+      "epoch": 0.32316371802596267,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.711621658587817e-05,
+      "loss": 1.1591,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3237114531412609,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.7102769827218773e-05,
+      "loss": 1.1353,
+      "step": 2955
+    },
+    {
+      "epoch": 0.32425918825655914,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.7089297099738703e-05,
+      "loss": 1.186,
+      "step": 2960
+    },
+    {
+      "epoch": 0.32480692337185735,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.7075798452696333e-05,
+      "loss": 1.1698,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3253546584871556,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.706227393544479e-05,
+      "loss": 1.1474,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3259023936024539,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.7048723597431804e-05,
+      "loss": 1.1462,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3264501287177521,
+      "grad_norm": 1.25,
+      "learning_rate": 1.703514748819948e-05,
+      "loss": 1.1458,
+      "step": 2980
+    },
+    {
+      "epoch": 0.32699786383305035,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.702154565738418e-05,
+      "loss": 1.1661,
+      "step": 2985
+    },
+    {
+      "epoch": 0.32754559894834856,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7007918154716286e-05,
+      "loss": 1.1273,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3280933340636468,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.6994265030020056e-05,
+      "loss": 1.1597,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3286410691789451,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.69805863332134e-05,
+      "loss": 1.1346,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3291888042942433,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.6966882114307756e-05,
+      "loss": 1.2173,
+      "step": 3005
+    },
+    {
+      "epoch": 0.32973653940954156,
+      "grad_norm": 1.25,
+      "learning_rate": 1.695315242340785e-05,
+      "loss": 1.1515,
+      "step": 3010
+    },
+    {
+      "epoch": 0.33028427452483977,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.6939397310711557e-05,
+      "loss": 1.1896,
+      "step": 3015
+    },
+    {
+      "epoch": 0.33083200964013804,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.6925616826509678e-05,
+      "loss": 1.16,
+      "step": 3020
+    },
+    {
+      "epoch": 0.33137974475543625,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6911811021185795e-05,
+      "loss": 1.1782,
+      "step": 3025
+    },
+    {
+      "epoch": 0.3319274798707345,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.689797994521605e-05,
+      "loss": 1.1129,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3324752149860328,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.688412364916899e-05,
+      "loss": 1.1878,
+      "step": 3035
+    },
+    {
+      "epoch": 0.333022950101331,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6870242183705374e-05,
+      "loss": 1.1491,
+      "step": 3040
+    },
+    {
+      "epoch": 0.33357068521662925,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6856335599577973e-05,
+      "loss": 1.2087,
+      "step": 3045
+    },
+    {
+      "epoch": 0.33411842033192746,
+      "grad_norm": 1.25,
+      "learning_rate": 1.68424039476314e-05,
+      "loss": 1.1259,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3346661554472257,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.6828447278801923e-05,
+      "loss": 1.1633,
+      "step": 3055
+    },
+    {
+      "epoch": 0.335213890562524,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.6814465644117274e-05,
+      "loss": 1.1689,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3357616256778222,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.6800459094696458e-05,
+      "loss": 1.2037,
+      "step": 3065
+    },
+    {
+      "epoch": 0.33630936079312046,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6786427681749585e-05,
+      "loss": 1.166,
+      "step": 3070
+    },
+    {
+      "epoch": 0.33685709590841867,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.677237145657766e-05,
+      "loss": 1.1427,
+      "step": 3075
+    },
+    {
+      "epoch": 0.33740483102371693,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6758290470572404e-05,
+      "loss": 1.1598,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3379525661390152,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.674418477521607e-05,
+      "loss": 1.1667,
+      "step": 3085
+    },
+    {
+      "epoch": 0.3385003012543134,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.673005442208126e-05,
+      "loss": 1.1389,
+      "step": 3090
+    },
+    {
+      "epoch": 0.33904803636961167,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6715899462830716e-05,
+      "loss": 1.1899,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3395957714849099,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.6701719949217154e-05,
+      "loss": 1.1621,
+      "step": 3100
+    },
+    {
+      "epoch": 0.34014350660020815,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6687515933083055e-05,
+      "loss": 1.1612,
+      "step": 3105
+    },
+    {
+      "epoch": 0.34069124171550635,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.66732874663605e-05,
+      "loss": 1.1507,
+      "step": 3110
+    },
+    {
+      "epoch": 0.3412389768308046,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.665903460107094e-05,
+      "loss": 1.1442,
+      "step": 3115
+    },
+    {
+      "epoch": 0.3417867119461029,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.664475738932506e-05,
+      "loss": 1.1422,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3423344470614011,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.6630455883322535e-05,
+      "loss": 1.1343,
+      "step": 3125
+    },
+    {
+      "epoch": 0.34288218217669936,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.6616130135351884e-05,
+      "loss": 1.1564,
+      "step": 3130
+    },
+    {
+      "epoch": 0.34342991729199757,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.660178019779024e-05,
+      "loss": 1.1465,
+      "step": 3135
+    },
+    {
+      "epoch": 0.34397765240729583,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6587406123103194e-05,
+      "loss": 1.1779,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3445253875225941,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.657300796384457e-05,
+      "loss": 1.1625,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3450731226378923,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6558585772656255e-05,
+      "loss": 1.1463,
+      "step": 3150
+    },
+    {
+      "epoch": 0.34562085775319057,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6544139602268008e-05,
+      "loss": 1.1371,
+      "step": 3155
+    },
+    {
+      "epoch": 0.3461685928684888,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.652966950549725e-05,
+      "loss": 1.179,
+      "step": 3160
+    },
+    {
+      "epoch": 0.34671632798378704,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.6515175535248883e-05,
+      "loss": 1.1811,
+      "step": 3165
+    },
+    {
+      "epoch": 0.3472640630990853,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.6500657744515098e-05,
+      "loss": 1.1595,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3478117982143835,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6486116186375175e-05,
+      "loss": 1.1437,
+      "step": 3175
+    },
+    {
+      "epoch": 0.3483595333296818,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6471550913995286e-05,
+      "loss": 1.1407,
+      "step": 3180
+    },
+    {
+      "epoch": 0.34890726844498,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6456961980628317e-05,
+      "loss": 1.1064,
+      "step": 3185
+    },
+    {
+      "epoch": 0.34945500356027825,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6442349439613648e-05,
+      "loss": 1.1789,
+      "step": 3190
+    },
+    {
+      "epoch": 0.35000273867557646,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6427713344376987e-05,
+      "loss": 1.138,
+      "step": 3195
+    },
+    {
+      "epoch": 0.35055047379087473,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.6413053748430145e-05,
+      "loss": 1.1574,
+      "step": 3200
+    },
+    {
+      "epoch": 0.351098208906173,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.639837070537087e-05,
+      "loss": 1.1448,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3516459440214712,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6383664268882632e-05,
+      "loss": 1.151,
+      "step": 3210
+    },
+    {
+      "epoch": 0.35219367913676947,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.636893449273442e-05,
+      "loss": 1.1196,
+      "step": 3215
+    },
+    {
+      "epoch": 0.3527414142520677,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.635418143078057e-05,
+      "loss": 1.1732,
+      "step": 3220
+    },
+    {
+      "epoch": 0.35328914936736594,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6339405136960544e-05,
+      "loss": 1.1382,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3538368844826642,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.6324605665298755e-05,
+      "loss": 1.1572,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3543846195979624,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.630978306990435e-05,
+      "loss": 1.1503,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3549323547132607,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6294937404971016e-05,
+      "loss": 1.1678,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3554800898285589,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6280068724776795e-05,
+      "loss": 1.1897,
+      "step": 3245
+    },
+    {
+      "epoch": 0.35602782494385715,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.6265177083683875e-05,
+      "loss": 1.1742,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3565755600591554,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6250262536138383e-05,
+      "loss": 1.1802,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3571232951744536,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.6235325136670208e-05,
+      "loss": 1.1628,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3576710302897519,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6220364939892783e-05,
+      "loss": 1.1598,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3582187654050501,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.6205382000502887e-05,
+      "loss": 1.1226,
+      "step": 3270
+    },
+    {
+      "epoch": 0.35876650052034836,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.619037637328046e-05,
+      "loss": 1.1599,
+      "step": 3275
+    },
+    {
+      "epoch": 0.35931423563564663,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.617534811308839e-05,
+      "loss": 1.1609,
+      "step": 3280
+    },
+    {
+      "epoch": 0.35986197075094484,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.61602972748723e-05,
+      "loss": 1.1611,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3604097058662431,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6145223913660378e-05,
+      "loss": 1.1171,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3609574409815413,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.613012808456316e-05,
+      "loss": 1.1123,
+      "step": 3295
+    },
+    {
+      "epoch": 0.3615051760968396,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.6115009842773322e-05,
+      "loss": 1.1953,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3620529112121378,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.609986924356548e-05,
+      "loss": 1.0976,
+      "step": 3305
+    },
+    {
+      "epoch": 0.36260064632743605,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.6084706342295994e-05,
+      "loss": 1.1681,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3631483814427343,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.6069521194402776e-05,
+      "loss": 1.1529,
+      "step": 3315
+    },
+    {
+      "epoch": 0.3636961165580325,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.605431385540506e-05,
+      "loss": 1.1988,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3642438516733308,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6039084380903222e-05,
+      "loss": 1.1394,
+      "step": 3325
+    },
+    {
+      "epoch": 0.364791586788629,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.602383282657857e-05,
+      "loss": 1.1255,
+      "step": 3330
+    },
+    {
+      "epoch": 0.36533932190392726,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6008559248193128e-05,
+      "loss": 1.1545,
+      "step": 3335
+    },
+    {
+      "epoch": 0.3658870570192255,
+      "grad_norm": 1.25,
+      "learning_rate": 1.599326370158946e-05,
+      "loss": 1.1661,
+      "step": 3340
+    },
+    {
+      "epoch": 0.36643479213452373,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.597794624269043e-05,
+      "loss": 1.1714,
+      "step": 3345
+    },
+    {
+      "epoch": 0.366982527249822,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5962606927499043e-05,
+      "loss": 1.1919,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3675302623651202,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5947245812098186e-05,
+      "loss": 1.1691,
+      "step": 3355
+    },
+    {
+      "epoch": 0.3680779974804185,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5931862952650466e-05,
+      "loss": 1.1454,
+      "step": 3360
+    },
+    {
+      "epoch": 0.36862573259571674,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.591645840539799e-05,
+      "loss": 1.1235,
+      "step": 3365
+    },
+    {
+      "epoch": 0.36917346771101495,
+      "grad_norm": 1.125,
+      "learning_rate": 1.5901032226662154e-05,
+      "loss": 1.1144,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3697212028263132,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.588558447284344e-05,
+      "loss": 1.0976,
+      "step": 3375
+    },
+    {
+      "epoch": 0.3702689379416114,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5870115200421223e-05,
+      "loss": 1.1143,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3708166730569097,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5854624465953537e-05,
+      "loss": 1.1874,
+      "step": 3385
+    },
+    {
+      "epoch": 0.3713644081722079,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5839112326076905e-05,
+      "loss": 1.1331,
+      "step": 3390
+    },
+    {
+      "epoch": 0.37191214328750616,
+      "grad_norm": 1.25,
+      "learning_rate": 1.582357883750609e-05,
+      "loss": 1.2034,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3724598784028044,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5808024057033927e-05,
+      "loss": 1.1387,
+      "step": 3400
+    },
+    {
+      "epoch": 0.37300761351810263,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.579244804153108e-05,
+      "loss": 1.1715,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3735553486334009,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5776850847945867e-05,
+      "loss": 1.1801,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3741030837486991,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.5761232533304034e-05,
+      "loss": 1.092,
+      "step": 3415
+    },
+    {
+      "epoch": 0.37465081886399737,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.5745593154708543e-05,
+      "loss": 1.1536,
+      "step": 3420
+    },
+    {
+      "epoch": 0.37519855397929563,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5729932769339366e-05,
+      "loss": 1.1283,
+      "step": 3425
+    },
+    {
+      "epoch": 0.37574628909459384,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.571425143445329e-05,
+      "loss": 1.1645,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3762940242098921,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.5698549207383687e-05,
+      "loss": 1.1203,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3768417593251903,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5682826145540324e-05,
+      "loss": 1.1555,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3773894944404886,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.566708230640913e-05,
+      "loss": 1.1614,
+      "step": 3445
+    },
+    {
+      "epoch": 0.37793722955578685,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5651317747552014e-05,
+      "loss": 1.1987,
+      "step": 3450
+    },
+    {
+      "epoch": 0.37848496467108506,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.5635532526606625e-05,
+      "loss": 1.1825,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3790326997863833,
+      "grad_norm": 1.25,
+      "learning_rate": 1.5619726701286167e-05,
+      "loss": 1.1229,
+      "step": 3460
+    },
+    {
+      "epoch": 0.37958043490168153,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5603900329379168e-05,
+      "loss": 1.1509,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3801281700169798,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5588053468749285e-05,
+      "loss": 1.185,
+      "step": 3470
+    },
+    {
+      "epoch": 0.380675905132278,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5572186177335084e-05,
+      "loss": 1.1458,
+      "step": 3475
+    },
+    {
+      "epoch": 0.38122364024757627,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.555629851314982e-05,
+      "loss": 1.1651,
+      "step": 3480
+    },
+    {
+      "epoch": 0.38177137536287453,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5540390534281245e-05,
+      "loss": 1.1313,
+      "step": 3485
+    },
+    {
+      "epoch": 0.38231911047817274,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5524462298891377e-05,
+      "loss": 1.2005,
+      "step": 3490
+    },
+    {
+      "epoch": 0.382866845593471,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5508513865216306e-05,
+      "loss": 1.1227,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3834145807087692,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.5492545291565953e-05,
+      "loss": 1.1439,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3839623158240675,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5476556636323893e-05,
+      "loss": 1.1816,
+      "step": 3505
+    },
+    {
+      "epoch": 0.38451005093936574,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5460547957947105e-05,
+      "loss": 1.124,
+      "step": 3510
+    },
+    {
+      "epoch": 0.38505778605466395,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5444519314965782e-05,
+      "loss": 1.1489,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3856055211699622,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.542847076598312e-05,
+      "loss": 1.1407,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3861532562852604,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5412402369675082e-05,
+      "loss": 1.2599,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3867009914005587,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5396314184790194e-05,
+      "loss": 1.1772,
+      "step": 3530
+    },
+    {
+      "epoch": 0.38724872651585696,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5380206270149353e-05,
+      "loss": 1.1525,
+      "step": 3535
+    },
+    {
+      "epoch": 0.38779646163115516,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.536407868464556e-05,
+      "loss": 1.1601,
+      "step": 3540
+    },
+    {
+      "epoch": 0.38834419674645343,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.534793148724376e-05,
+      "loss": 1.1402,
+      "step": 3545
+    },
+    {
+      "epoch": 0.38889193186175164,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5331764736980602e-05,
+      "loss": 1.1789,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3894396669770499,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.5315578492964203e-05,
+      "loss": 1.1397,
+      "step": 3555
+    },
+    {
+      "epoch": 0.38998740209234817,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5299372814373967e-05,
+      "loss": 1.183,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3905351372076464,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5283147760460354e-05,
+      "loss": 1.1203,
+      "step": 3565
+    },
+    {
+      "epoch": 0.39108287232294464,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.5266903390544662e-05,
+      "loss": 1.0783,
+      "step": 3570
+    },
+    {
+      "epoch": 0.39163060743824285,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5250639764018807e-05,
+      "loss": 1.1258,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3921783425535411,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5234356940345115e-05,
+      "loss": 1.158,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3927260776688393,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5218054979056093e-05,
+      "loss": 1.1331,
+      "step": 3585
+    },
+    {
+      "epoch": 0.3932738127841376,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5201733939754228e-05,
+      "loss": 1.2043,
+      "step": 3590
+    },
+    {
+      "epoch": 0.39382154789943585,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5185393882111756e-05,
+      "loss": 1.1403,
+      "step": 3595
+    },
+    {
+      "epoch": 0.39436928301473406,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5169034865870438e-05,
+      "loss": 1.1392,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3949170181300323,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.5152656950841361e-05,
+      "loss": 1.1757,
+      "step": 3605
+    },
+    {
+      "epoch": 0.39546475324533054,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5136260196904704e-05,
+      "loss": 1.1485,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3960124883606288,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5119844664009526e-05,
+      "loss": 1.1333,
+      "step": 3615
+    },
+    {
+      "epoch": 0.39656022347592707,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.510341041217355e-05,
+      "loss": 1.1758,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3971079585912253,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.508695750148292e-05,
+      "loss": 1.1336,
+      "step": 3625
+    },
+    {
+      "epoch": 0.39765569370652354,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.5070485992092023e-05,
+      "loss": 1.1332,
+      "step": 3630
+    },
+    {
+      "epoch": 0.39820342882182175,
+      "grad_norm": 1.25,
+      "learning_rate": 1.505399594422323e-05,
+      "loss": 1.2085,
+      "step": 3635
+    },
+    {
+      "epoch": 0.39875116393712,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5037487418166696e-05,
+      "loss": 1.1608,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3992988990524183,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5020960474280137e-05,
+      "loss": 1.1443,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3998466341677165,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5004415172988606e-05,
+      "loss": 1.1886,
+      "step": 3650
+    },
+    {
+      "epoch": 0.40039436928301475,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4987851574784271e-05,
+      "loss": 1.1719,
+      "step": 3655
+    },
+    {
+      "epoch": 0.40094210439831296,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.4971269740226203e-05,
+      "loss": 1.1735,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4014898395136112,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4954669729940137e-05,
+      "loss": 1.141,
+      "step": 3665
+    },
+    {
+      "epoch": 0.40203757462890943,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4938051604618275e-05,
+      "loss": 1.1463,
+      "step": 3670
+    },
+    {
+      "epoch": 0.4025853097442077,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.4921415425019039e-05,
+      "loss": 1.1986,
+      "step": 3675
+    },
+    {
+      "epoch": 0.40313304485950596,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4904761251966864e-05,
+      "loss": 1.15,
+      "step": 3680
+    },
+    {
+      "epoch": 0.40368077997480417,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4888089146351971e-05,
+      "loss": 1.1433,
+      "step": 3685
+    },
+    {
+      "epoch": 0.40422851509010244,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4871399169130157e-05,
+      "loss": 1.1749,
+      "step": 3690
+    },
+    {
+      "epoch": 0.40477625020540065,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4854691381322538e-05,
+      "loss": 1.1315,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4053239853206989,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4837965844015359e-05,
+      "loss": 1.1184,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4058717204359972,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4821222618359769e-05,
+      "loss": 1.1398,
+      "step": 3705
+    },
+    {
+      "epoch": 0.4064194555512954,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.480446176557158e-05,
+      "loss": 1.1341,
+      "step": 3710
+    },
+    {
+      "epoch": 0.40696719066659365,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.4787683346931047e-05,
+      "loss": 1.1129,
+      "step": 3715
+    },
+    {
+      "epoch": 0.40751492578189186,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4770887423782654e-05,
+      "loss": 1.1532,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4080626608971901,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.4754074057534885e-05,
+      "loss": 1.0848,
+      "step": 3725
+    },
+    {
+      "epoch": 0.4086103960124884,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4737243309659998e-05,
+      "loss": 1.1446,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4091581311277866,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4720395241693796e-05,
+      "loss": 1.1819,
+      "step": 3735
+    },
+    {
+      "epoch": 0.40970586624308486,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4703529915235417e-05,
+      "loss": 1.1483,
+      "step": 3740
+    },
+    {
+      "epoch": 0.41025360135838307,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.468664739194709e-05,
+      "loss": 1.1191,
+      "step": 3745
+    },
+    {
+      "epoch": 0.41080133647368133,
+      "grad_norm": 1.25,
+      "learning_rate": 1.4669747733553917e-05,
+      "loss": 1.1227,
+      "step": 3750
+    },
+    {
+      "epoch": 0.41134907158897954,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4652831001843656e-05,
+      "loss": 1.1884,
+      "step": 3755
+    },
+    {
+      "epoch": 0.4118968067042778,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4635897258666484e-05,
+      "loss": 1.1812,
+      "step": 3760
+    },
+    {
+      "epoch": 0.41244454181957607,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4618946565934775e-05,
+      "loss": 1.1986,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4129922769348743,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4601978985622874e-05,
+      "loss": 1.2179,
+      "step": 3770
+    },
+    {
+      "epoch": 0.41354001205017255,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4584994579766865e-05,
+      "loss": 1.1142,
+      "step": 3775
+    },
+    {
+      "epoch": 0.41408774716547075,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4567993410464354e-05,
+      "loss": 1.1743,
+      "step": 3780
+    },
+    {
+      "epoch": 0.414635482280769,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4550975539874233e-05,
+      "loss": 1.2144,
+      "step": 3785
+    },
+    {
+      "epoch": 0.4151832173960673,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4533941030216466e-05,
+      "loss": 1.1351,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4157309525113655,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4516889943771835e-05,
+      "loss": 1.1299,
+      "step": 3795
+    },
+    {
+      "epoch": 0.41627868762666376,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4499822342881744e-05,
+      "loss": 1.1562,
+      "step": 3800
+    },
+    {
+      "epoch": 0.41682642274196197,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4482738289947968e-05,
+      "loss": 1.1797,
+      "step": 3805
+    },
+    {
+      "epoch": 0.41737415785726023,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4465637847432444e-05,
+      "loss": 1.1477,
+      "step": 3810
+    },
+    {
+      "epoch": 0.4179218929725585,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.4448521077857013e-05,
+      "loss": 1.1662,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4184696280878567,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4431388043803227e-05,
+      "loss": 1.1298,
+      "step": 3820
+    },
+    {
+      "epoch": 0.41901736320315497,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.44142388079121e-05,
+      "loss": 1.1517,
+      "step": 3825
+    },
+    {
+      "epoch": 0.4195650983184532,
+      "grad_norm": 1.25,
+      "learning_rate": 1.439707343288388e-05,
+      "loss": 1.1491,
+      "step": 3830
+    },
+    {
+      "epoch": 0.42011283343375144,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4379891981477816e-05,
+      "loss": 1.1864,
+      "step": 3835
+    },
+    {
+      "epoch": 0.4206605685490497,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4362694516511946e-05,
+      "loss": 1.1392,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4212083036643479,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.434548110086285e-05,
+      "loss": 1.1222,
+      "step": 3845
+    },
+    {
+      "epoch": 0.4217560387796462,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.4328251797465434e-05,
+      "loss": 1.1979,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4223037738949444,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4311006669312672e-05,
+      "loss": 1.2049,
+      "step": 3855
+    },
+    {
+      "epoch": 0.42285150901024265,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.4293745779455418e-05,
+      "loss": 1.1786,
+      "step": 3860
+    },
+    {
+      "epoch": 0.42339924412554086,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4276469191002147e-05,
+      "loss": 1.1306,
+      "step": 3865
+    },
+    {
+      "epoch": 0.42394697924083913,
+      "grad_norm": 1.25,
+      "learning_rate": 1.4259176967118719e-05,
+      "loss": 1.1525,
+      "step": 3870
+    },
+    {
+      "epoch": 0.4244947143561374,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4241869171028178e-05,
+      "loss": 1.153,
+      "step": 3875
+    },
+    {
+      "epoch": 0.4250424494714356,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4224545866010484e-05,
+      "loss": 1.1271,
+      "step": 3880
+    },
+    {
+      "epoch": 0.42559018458673387,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4207207115402316e-05,
+      "loss": 1.1419,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4261379197020321,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.4189852982596813e-05,
+      "loss": 1.1326,
+      "step": 3890
+    },
+    {
+      "epoch": 0.42668565481733034,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4172483531043358e-05,
+      "loss": 1.1815,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4272333899326286,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.4155098824247341e-05,
+      "loss": 1.1368,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4277811250479268,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.4137698925769931e-05,
+      "loss": 1.218,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4283288601632251,
+      "grad_norm": 1.25,
+      "learning_rate": 1.4120283899227839e-05,
+      "loss": 1.1762,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4288765952785233,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4102853808293077e-05,
+      "loss": 1.15,
+      "step": 3915
+    },
+    {
+      "epoch": 0.42942433039382155,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.408540871669275e-05,
+      "loss": 1.1668,
+      "step": 3920
+    },
+    {
+      "epoch": 0.4299720655091198,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.4067948688208799e-05,
+      "loss": 1.1276,
+      "step": 3925
+    },
+    {
+      "epoch": 0.430519800624418,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4050473786677774e-05,
+      "loss": 1.189,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4310675357397163,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.403298407599061e-05,
+      "loss": 1.1595,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4316152708550145,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4015479620092383e-05,
+      "loss": 1.1478,
+      "step": 3940
+    },
+    {
+      "epoch": 0.43216300597031276,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.3997960482982082e-05,
+      "loss": 1.1392,
+      "step": 3945
+    },
+    {
+      "epoch": 0.432710741085611,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3980426728712369e-05,
+      "loss": 1.1706,
+      "step": 3950
+    },
+    {
+      "epoch": 0.43325847620090924,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.396287842138935e-05,
+      "loss": 1.1913,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4338062113162075,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.394531562517234e-05,
+      "loss": 1.1721,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4343539464315057,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.3927738404273634e-05,
+      "loss": 1.2269,
+      "step": 3965
+    },
+    {
+      "epoch": 0.434901681546804,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.391014682295825e-05,
+      "loss": 1.1459,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4354494166621022,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3892540945543722e-05,
+      "loss": 1.117,
+      "step": 3975
+    },
+    {
+      "epoch": 0.43599715177740045,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3874920836399854e-05,
+      "loss": 1.1366,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4365448868926987,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3857286559948476e-05,
+      "loss": 1.178,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4370926220079969,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.383963818066322e-05,
+      "loss": 1.0893,
+      "step": 3990
+    },
+    {
+      "epoch": 0.4376403571232952,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3821975763069279e-05,
+      "loss": 1.1656,
+      "step": 3995
+    },
+    {
+      "epoch": 0.4381880922385934,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3804299371743174e-05,
+      "loss": 1.1316,
+      "step": 4000
+    },
+    {
+      "epoch": 0.43873582735389166,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.3786609071312511e-05,
+      "loss": 1.1865,
+      "step": 4005
+    },
+    {
+      "epoch": 0.4392835624691899,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.376890492645576e-05,
+      "loss": 1.1552,
+      "step": 4010
+    },
+    {
+      "epoch": 0.43983129758448813,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3751187001901995e-05,
+      "loss": 1.1758,
+      "step": 4015
+    },
+    {
+      "epoch": 0.4403790326997864,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.3733455362430684e-05,
+      "loss": 1.1494,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4409267678150846,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3715710072871426e-05,
+      "loss": 1.1465,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4414745029303829,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.369795119810374e-05,
+      "loss": 1.1678,
+      "step": 4030
+    },
+    {
+      "epoch": 0.4420222380456811,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3680178803056802e-05,
+      "loss": 1.1544,
+      "step": 4035
+    },
+    {
+      "epoch": 0.44256997316097935,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.366239295270923e-05,
+      "loss": 1.1633,
+      "step": 4040
+    },
+    {
+      "epoch": 0.4431177082762776,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.3644593712088829e-05,
+      "loss": 1.1644,
+      "step": 4045
+    },
+    {
+      "epoch": 0.4436654433915758,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3626781146272369e-05,
+      "loss": 1.1366,
+      "step": 4050
+    },
+    {
+      "epoch": 0.4442131785068741,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3608955320385333e-05,
+      "loss": 1.1761,
+      "step": 4055
+    },
+    {
+      "epoch": 0.4447609136221723,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.3591116299601684e-05,
+      "loss": 1.1194,
+      "step": 4060
+    },
+    {
+      "epoch": 0.44530864873747056,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3573264149143636e-05,
+      "loss": 1.1779,
+      "step": 4065
+    },
+    {
+      "epoch": 0.4458563838527688,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.3555398934281397e-05,
+      "loss": 1.1487,
+      "step": 4070
+    },
+    {
+      "epoch": 0.44640411896806703,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3537520720332943e-05,
+      "loss": 1.1687,
+      "step": 4075
+    },
+    {
+      "epoch": 0.4469518540833653,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.351962957266378e-05,
+      "loss": 1.1544,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4474995891986635,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3501725556686702e-05,
+      "loss": 1.1358,
+      "step": 4085
+    },
+    {
+      "epoch": 0.44804732431396177,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3483808737861547e-05,
+      "loss": 1.1723,
+      "step": 4090
+    },
+    {
+      "epoch": 0.44859505942926003,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.3465879181694966e-05,
+      "loss": 1.1463,
+      "step": 4095
+    },
+    {
+      "epoch": 0.44914279454455824,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.344793695374018e-05,
+      "loss": 1.15,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4496905296598565,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3429982119596737e-05,
+      "loss": 1.2038,
+      "step": 4105
+    },
+    {
+      "epoch": 0.4502382647751547,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.341201474491027e-05,
+      "loss": 1.1767,
+      "step": 4110
+    },
+    {
+      "epoch": 0.450785999890453,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3394034895372281e-05,
+      "loss": 1.173,
+      "step": 4115
+    },
+    {
+      "epoch": 0.4513337350057512,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.337604263671986e-05,
+      "loss": 1.183,
+      "step": 4120
+    },
+    {
+      "epoch": 0.45188147012104946,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3358038034735485e-05,
+      "loss": 1.1281,
+      "step": 4125
+    },
+    {
+      "epoch": 0.4524292052363477,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.3340021155246745e-05,
+      "loss": 1.187,
+      "step": 4130
+    },
+    {
+      "epoch": 0.45297694035164593,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3321992064126132e-05,
+      "loss": 1.1543,
+      "step": 4135
+    },
+    {
+      "epoch": 0.4535246754669442,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3303950827290781e-05,
+      "loss": 1.191,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4540724105822424,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.328589751070223e-05,
+      "loss": 1.2142,
+      "step": 4145
+    },
+    {
+      "epoch": 0.45462014569754067,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.3267832180366189e-05,
+      "loss": 1.1444,
+      "step": 4150
+    },
+    {
+      "epoch": 0.45516788081283893,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3249754902332285e-05,
+      "loss": 1.1635,
+      "step": 4155
+    },
+    {
+      "epoch": 0.45571561592813714,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3231665742693838e-05,
+      "loss": 1.156,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4562633510434354,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.3213564767587594e-05,
+      "loss": 1.2229,
+      "step": 4165
+    },
+    {
+      "epoch": 0.4568110861587336,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3195452043193511e-05,
+      "loss": 1.1456,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4573588212740319,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3177327635734497e-05,
+      "loss": 1.1082,
+      "step": 4175
+    },
+    {
+      "epoch": 0.45790655638933014,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3159191611476183e-05,
+      "loss": 1.1428,
+      "step": 4180
+    },
+    {
+      "epoch": 0.45845429150462835,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3141044036726663e-05,
+      "loss": 1.1755,
+      "step": 4185
+    },
+    {
+      "epoch": 0.4590020266199266,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3122884977836265e-05,
+      "loss": 1.1714,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4595497617352248,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.3104714501197308e-05,
+      "loss": 1.1976,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4600974968505231,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.3086532673243855e-05,
+      "loss": 1.1892,
+      "step": 4200
+    },
+    {
+      "epoch": 0.46064523196582136,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.306833956045147e-05,
+      "loss": 1.1116,
+      "step": 4205
+    },
+    {
+      "epoch": 0.46119296708111956,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.3050135229336974e-05,
+      "loss": 1.1555,
+      "step": 4210
+    },
+    {
+      "epoch": 0.46174070219641783,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3031919746458202e-05,
+      "loss": 1.1401,
+      "step": 4215
+    },
+    {
+      "epoch": 0.46228843731171604,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3013693178413773e-05,
+      "loss": 1.144,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4628361724270143,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2995455591842825e-05,
+      "loss": 1.1727,
+      "step": 4225
+    },
+    {
+      "epoch": 0.4633839075423125,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.2977207053424781e-05,
+      "loss": 1.2165,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4639316426576108,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2958947629879113e-05,
+      "loss": 1.1501,
+      "step": 4235
+    },
+    {
+      "epoch": 0.46447937777290904,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2940677387965083e-05,
+      "loss": 1.244,
+      "step": 4240
+    },
+    {
+      "epoch": 0.46502711288820725,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2922396394481513e-05,
+      "loss": 1.1096,
+      "step": 4245
+    },
+    {
+      "epoch": 0.4655748480035055,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2904104716266529e-05,
+      "loss": 1.145,
+      "step": 4250
+    },
+    {
+      "epoch": 0.4661225831188037,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2885802420197324e-05,
+      "loss": 1.1838,
+      "step": 4255
+    },
+    {
+      "epoch": 0.466670318234102,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2867489573189916e-05,
+      "loss": 1.1849,
+      "step": 4260
+    },
+    {
+      "epoch": 0.46721805334940025,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2849166242198887e-05,
+      "loss": 1.1366,
+      "step": 4265
+    },
+    {
+      "epoch": 0.46776578846469846,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2830832494217167e-05,
+      "loss": 1.1374,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4683135235799967,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2812488396275757e-05,
+      "loss": 1.1513,
+      "step": 4275
+    },
+    {
+      "epoch": 0.46886125869529494,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.2794134015443508e-05,
+      "loss": 1.1682,
+      "step": 4280
+    },
+    {
+      "epoch": 0.4694089938105932,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.2775769418826858e-05,
+      "loss": 1.1919,
+      "step": 4285
+    },
+    {
+      "epoch": 0.46995672892589146,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.2757394673569608e-05,
+      "loss": 1.1142,
+      "step": 4290
+    },
+    {
+      "epoch": 0.4705044640411897,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.2739009846852655e-05,
+      "loss": 1.1921,
+      "step": 4295
+    },
+    {
+      "epoch": 0.47105219915648794,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2720615005893758e-05,
+      "loss": 1.154,
+      "step": 4300
+    },
+    {
+      "epoch": 0.47159993427178615,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.2702210217947289e-05,
+      "loss": 1.1875,
+      "step": 4305
+    },
+    {
+      "epoch": 0.4721476693870844,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.2683795550303985e-05,
+      "loss": 1.1543,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4726954045023826,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.266537107029071e-05,
+      "loss": 1.1906,
+      "step": 4315
+    },
+    {
+      "epoch": 0.4732431396176809,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.2646936845270208e-05,
+      "loss": 1.0545,
+      "step": 4320
+    },
+    {
+      "epoch": 0.47379087473297915,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2628492942640835e-05,
+      "loss": 1.1769,
+      "step": 4325
+    },
+    {
+      "epoch": 0.47433860984827736,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2610039429836345e-05,
+      "loss": 1.1138,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4748863449635756,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.2591576374325622e-05,
+      "loss": 1.1499,
+      "step": 4335
+    },
+    {
+      "epoch": 0.47543408007887383,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.2573103843612447e-05,
+      "loss": 1.1579,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4759818151941721,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.2554621905235226e-05,
+      "loss": 1.1361,
+      "step": 4345
+    },
+    {
+      "epoch": 0.47652955030947036,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2536130626766783e-05,
+      "loss": 1.2137,
+      "step": 4350
+    },
+    {
+      "epoch": 0.47707728542476857,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2517630075814078e-05,
+      "loss": 1.1736,
+      "step": 4355
+    },
+    {
+      "epoch": 0.47762502054006684,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2499120320017977e-05,
+      "loss": 1.1649,
+      "step": 4360
+    },
+    {
+      "epoch": 0.47817275565536504,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2480601427052994e-05,
+      "loss": 1.1655,
+      "step": 4365
+    },
+    {
+      "epoch": 0.4787204907706633,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2462073464627058e-05,
+      "loss": 1.1312,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4792682258859616,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2443536500481252e-05,
+      "loss": 1.1196,
+      "step": 4375
+    },
+    {
+      "epoch": 0.4798159610012598,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2424990602389578e-05,
+      "loss": 1.1134,
+      "step": 4380
+    },
+    {
+      "epoch": 0.48036369611655805,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2406435838158686e-05,
+      "loss": 1.2028,
+      "step": 4385
+    },
+    {
+      "epoch": 0.48091143123185626,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.2387872275627659e-05,
+      "loss": 1.0878,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4814591663471545,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2369299982667744e-05,
+      "loss": 1.1652,
+      "step": 4395
+    },
+    {
+      "epoch": 0.48200690146245273,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.23507190271821e-05,
+      "loss": 1.1133,
+      "step": 4400
+    },
+    {
+      "epoch": 0.482554636577751,
+      "grad_norm": 1.375,
+      "learning_rate": 1.2332129477105562e-05,
+      "loss": 1.1953,
+      "step": 4405
+    },
+    {
+      "epoch": 0.48310237169304926,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2313531400404397e-05,
+      "loss": 1.1242,
+      "step": 4410
+    },
+    {
+      "epoch": 0.48365010680834747,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2294924865076029e-05,
+      "loss": 1.1628,
+      "step": 4415
+    },
+    {
+      "epoch": 0.48419784192364573,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.227630993914882e-05,
+      "loss": 1.1063,
+      "step": 4420
+    },
+    {
+      "epoch": 0.48474557703894394,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2257686690681812e-05,
+      "loss": 1.1294,
+      "step": 4425
+    },
+    {
+      "epoch": 0.4852933121542422,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2239055187764463e-05,
+      "loss": 1.1627,
+      "step": 4430
+    },
+    {
+      "epoch": 0.48584104726954047,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.222041549851642e-05,
+      "loss": 1.1864,
+      "step": 4435
+    },
+    {
+      "epoch": 0.4863887823848387,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2201767691087265e-05,
+      "loss": 1.113,
+      "step": 4440
+    },
+    {
+      "epoch": 0.48693651750013695,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2183111833656244e-05,
+      "loss": 1.1931,
+      "step": 4445
+    },
+    {
+      "epoch": 0.48748425261543515,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2164447994432054e-05,
+      "loss": 1.1258,
+      "step": 4450
+    },
+    {
+      "epoch": 0.4880319877307334,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.214577624165256e-05,
+      "loss": 1.1602,
+      "step": 4455
+    },
+    {
+      "epoch": 0.4885797228460317,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2127096643584573e-05,
+      "loss": 1.1441,
+      "step": 4460
+    },
+    {
+      "epoch": 0.4891274579613299,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.2108409268523578e-05,
+      "loss": 1.1582,
+      "step": 4465
+    },
+    {
+      "epoch": 0.48967519307662816,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2089714184793492e-05,
+      "loss": 1.1229,
+      "step": 4470
+    },
+    {
+      "epoch": 0.49022292819192637,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.207101146074643e-05,
+      "loss": 1.1002,
+      "step": 4475
+    },
+    {
+      "epoch": 0.49077066330722463,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.205230116476243e-05,
+      "loss": 1.1865,
+      "step": 4480
+    },
+    {
+      "epoch": 0.4913183984225229,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.203358336524921e-05,
+      "loss": 1.2106,
+      "step": 4485
+    },
+    {
+      "epoch": 0.4918661335378211,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.2014858130641936e-05,
+      "loss": 1.1417,
+      "step": 4490
+    },
+    {
+      "epoch": 0.49241386865311937,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1996125529402946e-05,
+      "loss": 1.1356,
+      "step": 4495
+    },
+    {
+      "epoch": 0.4929616037684176,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1977385630021523e-05,
+      "loss": 1.1409,
+      "step": 4500
+    },
+    {
+      "epoch": 0.49350933888371584,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1958638501013613e-05,
+      "loss": 1.129,
+      "step": 4505
+    },
+    {
+      "epoch": 0.49405707399901405,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1939884210921618e-05,
+      "loss": 1.1589,
+      "step": 4510
+    },
+    {
+      "epoch": 0.4946048091143123,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1921122828314109e-05,
+      "loss": 1.0934,
+      "step": 4515
+    },
+    {
+      "epoch": 0.4951525442296106,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.1902354421785591e-05,
+      "loss": 1.1568,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4957002793449088,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.188357905995625e-05,
+      "loss": 1.151,
+      "step": 4525
+    },
+    {
+      "epoch": 0.49624801446020705,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.18647968114717e-05,
+      "loss": 1.1709,
+      "step": 4530
+    },
+    {
+      "epoch": 0.49679574957550526,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.1846007745002734e-05,
+      "loss": 1.1514,
+      "step": 4535
+    },
+    {
+      "epoch": 0.49734348469080353,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.1827211929245075e-05,
+      "loss": 1.1487,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4978912198061018,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1808409432919124e-05,
+      "loss": 1.1874,
+      "step": 4545
+    },
+    {
+      "epoch": 0.4984389549214,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1789600324769696e-05,
+      "loss": 1.1591,
+      "step": 4550
+    },
+    {
+      "epoch": 0.49898669003669827,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.1770784673565796e-05,
+      "loss": 1.1598,
+      "step": 4555
+    },
+    {
+      "epoch": 0.4995344251519965,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1751962548100339e-05,
+      "loss": 1.1619,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5000821602672947,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1733134017189918e-05,
+      "loss": 1.1255,
+      "step": 4565
+    },
+    {
+      "epoch": 0.500629895382593,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.1714299149674538e-05,
+      "loss": 1.136,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5011776304978912,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.1695458014417382e-05,
+      "loss": 1.1483,
+      "step": 4575
+    },
+    {
+      "epoch": 0.5017253656131895,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1676610680304539e-05,
+      "loss": 1.1182,
+      "step": 4580
+    },
+    {
+      "epoch": 0.5022731007284877,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1657757216244767e-05,
+      "loss": 1.118,
+      "step": 4585
+    },
+    {
+      "epoch": 0.5028208358437859,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.163889769116923e-05,
+      "loss": 1.1345,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5033685709590842,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.162003217403127e-05,
+      "loss": 1.1254,
+      "step": 4595
+    },
+    {
+      "epoch": 0.5039163060743824,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1601160733806113e-05,
+      "loss": 1.125,
+      "step": 4600
+    },
+    {
+      "epoch": 0.5044640411896807,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.1582283439490652e-05,
+      "loss": 1.2014,
+      "step": 4605
+    },
+    {
+      "epoch": 0.505011776304979,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1563400360103188e-05,
+      "loss": 1.1424,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5055595114202771,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1544511564683165e-05,
+      "loss": 1.1401,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5061072465355754,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.152561712229093e-05,
+      "loss": 1.1839,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5066549816508736,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1506717102007474e-05,
+      "loss": 1.1592,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5072027167661719,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1487811572934184e-05,
+      "loss": 1.1217,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5077504518814702,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1468900604192585e-05,
+      "loss": 1.1532,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5082981869967683,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1449984264924094e-05,
+      "loss": 1.1942,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5088459221120666,
+      "grad_norm": 1.25,
+      "learning_rate": 1.143106262428976e-05,
+      "loss": 1.0977,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5093936572273648,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1412135751470017e-05,
+      "loss": 1.1754,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5099413923426631,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.1393203715664427e-05,
+      "loss": 1.131,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5104891274579614,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.137426658609143e-05,
+      "loss": 1.1093,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5110368625732595,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1355324431988086e-05,
+      "loss": 1.1413,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5115845976885578,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.1336377322609832e-05,
+      "loss": 1.1897,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5121323328038561,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.131742532723022e-05,
+      "loss": 1.1527,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5126800679191543,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1298468515140662e-05,
+      "loss": 1.1567,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5132278030344526,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1279506955650182e-05,
+      "loss": 1.1145,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5137755381497507,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1260540718085162e-05,
+      "loss": 1.1819,
+      "step": 4690
+    },
+    {
+      "epoch": 0.514323273265049,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1241569871789096e-05,
+      "loss": 1.1306,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5148710083803473,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1222594486122312e-05,
+      "loss": 1.1663,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5154187434956455,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1203614630461746e-05,
+      "loss": 1.1199,
+      "step": 4705
+    },
+    {
+      "epoch": 0.5159664786109438,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.118463037420067e-05,
+      "loss": 1.1542,
+      "step": 4710
+    },
+    {
+      "epoch": 0.516514213726242,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.116564178674846e-05,
+      "loss": 1.1904,
+      "step": 4715
+    },
+    {
+      "epoch": 0.5170619488415402,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1146648937530309e-05,
+      "loss": 1.1787,
+      "step": 4720
+    },
+    {
+      "epoch": 0.5176096839568385,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1127651895986999e-05,
+      "loss": 1.163,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5181574190721367,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1108650731574644e-05,
+      "loss": 1.0855,
+      "step": 4730
+    },
+    {
+      "epoch": 0.5187051541874349,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1089645513764429e-05,
+      "loss": 1.1962,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5192528893027332,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1070636312042352e-05,
+      "loss": 1.2023,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5198006244180314,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1051623195908987e-05,
+      "loss": 1.1834,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5203483595333297,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1032606234879217e-05,
+      "loss": 1.172,
+      "step": 4750
+    },
+    {
+      "epoch": 0.520896094648628,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1013585498481983e-05,
+      "loss": 1.1487,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5214438297639261,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.0994561056260016e-05,
+      "loss": 1.2174,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5219915648792244,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.0975532977769619e-05,
+      "loss": 1.1955,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5225392999945226,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0956501332580375e-05,
+      "loss": 1.0925,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5230870351098209,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.093746619027491e-05,
+      "loss": 1.181,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5236347702251192,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.0918427620448635e-05,
+      "loss": 1.1976,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5241825053404173,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0899385692709499e-05,
+      "loss": 1.1783,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5247302404557156,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0880340476677718e-05,
+      "loss": 1.0751,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5252779755710139,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0861292041985538e-05,
+      "loss": 1.1422,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5258257106863121,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.084224045827697e-05,
+      "loss": 1.1123,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5263734458016104,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.082318579520754e-05,
+      "loss": 1.1201,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5269211809169085,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.080412812244403e-05,
+      "loss": 1.1319,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5274689160322068,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0785067509664231e-05,
+      "loss": 1.1475,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5280166511475051,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.0766004026556676e-05,
+      "loss": 1.1697,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5285643862628033,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0746937742820397e-05,
+      "loss": 1.1328,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5291121213781016,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.072786872816466e-05,
+      "loss": 1.1438,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5296598564933997,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.070879705230873e-05,
+      "loss": 1.1648,
+      "step": 4835
+    },
+    {
+      "epoch": 0.530207591608698,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.0689722784981586e-05,
+      "loss": 1.113,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5307553267239963,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.0670645995921687e-05,
+      "loss": 1.1427,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5313030618392945,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0651566754876715e-05,
+      "loss": 1.1771,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5318507969545928,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0632485131603313e-05,
+      "loss": 1.1629,
+      "step": 4855
+    },
+    {
+      "epoch": 0.532398532069891,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0613401195866835e-05,
+      "loss": 1.2022,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5329462671851892,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.059431501744109e-05,
+      "loss": 1.1548,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5334940023004875,
+      "grad_norm": 1.25,
+      "learning_rate": 1.0575226666108086e-05,
+      "loss": 1.1427,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5340417374157858,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0556136211657784e-05,
+      "loss": 1.1265,
+      "step": 4875
+    },
+    {
+      "epoch": 0.534589472531084,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0537043723887811e-05,
+      "loss": 1.1737,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5351372076463822,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0517949272603257e-05,
+      "loss": 1.081,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5356849427616804,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.049885292761637e-05,
+      "loss": 1.1054,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5362326778769787,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.0479754758746332e-05,
+      "loss": 1.1456,
+      "step": 4895
+    },
+    {
+      "epoch": 0.536780412992277,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.0460654835818989e-05,
+      "loss": 1.1402,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5373281481075752,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0441553228666603e-05,
+      "loss": 1.1627,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5378758832228734,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0422450007127591e-05,
+      "loss": 1.1311,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5384236183381717,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0403345241046277e-05,
+      "loss": 1.215,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5389713534534699,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.0384239000272624e-05,
+      "loss": 1.1007,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5395190885687682,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0365131354661995e-05,
+      "loss": 1.0988,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5400668236840663,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0346022374074885e-05,
+      "loss": 1.1654,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5406145587993646,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.032691212837667e-05,
+      "loss": 1.1612,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5411622939146629,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.0307800687437352e-05,
+      "loss": 1.1963,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5417100290299611,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0288688121131308e-05,
+      "loss": 1.1222,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5422577641452594,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0269574499337016e-05,
+      "loss": 1.1935,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5428054992605575,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0250459891936831e-05,
+      "loss": 1.1486,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5433532343758558,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0231344368816694e-05,
+      "loss": 1.0891,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5439009694911541,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.0212227999865905e-05,
+      "loss": 1.1315,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5444487046064523,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0193110854976859e-05,
+      "loss": 1.1676,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5449964397217506,
+      "grad_norm": 1.25,
+      "learning_rate": 1.0173993004044777e-05,
+      "loss": 1.1696,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5455441748370488,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0154874516967466e-05,
+      "loss": 1.1742,
+      "step": 4980
+    },
+    {
+      "epoch": 0.546091909952347,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.0135755463645065e-05,
+      "loss": 1.1852,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5466396450676453,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.0116635913979778e-05,
+      "loss": 1.164,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5471873801829436,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0097515937875619e-05,
+      "loss": 1.1373,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5477351152982418,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0078395605238168e-05,
+      "loss": 1.1487,
+      "step": 5000
+    },
+    {
+      "epoch": 0.54828285041354,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0059274985974305e-05,
+      "loss": 1.1163,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5488305855288382,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.004015414999197e-05,
+      "loss": 1.1109,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5493783206441365,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.002103316719987e-05,
+      "loss": 1.1546,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5499260557594348,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.0001912107507273e-05,
+      "loss": 1.1904,
+      "step": 5020
+    },
+    {
+      "epoch": 0.550473790874733,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.982791040823714e-06,
+      "loss": 1.1877,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5510215259900312,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.963670037058764e-06,
+      "loss": 1.1568,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5515692611053294,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.944549166121753e-06,
+      "loss": 1.146,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5521169962206277,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.925428497921533e-06,
+      "loss": 1.1435,
+      "step": 5040
+    },
+    {
+      "epoch": 0.552664731335926,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.906308102366216e-06,
+      "loss": 1.1093,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5532124664512242,
+      "grad_norm": 1.25,
+      "learning_rate": 9.887188049362906e-06,
+      "loss": 1.1546,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5537602015665224,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.86806840881747e-06,
+      "loss": 1.1049,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5543079366818207,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.848949250634253e-06,
+      "loss": 1.1575,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5548556717971189,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.82983064471584e-06,
+      "loss": 1.0936,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5554034069124172,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.810712660962813e-06,
+      "loss": 1.1147,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5559511420277155,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.791595369273454e-06,
+      "loss": 1.1929,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5564988771430136,
+      "grad_norm": 1.25,
+      "learning_rate": 9.772478839543526e-06,
+      "loss": 1.1763,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5570466122583119,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.753363141666017e-06,
+      "loss": 1.1068,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5575943473736101,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.734248345530854e-06,
+      "loss": 1.1285,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5581420824889084,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.715134521024675e-06,
+      "loss": 1.1655,
+      "step": 5095
+    },
+    {
+      "epoch": 0.5586898176042066,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.696021738030575e-06,
+      "loss": 1.1562,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5592375527195048,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.676910066427825e-06,
+      "loss": 1.2125,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5597852878348031,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.657799576091646e-06,
+      "loss": 1.1052,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5603330229501013,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.638690336892936e-06,
+      "loss": 1.1403,
+      "step": 5115
+    },
+    {
+      "epoch": 0.5608807580653996,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.619582418698009e-06,
+      "loss": 1.115,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5614284931806978,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.600475891368365e-06,
+      "loss": 1.1525,
+      "step": 5125
+    },
+    {
+      "epoch": 0.561976228295996,
+      "grad_norm": 1.25,
+      "learning_rate": 9.58137082476041e-06,
+      "loss": 1.1344,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5625239634112943,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.562267288725205e-06,
+      "loss": 1.1938,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5630716985265926,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.543165353108232e-06,
+      "loss": 1.1521,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5636194336418908,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.524065087749097e-06,
+      "loss": 1.1257,
+      "step": 5145
+    },
+    {
+      "epoch": 0.564167168757189,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.50496656248132e-06,
+      "loss": 1.1702,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5647149038724872,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.485869847132055e-06,
+      "loss": 1.141,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5652626389877855,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.466775011521825e-06,
+      "loss": 1.1146,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5658103741030838,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.447682125464299e-06,
+      "loss": 1.1636,
+      "step": 5165
+    },
+    {
+      "epoch": 0.566358109218382,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.42859125876601e-06,
+      "loss": 1.1135,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5669058443336802,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.409502481226098e-06,
+      "loss": 1.1315,
+      "step": 5175
+    },
+    {
+      "epoch": 0.5674535794489785,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.39041586263608e-06,
+      "loss": 1.1671,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5680013145642767,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.37133147277958e-06,
+      "loss": 1.1187,
+      "step": 5185
+    },
+    {
+      "epoch": 0.568549049679575,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.35224938143206e-06,
+      "loss": 1.1596,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5690967847948732,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.333169658360588e-06,
+      "loss": 1.1323,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5696445199101714,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.31409237332357e-06,
+      "loss": 1.1726,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5701922550254697,
+      "grad_norm": 1.25,
+      "learning_rate": 9.2950175960705e-06,
+      "loss": 1.1918,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5707399901407679,
+      "grad_norm": 1.296875,
+      "learning_rate": 9.275945396341704e-06,
+      "loss": 1.1976,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5712877252560662,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.256875843868072e-06,
+      "loss": 1.1623,
+      "step": 5215
+    },
+    {
+      "epoch": 0.5718354603713645,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.237809008370836e-06,
+      "loss": 1.1411,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5723831954866626,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.218744959561285e-06,
+      "loss": 1.1414,
+      "step": 5225
+    },
+    {
+      "epoch": 0.5729309306019609,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.199683767140511e-06,
+      "loss": 1.124,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5734786657172591,
+      "grad_norm": 1.25,
+      "learning_rate": 9.18062550079917e-06,
+      "loss": 1.1288,
+      "step": 5235
+    },
+    {
+      "epoch": 0.5740264008325574,
+      "grad_norm": 1.25,
+      "learning_rate": 9.161570230217228e-06,
+      "loss": 1.179,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5745741359478557,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.14251802506368e-06,
+      "loss": 1.1691,
+      "step": 5245
+    },
+    {
+      "epoch": 0.5751218710631538,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.12346895499633e-06,
+      "loss": 1.1064,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5756696061784521,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.104423089661512e-06,
+      "loss": 1.154,
+      "step": 5255
+    },
+    {
+      "epoch": 0.5762173412937504,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.085380498693837e-06,
+      "loss": 1.1888,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5767650764090486,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.066341251715958e-06,
+      "loss": 1.1586,
+      "step": 5265
+    },
+    {
+      "epoch": 0.5773128115243469,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.047305418338295e-06,
+      "loss": 1.1831,
+      "step": 5270
+    },
+    {
+      "epoch": 0.577860546639645,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.028273068158782e-06,
+      "loss": 1.2184,
+      "step": 5275
+    },
+    {
+      "epoch": 0.5784082817549433,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.009244270762625e-06,
+      "loss": 1.1374,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5789560168702416,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.990219095722044e-06,
+      "loss": 1.1599,
+      "step": 5285
+    },
+    {
+      "epoch": 0.5795037519855398,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.971197612596006e-06,
+      "loss": 1.0858,
+      "step": 5290
+    },
+    {
+      "epoch": 0.580051487100838,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.952179890929982e-06,
+      "loss": 1.1553,
+      "step": 5295
+    },
+    {
+      "epoch": 0.5805992222161362,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.933166000255692e-06,
+      "loss": 1.1676,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5811469573314345,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.914156010090852e-06,
+      "loss": 1.1262,
+      "step": 5305
+    },
+    {
+      "epoch": 0.5816946924467328,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.895149989938915e-06,
+      "loss": 1.1667,
+      "step": 5310
+    },
+    {
+      "epoch": 0.582242427562031,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.876148009288813e-06,
+      "loss": 1.1792,
+      "step": 5315
+    },
+    {
+      "epoch": 0.5827901626773292,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.857150137614718e-06,
+      "loss": 1.1894,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5833378977926275,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.83815644437578e-06,
+      "loss": 1.148,
+      "step": 5325
+    },
+    {
+      "epoch": 0.5838856329079257,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.819166999015863e-06,
+      "loss": 1.1339,
+      "step": 5330
+    },
+    {
+      "epoch": 0.584433368023224,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.8001818709633e-06,
+      "loss": 1.0615,
+      "step": 5335
+    },
+    {
+      "epoch": 0.5849811031385223,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.781201129630655e-06,
+      "loss": 1.1195,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5855288382538204,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.762224844414433e-06,
+      "loss": 1.1996,
+      "step": 5345
+    },
+    {
+      "epoch": 0.5860765733691187,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.743253084694858e-06,
+      "loss": 1.1773,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5866243084844169,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.724285919835611e-06,
+      "loss": 1.1409,
+      "step": 5355
+    },
+    {
+      "epoch": 0.5871720435997152,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.70532341918356e-06,
+      "loss": 1.1911,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5877197787150135,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.686365652068536e-06,
+      "loss": 1.1553,
+      "step": 5365
+    },
+    {
+      "epoch": 0.5882675138303116,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.667412687803054e-06,
+      "loss": 1.1326,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5888152489456099,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.648464595682068e-06,
+      "loss": 1.1896,
+      "step": 5375
+    },
+    {
+      "epoch": 0.5893629840609081,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.629521444982726e-06,
+      "loss": 1.1159,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5899107191762064,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.610583304964105e-06,
+      "loss": 1.1625,
+      "step": 5385
+    },
+    {
+      "epoch": 0.5904584542915047,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.591650244866957e-06,
+      "loss": 1.1655,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5910061894068028,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.572722333913473e-06,
+      "loss": 1.1404,
+      "step": 5395
+    },
+    {
+      "epoch": 0.5915539245221011,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.553799641307003e-06,
+      "loss": 1.1366,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5921016596373994,
+      "grad_norm": 1.25,
+      "learning_rate": 8.534882236231837e-06,
+      "loss": 1.1413,
+      "step": 5405
+    },
+    {
+      "epoch": 0.5926493947526976,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.515970187852916e-06,
+      "loss": 1.1671,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5931971298679959,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.4970635653156e-06,
+      "loss": 1.1465,
+      "step": 5415
+    },
+    {
+      "epoch": 0.593744864983294,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.478162437745418e-06,
+      "loss": 1.1198,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5942926000985923,
+      "grad_norm": 1.25,
+      "learning_rate": 8.459266874247805e-06,
+      "loss": 1.1324,
+      "step": 5425
+    },
+    {
+      "epoch": 0.5948403352138906,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.440376943907847e-06,
+      "loss": 1.1416,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5953880703291888,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.42149271579004e-06,
+      "loss": 1.1581,
+      "step": 5435
+    },
+    {
+      "epoch": 0.5959358054444871,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.402614258938038e-06,
+      "loss": 1.1468,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5964835405597853,
+      "grad_norm": 1.25,
+      "learning_rate": 8.38374164237438e-06,
+      "loss": 1.1138,
+      "step": 5445
+    },
+    {
+      "epoch": 0.5970312756750835,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.36487493510026e-06,
+      "loss": 1.1697,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5975790107903818,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.346014206095272e-06,
+      "loss": 1.1659,
+      "step": 5455
+    },
+    {
+      "epoch": 0.59812674590568,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.32715952431714e-06,
+      "loss": 1.163,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5986744810209782,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.308310958701492e-06,
+      "loss": 1.1482,
+      "step": 5465
+    },
+    {
+      "epoch": 0.5992222161362765,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.289468578161581e-06,
+      "loss": 1.1804,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5997699512515747,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.270632451588053e-06,
+      "loss": 1.143,
+      "step": 5475
+    },
+    {
+      "epoch": 0.600317686366873,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.251802647848696e-06,
+      "loss": 1.1491,
+      "step": 5480
+    },
+    {
+      "epoch": 0.6008654214821713,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.232979235788167e-06,
+      "loss": 1.2016,
+      "step": 5485
+    },
+    {
+      "epoch": 0.6014131565974694,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.214162284227758e-06,
+      "loss": 1.1132,
+      "step": 5490
+    },
+    {
+      "epoch": 0.6019608917127677,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.195351861965151e-06,
+      "loss": 1.1075,
+      "step": 5495
+    },
+    {
+      "epoch": 0.6025086268280659,
+      "grad_norm": 1.25,
+      "learning_rate": 8.176548037774137e-06,
+      "loss": 1.152,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6030563619433642,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.157750880404402e-06,
+      "loss": 1.1574,
+      "step": 5505
+    },
+    {
+      "epoch": 0.6036040970586625,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.13896045858125e-06,
+      "loss": 1.1635,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6041518321739606,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.12017684100535e-06,
+      "loss": 1.1093,
+      "step": 5515
+    },
+    {
+      "epoch": 0.6046995672892589,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.101400096352508e-06,
+      "loss": 1.1926,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6052473024045572,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.082630293273394e-06,
+      "loss": 1.1289,
+      "step": 5525
+    },
+    {
+      "epoch": 0.6057950375198554,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.063867500393296e-06,
+      "loss": 1.1551,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6063427726351537,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.045111786311878e-06,
+      "loss": 1.1592,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6068905077504518,
+      "grad_norm": 1.25,
+      "learning_rate": 8.026363219602921e-06,
+      "loss": 1.2031,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6074382428657501,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.007621868814073e-06,
+      "loss": 1.2304,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6079859779810484,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.9888878024666e-06,
+      "loss": 1.1297,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6085337130963466,
+      "grad_norm": 1.25,
+      "learning_rate": 7.970161089055127e-06,
+      "loss": 1.1341,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6090814482116449,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.951441797047412e-06,
+      "loss": 1.0772,
+      "step": 5560
+    },
+    {
+      "epoch": 0.609629183326943,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.93272999488407e-06,
+      "loss": 1.141,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6101769184422413,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.914025750978324e-06,
+      "loss": 1.1046,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6107246535575396,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.895329133715779e-06,
+      "loss": 1.1575,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6112723886728378,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.876640211454148e-06,
+      "loss": 1.1169,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6118201237881361,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.857959052523005e-06,
+      "loss": 1.157,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6123678589034343,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.839285725223545e-06,
+      "loss": 1.1819,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6129155940187325,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.820620297828337e-06,
+      "loss": 1.16,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6134633291340308,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.801962838581051e-06,
+      "loss": 1.1279,
+      "step": 5600
+    },
+    {
+      "epoch": 0.614011064249329,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.783313415696231e-06,
+      "loss": 1.1946,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6145587993646273,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.76467209735905e-06,
+      "loss": 1.1686,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6151065344799255,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.74603895172503e-06,
+      "loss": 1.206,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6156542695952237,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.727414046919825e-06,
+      "loss": 1.1534,
+      "step": 5620
+    },
+    {
+      "epoch": 0.616202004710522,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.70879745103896e-06,
+      "loss": 1.1479,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6167497398258203,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.690189232147566e-06,
+      "loss": 1.2199,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6172974749411185,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.671589458280172e-06,
+      "loss": 1.1312,
+      "step": 5635
+    },
+    {
+      "epoch": 0.6178452100564167,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.65299819744041e-06,
+      "loss": 1.1763,
+      "step": 5640
+    },
+    {
+      "epoch": 0.618392945171715,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.634415517600789e-06,
+      "loss": 1.1124,
+      "step": 5645
+    },
+    {
+      "epoch": 0.6189406802870132,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.61584148670246e-06,
+      "loss": 1.1387,
+      "step": 5650
+    },
+    {
+      "epoch": 0.6194884154023115,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.59727617265493e-06,
+      "loss": 1.1751,
+      "step": 5655
+    },
+    {
+      "epoch": 0.6200361505176096,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.578719643335854e-06,
+      "loss": 1.1762,
+      "step": 5660
+    },
+    {
+      "epoch": 0.6205838856329079,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.560171966590762e-06,
+      "loss": 1.1825,
+      "step": 5665
+    },
+    {
+      "epoch": 0.6211316207482062,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.541633210232812e-06,
+      "loss": 1.161,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6216793558635044,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.523103442042556e-06,
+      "loss": 1.097,
+      "step": 5675
+    },
+    {
+      "epoch": 0.6222270909788027,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.504582729767687e-06,
+      "loss": 1.1358,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6227748260941008,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.486071141122774e-06,
+      "loss": 1.2019,
+      "step": 5685
+    },
+    {
+      "epoch": 0.6233225612093991,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.4675687437890375e-06,
+      "loss": 1.1657,
+      "step": 5690
+    },
+    {
+      "epoch": 0.6238702963246974,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.4490756054141e-06,
+      "loss": 1.1705,
+      "step": 5695
+    },
+    {
+      "epoch": 0.6244180314399956,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.430591793611715e-06,
+      "loss": 1.1294,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6249657665552939,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.412117375961554e-06,
+      "loss": 1.1325,
+      "step": 5705
+    },
+    {
+      "epoch": 0.625513501670592,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.393652420008923e-06,
+      "loss": 1.1276,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6260612367858903,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.375196993264555e-06,
+      "loss": 1.1718,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6266089719011886,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.356751163204332e-06,
+      "loss": 1.11,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6271567070164868,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.33831499726905e-06,
+      "loss": 1.1356,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6277044421317851,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.319888562864165e-06,
+      "loss": 1.1446,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6282521772470833,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.301471927359571e-06,
+      "loss": 1.1267,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6287999123623815,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.283065158089315e-06,
+      "loss": 1.1637,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6293476474776798,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.264668322351385e-06,
+      "loss": 1.1372,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6298953825929781,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.24628148740745e-06,
+      "loss": 1.1375,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6304431177082763,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.227904720482601e-06,
+      "loss": 1.117,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6309908528235745,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.20953808876514e-06,
+      "loss": 1.1562,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6315385879388727,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.191181659406297e-06,
+      "loss": 1.12,
+      "step": 5765
+    },
+    {
+      "epoch": 0.632086323054171,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.172835499520002e-06,
+      "loss": 1.1717,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6326340581694693,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.154499676182648e-06,
+      "loss": 1.1618,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6331817932847675,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.136174256432828e-06,
+      "loss": 1.1518,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6337295284000657,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.117859307271095e-06,
+      "loss": 1.1344,
+      "step": 5785
+    },
+    {
+      "epoch": 0.634277263515364,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.099554895659734e-06,
+      "loss": 1.1029,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6348249986306622,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.081261088522482e-06,
+      "loss": 1.1324,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6353727337459605,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.062977952744326e-06,
+      "loss": 1.0928,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6359204688612587,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.044705555171223e-06,
+      "loss": 1.1352,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6364682039765569,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.02644396260987e-06,
+      "loss": 1.1719,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6370159390918552,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.008193241827468e-06,
+      "loss": 1.1325,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6375636742071534,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.989953459551463e-06,
+      "loss": 1.1598,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6381114093224517,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.971724682469303e-06,
+      "loss": 1.1432,
+      "step": 5825
+    },
+    {
+      "epoch": 0.63865914443775,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.953506977228211e-06,
+      "loss": 1.1366,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6392068795530481,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.935300410434927e-06,
+      "loss": 1.1742,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6397546146683464,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.917105048655454e-06,
+      "loss": 1.2324,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6403023497836446,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.898920958414843e-06,
+      "loss": 1.1062,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6408500848989429,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.880748206196934e-06,
+      "loss": 1.1941,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6413978200142411,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.862586858444102e-06,
+      "loss": 1.1457,
+      "step": 5855
+    },
+    {
+      "epoch": 0.6419455551295393,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.84443698155703e-06,
+      "loss": 1.1622,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6424932902448376,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.826298641894473e-06,
+      "loss": 1.1634,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6430410253601359,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.808171905772986e-06,
+      "loss": 1.1232,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6435887604754341,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.790056839466715e-06,
+      "loss": 1.164,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6441364955907323,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.7719535092071296e-06,
+      "loss": 1.1156,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6446842307060305,
+      "grad_norm": 1.25,
+      "learning_rate": 6.753861981182793e-06,
+      "loss": 1.1748,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6452319658213288,
+      "grad_norm": 1.25,
+      "learning_rate": 6.735782321539125e-06,
+      "loss": 1.105,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6457797009366271,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.717714596378138e-06,
+      "loss": 1.1865,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6463274360519253,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.699658871758223e-06,
+      "loss": 1.1225,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6468751711672235,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.681615213693895e-06,
+      "loss": 1.0707,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6474229062825217,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.6635836881555374e-06,
+      "loss": 1.1432,
+      "step": 5910
+    },
+    {
+      "epoch": 0.64797064139782,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.645564361069193e-06,
+      "loss": 1.1497,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6485183765131183,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.627557298316296e-06,
+      "loss": 1.1894,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6490661116284165,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.609562565733438e-06,
+      "loss": 1.1338,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6496138467437147,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.591580229112135e-06,
+      "loss": 1.0914,
+      "step": 5930
+    },
+    {
+      "epoch": 0.650161581859013,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.573610354198587e-06,
+      "loss": 1.1414,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6507093169743112,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.555653006693417e-06,
+      "loss": 1.1661,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6512570520896095,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.537708252251454e-06,
+      "loss": 1.2106,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6518047872049078,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.519776156481492e-06,
+      "loss": 1.127,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6523525223202059,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.501856784946027e-06,
+      "loss": 1.1524,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6529002574355042,
+      "grad_norm": 1.25,
+      "learning_rate": 6.483950203161047e-06,
+      "loss": 1.1249,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6534479925508024,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.466056476595768e-06,
+      "loss": 1.1287,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6539957276661007,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.448175670672413e-06,
+      "loss": 1.1386,
+      "step": 5970
+    },
+    {
+      "epoch": 0.654543462781399,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.43030785076597e-06,
+      "loss": 1.21,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6550911978966971,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.412453082203933e-06,
+      "loss": 1.1321,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6556389330119954,
+      "grad_norm": 1.25,
+      "learning_rate": 6.394611430266087e-06,
+      "loss": 1.1684,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6561866681272936,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.376782960184266e-06,
+      "loss": 1.1898,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6567344032425919,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.358967737142096e-06,
+      "loss": 1.1738,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6572821383578902,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.34116582627478e-06,
+      "loss": 1.1621,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6578298734731883,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.323377292668848e-06,
+      "loss": 1.1685,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6583776085884866,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.305602201361915e-06,
+      "loss": 1.1973,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6589253437037849,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.2878406173424576e-06,
+      "loss": 1.0964,
+      "step": 6015
+    },
+    {
+      "epoch": 0.6594730788190831,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.270092605549564e-06,
+      "loss": 1.1605,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6600208139343813,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.25235823087269e-06,
+      "loss": 1.1705,
+      "step": 6025
+    },
+    {
+      "epoch": 0.6605685490496795,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.2346375581514525e-06,
+      "loss": 1.1656,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6611162841649778,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.216930652175358e-06,
+      "loss": 1.1893,
+      "step": 6035
+    },
+    {
+      "epoch": 0.6616640192802761,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.199237577683577e-06,
+      "loss": 1.1163,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6622117543955743,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.181558399364722e-06,
+      "loss": 1.1637,
+      "step": 6045
+    },
+    {
+      "epoch": 0.6627594895108725,
+      "grad_norm": 1.3359375,
+      "learning_rate": 6.163893181856586e-06,
+      "loss": 1.2075,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6633072246261708,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.146241989745932e-06,
+      "loss": 1.1517,
+      "step": 6055
+    },
+    {
+      "epoch": 0.663854959741469,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.128604887568237e-06,
+      "loss": 1.113,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6644026948567673,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.1109819398074564e-06,
+      "loss": 1.1239,
+      "step": 6065
+    },
+    {
+      "epoch": 0.6649504299720655,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.09337321089581e-06,
+      "loss": 1.1289,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6654981650873637,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.0757787652135195e-06,
+      "loss": 1.1207,
+      "step": 6075
+    },
+    {
+      "epoch": 0.666045900202662,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.05819866708859e-06,
+      "loss": 1.1825,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6665936353179602,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.040632980796566e-06,
+      "loss": 1.1496,
+      "step": 6085
+    },
+    {
+      "epoch": 0.6671413704332585,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.023081770560307e-06,
+      "loss": 1.1455,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6676891055485568,
+      "grad_norm": 1.25,
+      "learning_rate": 6.005545100549739e-06,
+      "loss": 1.1437,
+      "step": 6095
+    },
+    {
+      "epoch": 0.6682368406638549,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.9880230348816245e-06,
+      "loss": 1.1507,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6687845757791532,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.970515637619346e-06,
+      "loss": 1.096,
+      "step": 6105
+    },
+    {
+      "epoch": 0.6693323108944514,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.953022972772633e-06,
+      "loss": 1.1749,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6698800460097497,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.935545104297373e-06,
+      "loss": 1.1504,
+      "step": 6115
+    },
+    {
+      "epoch": 0.670427781125048,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.918082096095339e-06,
+      "loss": 1.1512,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6709755162403461,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.900634012013977e-06,
+      "loss": 1.1532,
+      "step": 6125
+    },
+    {
+      "epoch": 0.6715232513556444,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.883200915846181e-06,
+      "loss": 1.0785,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6720709864709427,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.865782871330026e-06,
+      "loss": 1.1757,
+      "step": 6135
+    },
+    {
+      "epoch": 0.6726187215862409,
+      "grad_norm": 1.2734375,
+      "learning_rate": 5.848379942148568e-06,
+      "loss": 1.2096,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6731664567015392,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.830992191929603e-06,
+      "loss": 1.1297,
+      "step": 6145
+    },
+    {
+      "epoch": 0.6737141918168373,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.813619684245413e-06,
+      "loss": 1.1174,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6742619269321356,
+      "grad_norm": 1.25,
+      "learning_rate": 5.7962624826125725e-06,
+      "loss": 1.1527,
+      "step": 6155
+    },
+    {
+      "epoch": 0.6748096620474339,
+      "grad_norm": 1.28125,
+      "learning_rate": 5.7789206504916815e-06,
+      "loss": 1.1952,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6753573971627321,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.761594251287149e-06,
+      "loss": 1.1212,
+      "step": 6165
+    },
+    {
+      "epoch": 0.6759051322780304,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.74428334834696e-06,
+      "loss": 1.142,
+      "step": 6170
+    },
+    {
+      "epoch": 0.6764528673933285,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.726988004962443e-06,
+      "loss": 1.137,
+      "step": 6175
+    },
+    {
+      "epoch": 0.6770006025086268,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.709708284368038e-06,
+      "loss": 1.1304,
+      "step": 6180
+    },
+    {
+      "epoch": 0.6775483376239251,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.692444249741064e-06,
+      "loss": 1.101,
+      "step": 6185
+    },
+    {
+      "epoch": 0.6780960727392233,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.675195964201492e-06,
+      "loss": 1.1385,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6786438078545216,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.65796349081171e-06,
+      "loss": 1.1179,
+      "step": 6195
+    },
+    {
+      "epoch": 0.6791915429698198,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.640746892576296e-06,
+      "loss": 1.1782,
+      "step": 6200
+    },
+    {
+      "epoch": 0.679739278085118,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.623546232441786e-06,
+      "loss": 1.1495,
+      "step": 6205
+    },
+    {
+      "epoch": 0.6802870132004163,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.6063615732964425e-06,
+      "loss": 1.1382,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6808347483157146,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.589192977970028e-06,
+      "loss": 1.0981,
+      "step": 6215
+    },
+    {
+      "epoch": 0.6813824834310127,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.57204050923357e-06,
+      "loss": 1.1724,
+      "step": 6220
+    },
+    {
+      "epoch": 0.681930218546311,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.5549042297991366e-06,
+      "loss": 1.1736,
+      "step": 6225
+    },
+    {
+      "epoch": 0.6824779536616092,
+      "grad_norm": 1.25,
+      "learning_rate": 5.537784202319607e-06,
+      "loss": 1.1174,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6830256887769075,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.520680489388437e-06,
+      "loss": 1.1457,
+      "step": 6235
+    },
+    {
+      "epoch": 0.6835734238922058,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.503593153539437e-06,
+      "loss": 1.1336,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6841211590075039,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5.486522257246538e-06,
+      "loss": 1.2187,
+      "step": 6245
+    },
+    {
+      "epoch": 0.6846688941228022,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5.4694678629235696e-06,
+      "loss": 1.1642,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6852166292381004,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.452430032924017e-06,
+      "loss": 1.1529,
+      "step": 6255
+    },
+    {
+      "epoch": 0.6857643643533987,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.4354088295408265e-06,
+      "loss": 1.1982,
+      "step": 6260
+    },
+    {
+      "epoch": 0.686312099468697,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5.418404315006125e-06,
+      "loss": 1.0957,
+      "step": 6265
+    },
+    {
+      "epoch": 0.6868598345839951,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.401416551491039e-06,
+      "loss": 1.1719,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6874075696992934,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.384445601105462e-06,
+      "loss": 1.0921,
+      "step": 6275
+    },
+    {
+      "epoch": 0.6879553048145917,
+      "grad_norm": 1.25,
+      "learning_rate": 5.367491525897787e-06,
+      "loss": 1.1153,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6885030399298899,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.3505543878547315e-06,
+      "loss": 1.1524,
+      "step": 6285
+    },
+    {
+      "epoch": 0.6890507750451882,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.33363424890108e-06,
+      "loss": 1.1685,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6895985101604863,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.316731170899467e-06,
+      "loss": 1.1593,
+      "step": 6295
+    },
+    {
+      "epoch": 0.6901462452757846,
+      "grad_norm": 1.328125,
+      "learning_rate": 5.299845215650157e-06,
+      "loss": 1.1513,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6906939803910829,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.282976444890793e-06,
+      "loss": 1.1787,
+      "step": 6305
+    },
+    {
+      "epoch": 0.6912417155063811,
+      "grad_norm": 1.2734375,
+      "learning_rate": 5.2661249202962e-06,
+      "loss": 1.125,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6917894506216794,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.249290703478158e-06,
+      "loss": 1.1662,
+      "step": 6315
+    },
+    {
+      "epoch": 0.6923371857369776,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.232473855985149e-06,
+      "loss": 1.1612,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6928849208522758,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.215674439302155e-06,
+      "loss": 1.1552,
+      "step": 6325
+    },
+    {
+      "epoch": 0.6934326559675741,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.198892514850444e-06,
+      "loss": 1.1097,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6939803910828723,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.182128143987302e-06,
+      "loss": 1.1337,
+      "step": 6335
+    },
+    {
+      "epoch": 0.6945281261981706,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.165381388005861e-06,
+      "loss": 1.148,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6950758613134688,
+      "grad_norm": 1.34375,
+      "learning_rate": 5.148652308134844e-06,
+      "loss": 1.2329,
+      "step": 6345
+    },
+    {
+      "epoch": 0.695623596428767,
+      "grad_norm": 1.328125,
+      "learning_rate": 5.131940965538329e-06,
+      "loss": 1.1538,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6961713315440653,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.115247421315572e-06,
+      "loss": 1.2042,
+      "step": 6355
+    },
+    {
+      "epoch": 0.6967190666593636,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.098571736500744e-06,
+      "loss": 1.1985,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6972668017746618,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.081913972062704e-06,
+      "loss": 1.1183,
+      "step": 6365
+    },
+    {
+      "epoch": 0.69781453688996,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.0652741889048205e-06,
+      "loss": 1.179,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6983622720052582,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.048652447864689e-06,
+      "loss": 1.1346,
+      "step": 6375
+    },
+    {
+      "epoch": 0.6989100071205565,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.032048809713967e-06,
+      "loss": 1.1022,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6994577422358548,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.015463335158114e-06,
+      "loss": 1.179,
+      "step": 6385
+    },
+    {
+      "epoch": 0.7000054773511529,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.998896084836167e-06,
+      "loss": 1.1692,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7005532124664512,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.982347119320557e-06,
+      "loss": 1.1716,
+      "step": 6395
+    },
+    {
+      "epoch": 0.7011009475817495,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.965816499116849e-06,
+      "loss": 1.1171,
+      "step": 6400
+    },
+    {
+      "epoch": 0.7016486826970477,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.949304284663535e-06,
+      "loss": 1.1577,
+      "step": 6405
+    },
+    {
+      "epoch": 0.702196417812346,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.932810536331817e-06,
+      "loss": 1.1562,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7027441529276441,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.916335314425376e-06,
+      "loss": 1.1456,
+      "step": 6415
+    },
+    {
+      "epoch": 0.7032918880429424,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.899878679180167e-06,
+      "loss": 1.1799,
+      "step": 6420
+    },
+    {
+      "epoch": 0.7038396231582407,
+      "grad_norm": 1.25,
+      "learning_rate": 4.8834406907641784e-06,
+      "loss": 1.1339,
+      "step": 6425
+    },
+    {
+      "epoch": 0.7043873582735389,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.8670214092772316e-06,
+      "loss": 1.1551,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7049350933888372,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.850620894750746e-06,
+      "loss": 1.1504,
+      "step": 6435
+    },
+    {
+      "epoch": 0.7054828285041354,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.834239207147532e-06,
+      "loss": 1.2009,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7060305636194336,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.817876406361561e-06,
+      "loss": 1.1192,
+      "step": 6445
+    },
+    {
+      "epoch": 0.7065782987347319,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.801532552217756e-06,
+      "loss": 1.1502,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7071260338500301,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.785207704471763e-06,
+      "loss": 1.142,
+      "step": 6455
+    },
+    {
+      "epoch": 0.7076737689653284,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.76890192280974e-06,
+      "loss": 1.1281,
+      "step": 6460
+    },
+    {
+      "epoch": 0.7082215040806266,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.7526152668481385e-06,
+      "loss": 1.1053,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7087692391959248,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.736347796133481e-06,
+      "loss": 1.163,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7093169743112231,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.7200995701421455e-06,
+      "loss": 1.2034,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7098647094265214,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.703870648280151e-06,
+      "loss": 1.1658,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7104124445418196,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.687661089882934e-06,
+      "loss": 1.1676,
+      "step": 6485
+    },
+    {
+      "epoch": 0.7109601796571178,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.671470954215139e-06,
+      "loss": 1.1919,
+      "step": 6490
+    },
+    {
+      "epoch": 0.711507914772416,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.655300300470395e-06,
+      "loss": 1.1288,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7120556498877143,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.639149187771102e-06,
+      "loss": 1.1863,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7126033850030126,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.6230176751682185e-06,
+      "loss": 1.1942,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7131511201183108,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.606905821641036e-06,
+      "loss": 1.1683,
+      "step": 6510
+    },
+    {
+      "epoch": 0.713698855233609,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.590813686096981e-06,
+      "loss": 1.0678,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7142465903489073,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.5747413273713715e-06,
+      "loss": 1.1841,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7147943254642055,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.558688804227229e-06,
+      "loss": 1.1208,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7153420605795038,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.542656175355054e-06,
+      "loss": 1.1631,
+      "step": 6530
+    },
+    {
+      "epoch": 0.715889795694802,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.526643499372604e-06,
+      "loss": 1.1604,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7164375308101002,
+      "grad_norm": 1.328125,
+      "learning_rate": 4.510650834824692e-06,
+      "loss": 1.1719,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7169852659253985,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.49467824018296e-06,
+      "loss": 1.1601,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7175330010406967,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.4787257738456745e-06,
+      "loss": 1.1557,
+      "step": 6550
+    },
+    {
+      "epoch": 0.718080736155995,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.4627934941375185e-06,
+      "loss": 1.102,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7186284712712933,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.446881459309351e-06,
+      "loss": 1.0873,
+      "step": 6560
+    },
+    {
+      "epoch": 0.7191762063865914,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.430989727538023e-06,
+      "loss": 1.1641,
+      "step": 6565
+    },
+    {
+      "epoch": 0.7197239415018897,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.415118356926163e-06,
+      "loss": 1.1565,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7202716766171879,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.399267405501938e-06,
+      "loss": 1.174,
+      "step": 6575
+    },
+    {
+      "epoch": 0.7208194117324862,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.383436931218867e-06,
+      "loss": 1.1657,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7213671468477844,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.367626991955615e-06,
+      "loss": 1.1566,
+      "step": 6585
+    },
+    {
+      "epoch": 0.7219148819630826,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.351837645515739e-06,
+      "loss": 1.1382,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7224626170783809,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.336068949627534e-06,
+      "loss": 1.1729,
+      "step": 6595
+    },
+    {
+      "epoch": 0.7230103521936792,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.320320961943781e-06,
+      "loss": 1.1313,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7235580873089774,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.304593740041539e-06,
+      "loss": 1.1615,
+      "step": 6605
+    },
+    {
+      "epoch": 0.7241058224242756,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.2888873414219675e-06,
+      "loss": 1.1193,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7246535575395738,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.273201823510066e-06,
+      "loss": 1.1548,
+      "step": 6615
+    },
+    {
+      "epoch": 0.7252012926548721,
+      "grad_norm": 1.25,
+      "learning_rate": 4.2575372436545125e-06,
+      "loss": 1.1503,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7257490277701704,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.241893659127428e-06,
+      "loss": 1.0939,
+      "step": 6625
+    },
+    {
+      "epoch": 0.7262967628854686,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.226271127124153e-06,
+      "loss": 1.1613,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7268444980007668,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.210669704763084e-06,
+      "loss": 1.1476,
+      "step": 6635
+    },
+    {
+      "epoch": 0.727392233116065,
+      "grad_norm": 1.34375,
+      "learning_rate": 4.195089449085424e-06,
+      "loss": 1.1295,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7279399682313633,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.1795304170549784e-06,
+      "loss": 1.1707,
+      "step": 6645
+    },
+    {
+      "epoch": 0.7284877033466616,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.163992665557975e-06,
+      "loss": 1.1444,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7290354384619598,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.148476251402821e-06,
+      "loss": 1.2146,
+      "step": 6655
+    },
+    {
+      "epoch": 0.729583173577258,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.1329812313199195e-06,
+      "loss": 1.1645,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7301309086925563,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.117507661961451e-06,
+      "loss": 1.1362,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7306786438078545,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.102055599901168e-06,
+      "loss": 1.1424,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7312263789231528,
+      "grad_norm": 1.25,
+      "learning_rate": 4.08662510163419e-06,
+      "loss": 1.1322,
+      "step": 6675
+    },
+    {
+      "epoch": 0.731774114038451,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.071216223576795e-06,
+      "loss": 1.1155,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7323218491537492,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.055829022066216e-06,
+      "loss": 1.126,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7328695842690475,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.040463553360431e-06,
+      "loss": 1.1659,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7334173193843457,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.025119873637962e-06,
+      "loss": 1.1248,
+      "step": 6695
+    },
+    {
+      "epoch": 0.733965054499644,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.009798038997664e-06,
+      "loss": 1.1073,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7345127896149423,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.9944981054585276e-06,
+      "loss": 1.1886,
+      "step": 6705
+    },
+    {
+      "epoch": 0.7350605247302404,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.979220128959463e-06,
+      "loss": 1.1453,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7356082598455387,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.96396416535911e-06,
+      "loss": 1.1076,
+      "step": 6715
+    },
+    {
+      "epoch": 0.736155994960837,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.948730270435618e-06,
+      "loss": 1.1978,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7367037300761352,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.933518499886455e-06,
+      "loss": 1.1705,
+      "step": 6725
+    },
+    {
+      "epoch": 0.7372514651914335,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.918328909328198e-06,
+      "loss": 1.1433,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7377992003067316,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.9031615542963305e-06,
+      "loss": 1.1205,
+      "step": 6735
+    },
+    {
+      "epoch": 0.7383469354220299,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.8880164902450375e-06,
+      "loss": 1.2021,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7388946705373282,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.8728937725470084e-06,
+      "loss": 1.1087,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7394424056526264,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.857793456493226e-06,
+      "loss": 1.1444,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7399901407679247,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.842715597292773e-06,
+      "loss": 1.1848,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7405378758832228,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.8276602500726265e-06,
+      "loss": 1.1686,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7410856109985211,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.8126274698774513e-06,
+      "loss": 1.1225,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7416333461138194,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.797617311669417e-06,
+      "loss": 1.2015,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7421810812291176,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.7826298303279652e-06,
+      "loss": 1.1643,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7427288163444158,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.7676650806496386e-06,
+      "loss": 1.1988,
+      "step": 6780
+    },
+    {
+      "epoch": 0.743276551459714,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.7527231173478684e-06,
+      "loss": 1.1904,
+      "step": 6785
+    },
+    {
+      "epoch": 0.7438242865750123,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.7378039950527733e-06,
+      "loss": 1.1832,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7443720216903106,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.722907768310965e-06,
+      "loss": 1.1459,
+      "step": 6795
+    },
+    {
+      "epoch": 0.7449197568056088,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.7080344915853406e-06,
+      "loss": 1.1343,
+      "step": 6800
+    },
+    {
+      "epoch": 0.745467491920907,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.6931842192548884e-06,
+      "loss": 1.161,
+      "step": 6805
+    },
+    {
+      "epoch": 0.7460152270362053,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.6783570056145025e-06,
+      "loss": 1.089,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7465629621515035,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.6635529048747498e-06,
+      "loss": 1.1435,
+      "step": 6815
+    },
+    {
+      "epoch": 0.7471106972668018,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.648771971161702e-06,
+      "loss": 1.1859,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7476584323821001,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.6340142585167415e-06,
+      "loss": 1.1555,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7482061674973982,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.6192798208963255e-06,
+      "loss": 1.1323,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7487539026126965,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.6045687121718365e-06,
+      "loss": 1.1747,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7493016377279947,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.589880986129356e-06,
+      "loss": 1.1286,
+      "step": 6840
+    },
+    {
+      "epoch": 0.749849372843293,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.5752166964694613e-06,
+      "loss": 1.1288,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7503971079585913,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.5605758968070658e-06,
+      "loss": 1.2077,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7509448430738894,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.5459586406711875e-06,
+      "loss": 1.0527,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7514925781891877,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.531364981504758e-06,
+      "loss": 1.1637,
+      "step": 6860
+    },
+    {
+      "epoch": 0.752040313304486,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.5167949726644545e-06,
+      "loss": 1.1552,
+      "step": 6865
+    },
+    {
+      "epoch": 0.7525880484197842,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.5022486674204626e-06,
+      "loss": 1.1109,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7531357835350825,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.4877261189563237e-06,
+      "loss": 1.1444,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7536835186503806,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.4732273803687133e-06,
+      "loss": 1.1577,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7542312537656789,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.4587525046672433e-06,
+      "loss": 1.1359,
+      "step": 6885
+    },
+    {
+      "epoch": 0.7547789888809772,
+      "grad_norm": 1.25,
+      "learning_rate": 3.4443015447742977e-06,
+      "loss": 1.1344,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7553267239962754,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.4298745535248135e-06,
+      "loss": 1.1513,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7558744591115737,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.4154715836660814e-06,
+      "loss": 1.1569,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7564221942268718,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.40109268785759e-06,
+      "loss": 1.1421,
+      "step": 6905
+    },
+    {
+      "epoch": 0.7569699293421701,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.3867379186707927e-06,
+      "loss": 1.1312,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7575176644574684,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.372407328588938e-06,
+      "loss": 1.1694,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7580653995727666,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.35810097000687e-06,
+      "loss": 1.1709,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7586131346880649,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.3438188952308403e-06,
+      "loss": 1.15,
+      "step": 6925
+    },
+    {
+      "epoch": 0.7591608698033631,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.3295611564783136e-06,
+      "loss": 1.0951,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7597086049186613,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.315327805877784e-06,
+      "loss": 1.2345,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7602563400339596,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.3011188954685626e-06,
+      "loss": 1.1564,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7608040751492579,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.286934477200624e-06,
+      "loss": 1.1598,
+      "step": 6945
+    },
+    {
+      "epoch": 0.761351810264556,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.2727746029343856e-06,
+      "loss": 1.1203,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7618995453798543,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.258639324440527e-06,
+      "loss": 1.1721,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7624472804951525,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.2445286933998044e-06,
+      "loss": 1.1528,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7629950156104508,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.23044276140286e-06,
+      "loss": 1.124,
+      "step": 6965
+    },
+    {
+      "epoch": 0.7635427507257491,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.2163815799500288e-06,
+      "loss": 1.1527,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7640904858410472,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.202345200451158e-06,
+      "loss": 1.103,
+      "step": 6975
+    },
+    {
+      "epoch": 0.7646382209563455,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.1883336742254124e-06,
+      "loss": 1.16,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7651859560716437,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.174347052501091e-06,
+      "loss": 1.1707,
+      "step": 6985
+    },
+    {
+      "epoch": 0.765733691186942,
+      "grad_norm": 1.25,
+      "learning_rate": 3.160385386415438e-06,
+      "loss": 1.1882,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7662814263022403,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.146448727014454e-06,
+      "loss": 1.1877,
+      "step": 6995
+    },
+    {
+      "epoch": 0.7668291614175384,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.1325371252527127e-06,
+      "loss": 1.1277,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7673768965328367,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.1186506319931754e-06,
+      "loss": 1.1748,
+      "step": 7005
+    },
+    {
+      "epoch": 0.767924631648135,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.1047892980069984e-06,
+      "loss": 1.2315,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7684723667634332,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.0909531739733544e-06,
+      "loss": 1.2003,
+      "step": 7015
+    },
+    {
+      "epoch": 0.7690201018787315,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.0771423104792454e-06,
+      "loss": 1.1535,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7695678369940296,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.0633567580193147e-06,
+      "loss": 1.1765,
+      "step": 7025
+    },
+    {
+      "epoch": 0.7701155721093279,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.049596566995666e-06,
+      "loss": 1.1403,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7706633072246262,
+      "grad_norm": 1.25,
+      "learning_rate": 3.0358617877176787e-06,
+      "loss": 1.1534,
+      "step": 7035
+    },
+    {
+      "epoch": 0.7712110423399244,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.0221524704018212e-06,
+      "loss": 1.1534,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7717587774552227,
+      "grad_norm": 1.25,
+      "learning_rate": 3.0084686651714647e-06,
+      "loss": 1.1507,
+      "step": 7045
+    },
+    {
+      "epoch": 0.7723065125705209,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.99481042205672e-06,
+      "loss": 1.1906,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7728542476858191,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.9811777909942185e-06,
+      "loss": 1.1704,
+      "step": 7055
+    },
+    {
+      "epoch": 0.7734019828011174,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.9675708218269596e-06,
+      "loss": 1.1865,
+      "step": 7060
+    },
+    {
+      "epoch": 0.7739497179164156,
+      "grad_norm": 1.25,
+      "learning_rate": 2.9539895643041283e-06,
+      "loss": 1.1695,
+      "step": 7065
+    },
+    {
+      "epoch": 0.7744974530317139,
+      "grad_norm": 1.25,
+      "learning_rate": 2.940434068080883e-06,
+      "loss": 1.1652,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7750451881470121,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.9269043827182084e-06,
+      "loss": 1.1423,
+      "step": 7075
+    },
+    {
+      "epoch": 0.7755929232623103,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.9134005576827273e-06,
+      "loss": 1.1913,
+      "step": 7080
+    },
+    {
+      "epoch": 0.7761406583776086,
+      "grad_norm": 1.25,
+      "learning_rate": 2.899922642346491e-06,
+      "loss": 1.1596,
+      "step": 7085
+    },
+    {
+      "epoch": 0.7766883934929069,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.8864706859868443e-06,
+      "loss": 1.1581,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7772361286082051,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.8730447377862116e-06,
+      "loss": 1.1263,
+      "step": 7095
+    },
+    {
+      "epoch": 0.7777838637235033,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.859644846831918e-06,
+      "loss": 1.1653,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7783315988388015,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.8462710621160427e-06,
+      "loss": 1.1609,
+      "step": 7105
+    },
+    {
+      "epoch": 0.7788793339540998,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.8329234325351917e-06,
+      "loss": 1.0972,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7794270690693981,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.8196020068903574e-06,
+      "loss": 1.1194,
+      "step": 7115
+    },
+    {
+      "epoch": 0.7799748041846963,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.8063068338867315e-06,
+      "loss": 1.17,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7805225392999945,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.7930379621335035e-06,
+      "loss": 1.1294,
+      "step": 7125
+    },
+    {
+      "epoch": 0.7810702744152928,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.779795440143721e-06,
+      "loss": 1.1411,
+      "step": 7130
+    },
+    {
+      "epoch": 0.781618009530591,
+      "grad_norm": 1.2890625,
+      "learning_rate": 2.766579316334085e-06,
+      "loss": 1.2037,
+      "step": 7135
+    },
+    {
+      "epoch": 0.7821657446458893,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.7533896390247706e-06,
+      "loss": 1.1298,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7827134797611874,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.7402264564392788e-06,
+      "loss": 1.1657,
+      "step": 7145
+    },
+    {
+      "epoch": 0.7832612148764857,
+      "grad_norm": 1.25,
+      "learning_rate": 2.7270898167042325e-06,
+      "loss": 1.1313,
+      "step": 7150
+    },
+    {
+      "epoch": 0.783808949991784,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.7139797678492032e-06,
+      "loss": 1.1042,
+      "step": 7155
+    },
+    {
+      "epoch": 0.7843566851070822,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.700896357806555e-06,
+      "loss": 1.2117,
+      "step": 7160
+    },
+    {
+      "epoch": 0.7849044202223805,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.6878396344112513e-06,
+      "loss": 1.1121,
+      "step": 7165
+    },
+    {
+      "epoch": 0.7854521553376786,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.6748096454006812e-06,
+      "loss": 1.1624,
+      "step": 7170
+    },
+    {
+      "epoch": 0.7859998904529769,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.6618064384144925e-06,
+      "loss": 1.0864,
+      "step": 7175
+    },
+    {
+      "epoch": 0.7865476255682752,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.648830060994414e-06,
+      "loss": 1.1666,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7870953606835734,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.63588056058408e-06,
+      "loss": 1.0627,
+      "step": 7185
+    },
+    {
+      "epoch": 0.7876430957988717,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.622957984528861e-06,
+      "loss": 1.1774,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7881908309141699,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.6100623800756787e-06,
+      "loss": 1.117,
+      "step": 7195
+    },
+    {
+      "epoch": 0.7887385660294681,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.5971937943728563e-06,
+      "loss": 1.1719,
+      "step": 7200
+    },
+    {
+      "epoch": 0.7892863011447664,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.5843522744699246e-06,
+      "loss": 1.1124,
+      "step": 7205
+    },
+    {
+      "epoch": 0.7898340362600647,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.5715378673174575e-06,
+      "loss": 1.1658,
+      "step": 7210
+    },
+    {
+      "epoch": 0.7903817713753629,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.558750619766902e-06,
+      "loss": 1.1308,
+      "step": 7215
+    },
+    {
+      "epoch": 0.7909295064906611,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.545990578570404e-06,
+      "loss": 1.1377,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7914772416059593,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.533257790380642e-06,
+      "loss": 1.1657,
+      "step": 7225
+    },
+    {
+      "epoch": 0.7920249767212576,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.5205523017506496e-06,
+      "loss": 1.1398,
+      "step": 7230
+    },
+    {
+      "epoch": 0.7925727118365559,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.5078741591336496e-06,
+      "loss": 1.193,
+      "step": 7235
+    },
+    {
+      "epoch": 0.7931204469518541,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.495223408882886e-06,
+      "loss": 1.168,
+      "step": 7240
+    },
+    {
+      "epoch": 0.7936681820671523,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.4826000972514498e-06,
+      "loss": 1.1467,
+      "step": 7245
+    },
+    {
+      "epoch": 0.7942159171824505,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.4700042703921132e-06,
+      "loss": 1.154,
+      "step": 7250
+    },
+    {
+      "epoch": 0.7947636522977488,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.4574359743571606e-06,
+      "loss": 1.1555,
+      "step": 7255
+    },
+    {
+      "epoch": 0.7953113874130471,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.4448952550982142e-06,
+      "loss": 1.1517,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7958591225283453,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.4323821584660846e-06,
+      "loss": 1.1558,
+      "step": 7265
+    },
+    {
+      "epoch": 0.7964068576436435,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.4198967302105712e-06,
+      "loss": 1.106,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7969545927589418,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.4074390159803273e-06,
+      "loss": 1.17,
+      "step": 7275
+    },
+    {
+      "epoch": 0.79750232787424,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.395009061322675e-06,
+      "loss": 1.1446,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7980500629895383,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.3826069116834426e-06,
+      "loss": 1.1757,
+      "step": 7285
+    },
+    {
+      "epoch": 0.7985977981048366,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.370232612406801e-06,
+      "loss": 1.1398,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7991455332201347,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.3578862087350927e-06,
+      "loss": 1.1763,
+      "step": 7295
+    },
+    {
+      "epoch": 0.799693268335433,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.3455677458086702e-06,
+      "loss": 1.1386,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8002410034507312,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.3332772686657402e-06,
+      "loss": 1.1302,
+      "step": 7305
+    },
+    {
+      "epoch": 0.8007887385660295,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.3210148222421715e-06,
+      "loss": 1.1224,
+      "step": 7310
+    },
+    {
+      "epoch": 0.8013364736813277,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.3087804513713563e-06,
+      "loss": 1.1453,
+      "step": 7315
+    },
+    {
+      "epoch": 0.8018842087966259,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.29657420078405e-06,
+      "loss": 1.1204,
+      "step": 7320
+    },
+    {
+      "epoch": 0.8024319439119242,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.284396115108174e-06,
+      "loss": 1.1726,
+      "step": 7325
+    },
+    {
+      "epoch": 0.8029796790272224,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.272246238868687e-06,
+      "loss": 1.1466,
+      "step": 7330
+    },
+    {
+      "epoch": 0.8035274141425207,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.2601246164874168e-06,
+      "loss": 1.1598,
+      "step": 7335
+    },
+    {
+      "epoch": 0.8040751492578189,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.2480312922828717e-06,
+      "loss": 1.1559,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8046228843731171,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.2359663104701133e-06,
+      "loss": 1.1813,
+      "step": 7345
+    },
+    {
+      "epoch": 0.8051706194884154,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.2239297151605774e-06,
+      "loss": 1.1431,
+      "step": 7350
+    },
+    {
+      "epoch": 0.8057183546037137,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.2119215503618995e-06,
+      "loss": 1.1598,
+      "step": 7355
+    },
+    {
+      "epoch": 0.8062660897190119,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.1999418599777935e-06,
+      "loss": 1.1539,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8068138248343101,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.1879906878078427e-06,
+      "loss": 1.1623,
+      "step": 7365
+    },
+    {
+      "epoch": 0.8073615599496083,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.1760680775473742e-06,
+      "loss": 1.108,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8079092950649066,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.1641740727872963e-06,
+      "loss": 1.1486,
+      "step": 7375
+    },
+    {
+      "epoch": 0.8084570301802049,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.152308717013911e-06,
+      "loss": 1.1697,
+      "step": 7380
+    },
+    {
+      "epoch": 0.8090047652955031,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.1404720536087954e-06,
+      "loss": 1.1243,
+      "step": 7385
+    },
+    {
+      "epoch": 0.8095525004108013,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.128664125848615e-06,
+      "loss": 1.1442,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8101002355260996,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.116884976904965e-06,
+      "loss": 1.0855,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8106479706413978,
+      "grad_norm": 1.25,
+      "learning_rate": 2.105134649844235e-06,
+      "loss": 1.1801,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8111957057566961,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.093413187627431e-06,
+      "loss": 1.0925,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8117434408719943,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.081720633110017e-06,
+      "loss": 1.1548,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8122911759872925,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.070057029041783e-06,
+      "loss": 1.1049,
+      "step": 7415
+    },
+    {
+      "epoch": 0.8128389111025908,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.0584224180666567e-06,
+      "loss": 1.184,
+      "step": 7420
+    },
+    {
+      "epoch": 0.813386646217889,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.046816842722571e-06,
+      "loss": 1.1299,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8139343813331873,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.0352403454412984e-06,
+      "loss": 1.1397,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8144821164484856,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.0236929685482875e-06,
+      "loss": 1.1364,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8150298515637837,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.012174754262537e-06,
+      "loss": 1.1209,
+      "step": 7440
+    },
+    {
+      "epoch": 0.815577586679082,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.000685744696409e-06,
+      "loss": 1.1272,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8161253217943802,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9892259818554915e-06,
+      "loss": 1.2119,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8166730569096785,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9777955076384446e-06,
+      "loss": 1.1568,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8172207920249768,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.966394363836842e-06,
+      "loss": 1.1831,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8177685271402749,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.955022592135022e-06,
+      "loss": 1.1637,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8183162622555732,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9436802341099337e-06,
+      "loss": 1.1363,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8188639973708715,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9323673312309856e-06,
+      "loss": 1.1293,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8194117324861697,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9210839248598924e-06,
+      "loss": 1.1667,
+      "step": 7480
+    },
+    {
+      "epoch": 0.819959467601468,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 1.2437,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8205072027167661,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8986057665487656e-06,
+      "loss": 1.1136,
+      "step": 7490
+    },
+    {
+      "epoch": 0.8210549378320644,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.887411096792342e-06,
+      "loss": 1.1329,
+      "step": 7495
+    },
+    {
+      "epoch": 0.8216026729473627,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8762460879106925e-06,
+      "loss": 1.1745,
+      "step": 7500
+    },
+    {
+      "epoch": 0.8221504080626609,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8651107807248091e-06,
+      "loss": 1.1553,
+      "step": 7505
+    },
+    {
+      "epoch": 0.8226981431779591,
+      "grad_norm": 1.25,
+      "learning_rate": 1.8540052159470912e-06,
+      "loss": 1.1032,
+      "step": 7510
+    },
+    {
+      "epoch": 0.8232458782932573,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8429294341811933e-06,
+      "loss": 1.1256,
+      "step": 7515
+    },
+    {
+      "epoch": 0.8237936134085556,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8318834759218817e-06,
+      "loss": 1.1521,
+      "step": 7520
+    },
+    {
+      "epoch": 0.8243413485238539,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8208673815548794e-06,
+      "loss": 1.1444,
+      "step": 7525
+    },
+    {
+      "epoch": 0.8248890836391521,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8098811913567248e-06,
+      "loss": 1.0999,
+      "step": 7530
+    },
+    {
+      "epoch": 0.8254368187544503,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7989249454946234e-06,
+      "loss": 1.1726,
+      "step": 7535
+    },
+    {
+      "epoch": 0.8259845538697486,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7879986840262953e-06,
+      "loss": 1.1836,
+      "step": 7540
+    },
+    {
+      "epoch": 0.8265322889850468,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7771024468998377e-06,
+      "loss": 1.1058,
+      "step": 7545
+    },
+    {
+      "epoch": 0.8270800241003451,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.766236273953571e-06,
+      "loss": 1.1559,
+      "step": 7550
+    },
+    {
+      "epoch": 0.8276277592156434,
+      "grad_norm": 1.25,
+      "learning_rate": 1.7554002049158947e-06,
+      "loss": 1.1613,
+      "step": 7555
+    },
+    {
+      "epoch": 0.8281754943309415,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.7445942794051552e-06,
+      "loss": 1.1669,
+      "step": 7560
+    },
+    {
+      "epoch": 0.8287232294462398,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7338185369294725e-06,
+      "loss": 1.1707,
+      "step": 7565
+    },
+    {
+      "epoch": 0.829270964561538,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.7230730168866216e-06,
+      "loss": 1.1434,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8298186996768363,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7123577585638861e-06,
+      "loss": 1.1345,
+      "step": 7575
+    },
+    {
+      "epoch": 0.8303664347921346,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.7016728011378936e-06,
+      "loss": 1.1141,
+      "step": 7580
+    },
+    {
+      "epoch": 0.8309141699074327,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6910181836744955e-06,
+      "loss": 1.1606,
+      "step": 7585
+    },
+    {
+      "epoch": 0.831461905022731,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.68039394512862e-06,
+      "loss": 1.1388,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8320096401380292,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.669800124344112e-06,
+      "loss": 1.1396,
+      "step": 7595
+    },
+    {
+      "epoch": 0.8325573752533275,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.6592367600536209e-06,
+      "loss": 1.1241,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8331051103686258,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.648703890878427e-06,
+      "loss": 1.1524,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8336528454839239,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.6382015553283215e-06,
+      "loss": 1.1413,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8342005805992222,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6277297918014711e-06,
+      "loss": 1.1485,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8347483157145205,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.6172886385842457e-06,
+      "loss": 1.1215,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8352960508298187,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6068781338511131e-06,
+      "loss": 1.1049,
+      "step": 7625
+    },
+    {
+      "epoch": 0.835843785945117,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.5964983156644876e-06,
+      "loss": 1.1723,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8363915210604151,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5861492219745733e-06,
+      "loss": 1.1166,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8369392561757134,
+      "grad_norm": 1.25,
+      "learning_rate": 1.575830890619261e-06,
+      "loss": 1.1591,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8374869912910117,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5655433593239566e-06,
+      "loss": 1.1433,
+      "step": 7645
+    },
+    {
+      "epoch": 0.8380347264063099,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.5552866657014542e-06,
+      "loss": 1.1351,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8385824615216082,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5450608472518115e-06,
+      "loss": 1.1423,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8391301966369064,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5348659413621946e-06,
+      "loss": 1.1316,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8396779317522046,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5247019853067501e-06,
+      "loss": 1.1513,
+      "step": 7665
+    },
+    {
+      "epoch": 0.8402256668675029,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5145690162464666e-06,
+      "loss": 1.1162,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8407734019828011,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5044670712290399e-06,
+      "loss": 1.1247,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8413211370980994,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4943961871887368e-06,
+      "loss": 1.1506,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8418688722133976,
+      "grad_norm": 1.25,
+      "learning_rate": 1.484356400946264e-06,
+      "loss": 1.1769,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8424166073286958,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.47434774920862e-06,
+      "loss": 1.1816,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8429643424439941,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4643702685689832e-06,
+      "loss": 1.1576,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8435120775592924,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.4544239955065576e-06,
+      "loss": 1.109,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8440598126745905,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.444508966386451e-06,
+      "loss": 1.2185,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8446075477898888,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4346252174595354e-06,
+      "loss": 1.1538,
+      "step": 7710
+    },
+    {
+      "epoch": 0.845155282905187,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4247727848623205e-06,
+      "loss": 1.1246,
+      "step": 7715
+    },
+    {
+      "epoch": 0.8457030180204853,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.4149517046168182e-06,
+      "loss": 1.1427,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8462507531357836,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.40516201263041e-06,
+      "loss": 1.1379,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8467984882510817,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.395403744695717e-06,
+      "loss": 1.119,
+      "step": 7730
+    },
+    {
+      "epoch": 0.84734622336638,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.385676936490472e-06,
+      "loss": 1.15,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8478939584816783,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3759816235773838e-06,
+      "loss": 1.1581,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8484416935969765,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.366317841404009e-06,
+      "loss": 1.1532,
+      "step": 7745
+    },
+    {
+      "epoch": 0.8489894287122748,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.356685625302625e-06,
+      "loss": 1.1954,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8495371638275729,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3470850104900967e-06,
+      "loss": 1.1654,
+      "step": 7755
+    },
+    {
+      "epoch": 0.8500848989428712,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.3375160320677517e-06,
+      "loss": 1.2019,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8506326340581695,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3279787250212483e-06,
+      "loss": 1.132,
+      "step": 7765
+    },
+    {
+      "epoch": 0.8511803691734677,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3184731242204508e-06,
+      "loss": 1.2017,
+      "step": 7770
+    },
+    {
+      "epoch": 0.851728104288766,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3089992644193005e-06,
+      "loss": 1.1279,
+      "step": 7775
+    },
+    {
+      "epoch": 0.8522758394040642,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.2995571802556872e-06,
+      "loss": 1.1969,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8528235745193624,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2901469062513262e-06,
+      "loss": 1.1247,
+      "step": 7785
+    },
+    {
+      "epoch": 0.8533713096346607,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.2807684768116292e-06,
+      "loss": 1.1916,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8539190447499589,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2714219262255777e-06,
+      "loss": 1.1319,
+      "step": 7795
+    },
+    {
+      "epoch": 0.8544667798652572,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2621072886656005e-06,
+      "loss": 1.1725,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8550145149805554,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2528245981874488e-06,
+      "loss": 1.124,
+      "step": 7805
+    },
+    {
+      "epoch": 0.8555622500958536,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2435738887300653e-06,
+      "loss": 1.1971,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8561099852111519,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2343551941154763e-06,
+      "loss": 1.1879,
+      "step": 7815
+    },
+    {
+      "epoch": 0.8566577203264502,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.225168548048643e-06,
+      "loss": 1.1339,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8572054554417484,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.2160139841173579e-06,
+      "loss": 1.195,
+      "step": 7825
+    },
+    {
+      "epoch": 0.8577531905570466,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.206891535792124e-06,
+      "loss": 1.1395,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8583009256723448,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1978012364260116e-06,
+      "loss": 1.1273,
+      "step": 7835
+    },
+    {
+      "epoch": 0.8588486607876431,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1887431192545573e-06,
+      "loss": 1.1573,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8593963959029414,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1797172173956328e-06,
+      "loss": 1.1382,
+      "step": 7845
+    },
+    {
+      "epoch": 0.8599441310182396,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.170723563849323e-06,
+      "loss": 1.1812,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8604918661335378,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1617621914978184e-06,
+      "loss": 1.1322,
+      "step": 7855
+    },
+    {
+      "epoch": 0.861039601248836,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1528331331052688e-06,
+      "loss": 1.2111,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8615873363641343,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.1439364213176895e-06,
+      "loss": 1.114,
+      "step": 7865
+    },
+    {
+      "epoch": 0.8621350714794326,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.135072088662833e-06,
+      "loss": 1.1716,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8626828065947307,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.126240167550059e-06,
+      "loss": 1.1152,
+      "step": 7875
+    },
+    {
+      "epoch": 0.863230541710029,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1174406902702362e-06,
+      "loss": 1.1209,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8637782768253273,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1086736889956107e-06,
+      "loss": 1.1278,
+      "step": 7885
+    },
+    {
+      "epoch": 0.8643260119406255,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0999391957796824e-06,
+      "loss": 1.1796,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8648737470559238,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.0912372425571095e-06,
+      "loss": 1.1365,
+      "step": 7895
+    },
+    {
+      "epoch": 0.865421482171222,
+      "grad_norm": 1.375,
+      "learning_rate": 1.0825678611435753e-06,
+      "loss": 1.1597,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8659692172865202,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0739310832356664e-06,
+      "loss": 1.1346,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8665169524018185,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0653269404107802e-06,
+      "loss": 1.124,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8670646875171167,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.056755464126985e-06,
+      "loss": 1.1831,
+      "step": 7915
+    },
+    {
+      "epoch": 0.867612422632415,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.0482166857229204e-06,
+      "loss": 1.1719,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8681601577477132,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0397106364176768e-06,
+      "loss": 1.1345,
+      "step": 7925
+    },
+    {
+      "epoch": 0.8687078928630114,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.0312373473106741e-06,
+      "loss": 1.1847,
+      "step": 7930
+    },
+    {
+      "epoch": 0.8692556279783097,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.0227968493815698e-06,
+      "loss": 1.1735,
+      "step": 7935
+    },
+    {
+      "epoch": 0.869803363093608,
+      "grad_norm": 1.25,
+      "learning_rate": 1.0143891734901235e-06,
+      "loss": 1.1974,
+      "step": 7940
+    },
+    {
+      "epoch": 0.8703510982089062,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0060143503760877e-06,
+      "loss": 1.1209,
+      "step": 7945
+    },
+    {
+      "epoch": 0.8708988333242044,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.976724106591128e-07,
+      "loss": 1.1643,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8714465684395026,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.89363384838613e-07,
+      "loss": 1.1561,
+      "step": 7955
+    },
+    {
+      "epoch": 0.8719943035548009,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.810873032936653e-07,
+      "loss": 1.154,
+      "step": 7960
+    },
+    {
+      "epoch": 0.8725420386700992,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.728441962829006e-07,
+      "loss": 1.1731,
+      "step": 7965
+    },
+    {
+      "epoch": 0.8730897737853974,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.646340939443877e-07,
+      "loss": 1.1772,
+      "step": 7970
+    },
+    {
+      "epoch": 0.8736375089006956,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.564570262955252e-07,
+      "loss": 1.1721,
+      "step": 7975
+    },
+    {
+      "epoch": 0.8741852440159938,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.483130232329307e-07,
+      "loss": 1.1366,
+      "step": 7980
+    },
+    {
+      "epoch": 0.8747329791312921,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.40202114532337e-07,
+      "loss": 1.1504,
+      "step": 7985
+    },
+    {
+      "epoch": 0.8752807142465904,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.321243298484728e-07,
+      "loss": 1.1317,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8758284493618886,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.240796987149658e-07,
+      "loss": 1.1397,
+      "step": 7995
+    },
+    {
+      "epoch": 0.8763761844771868,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.160682505442242e-07,
+      "loss": 1.15,
+      "step": 8000
+    },
+    {
+      "epoch": 0.8769239195924851,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.080900146273386e-07,
+      "loss": 1.1657,
+      "step": 8005
+    },
+    {
+      "epoch": 0.8774716547077833,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.001450201339679e-07,
+      "loss": 1.1981,
+      "step": 8010
+    },
+    {
+      "epoch": 0.8780193898230816,
+      "grad_norm": 1.25,
+      "learning_rate": 8.92233296112236e-07,
+      "loss": 1.1738,
+      "step": 8015
+    },
+    {
+      "epoch": 0.8785671249383799,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.843548714886252e-07,
+      "loss": 1.0631,
+      "step": 8020
+    },
+    {
+      "epoch": 0.879114860053678,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.765097750678675e-07,
+      "loss": 1.1533,
+      "step": 8025
+    },
+    {
+      "epoch": 0.8796625951689763,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.686980355328467e-07,
+      "loss": 1.1551,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8802103302842745,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.609196814444843e-07,
+      "loss": 1.1485,
+      "step": 8035
+    },
+    {
+      "epoch": 0.8807580653995728,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.531747412416424e-07,
+      "loss": 1.14,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8813058005148711,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.454632432410137e-07,
+      "loss": 1.1257,
+      "step": 8045
+    },
+    {
+      "epoch": 0.8818535356301692,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.377852156370236e-07,
+      "loss": 1.1692,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8824012707454675,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.301406865017247e-07,
+      "loss": 1.1838,
+      "step": 8055
+    },
+    {
+      "epoch": 0.8829490058607657,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.225296837846919e-07,
+      "loss": 1.1881,
+      "step": 8060
+    },
+    {
+      "epoch": 0.883496740976064,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.149522353129224e-07,
+      "loss": 1.1945,
+      "step": 8065
+    },
+    {
+      "epoch": 0.8840444760913622,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.074083687907408e-07,
+      "loss": 1.1544,
+      "step": 8070
+    },
+    {
+      "epoch": 0.8845922112066604,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.998981117996796e-07,
+      "loss": 1.1729,
+      "step": 8075
+    },
+    {
+      "epoch": 0.8851399463219587,
+      "grad_norm": 1.3203125,
+      "learning_rate": 7.924214917983952e-07,
+      "loss": 1.1486,
+      "step": 8080
+    },
+    {
+      "epoch": 0.885687681437257,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.84978536122567e-07,
+      "loss": 1.2001,
+      "step": 8085
+    },
+    {
+      "epoch": 0.8862354165525552,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.775692719847816e-07,
+      "loss": 1.1905,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8867831516678534,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.701937264744564e-07,
+      "loss": 1.1848,
+      "step": 8095
+    },
+    {
+      "epoch": 0.8873308867831516,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.628519265577162e-07,
+      "loss": 1.1559,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8878786218984499,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.555438990773134e-07,
+      "loss": 1.1799,
+      "step": 8105
+    },
+    {
+      "epoch": 0.8884263570137482,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.482696707525272e-07,
+      "loss": 1.2088,
+      "step": 8110
+    },
+    {
+      "epoch": 0.8889740921290464,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.41029268179052e-07,
+      "loss": 1.1298,
+      "step": 8115
+    },
+    {
+      "epoch": 0.8895218272443446,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.338227178289148e-07,
+      "loss": 1.1782,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8900695623596429,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.266500460503778e-07,
+      "loss": 1.2125,
+      "step": 8125
+    },
+    {
+      "epoch": 0.8906172974749411,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.195112790678293e-07,
+      "loss": 1.1491,
+      "step": 8130
+    },
+    {
+      "epoch": 0.8911650325902394,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.124064429817057e-07,
+      "loss": 1.1438,
+      "step": 8135
+    },
+    {
+      "epoch": 0.8917127677055376,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.053355637683801e-07,
+      "loss": 1.1473,
+      "step": 8140
+    },
+    {
+      "epoch": 0.8922605028208358,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.98298667280074e-07,
+      "loss": 1.2112,
+      "step": 8145
+    },
+    {
+      "epoch": 0.8928082379361341,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.912957792447683e-07,
+      "loss": 1.1886,
+      "step": 8150
+    },
+    {
+      "epoch": 0.8933559730514323,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.843269252661011e-07,
+      "loss": 1.1485,
+      "step": 8155
+    },
+    {
+      "epoch": 0.8939037081667306,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.773921308232701e-07,
+      "loss": 1.1473,
+      "step": 8160
+    },
+    {
+      "epoch": 0.8944514432820289,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.70491421270959e-07,
+      "loss": 1.1345,
+      "step": 8165
+    },
+    {
+      "epoch": 0.894999178397327,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.636248218392216e-07,
+      "loss": 1.1708,
+      "step": 8170
+    },
+    {
+      "epoch": 0.8955469135126253,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.567923576334034e-07,
+      "loss": 1.1519,
+      "step": 8175
+    },
+    {
+      "epoch": 0.8960946486279235,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.499940536340488e-07,
+      "loss": 1.1515,
+      "step": 8180
+    },
+    {
+      "epoch": 0.8966423837432218,
+      "grad_norm": 1.3125,
+      "learning_rate": 6.432299346967996e-07,
+      "loss": 1.1465,
+      "step": 8185
+    },
+    {
+      "epoch": 0.8971901188585201,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.365000255523202e-07,
+      "loss": 1.1477,
+      "step": 8190
+    },
+    {
+      "epoch": 0.8977378539738182,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.298043508061946e-07,
+      "loss": 1.163,
+      "step": 8195
+    },
+    {
+      "epoch": 0.8982855890891165,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.231429349388396e-07,
+      "loss": 1.1559,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8988333242044148,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.165158023054196e-07,
+      "loss": 1.1445,
+      "step": 8205
+    },
+    {
+      "epoch": 0.899381059319713,
+      "grad_norm": 1.25,
+      "learning_rate": 6.099229771357517e-07,
+      "loss": 1.1405,
+      "step": 8210
+    },
+    {
+      "epoch": 0.8999287944350113,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.0336448353422e-07,
+      "loss": 1.1287,
+      "step": 8215
+    },
+    {
+      "epoch": 0.9004765295503094,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.968403454796889e-07,
+      "loss": 1.1489,
+      "step": 8220
+    },
+    {
+      "epoch": 0.9010242646656077,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.9035058682541e-07,
+      "loss": 1.1511,
+      "step": 8225
+    },
+    {
+      "epoch": 0.901571999780906,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.838952312989432e-07,
+      "loss": 1.1326,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9021197348962042,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.774743025020602e-07,
+      "loss": 1.1529,
+      "step": 8235
+    },
+    {
+      "epoch": 0.9026674700115024,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.710878239106677e-07,
+      "loss": 1.1237,
+      "step": 8240
+    },
+    {
+      "epoch": 0.9032152051268006,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.647358188747143e-07,
+      "loss": 1.1941,
+      "step": 8245
+    },
+    {
+      "epoch": 0.9037629402420989,
+      "grad_norm": 1.28125,
+      "learning_rate": 5.584183106181085e-07,
+      "loss": 1.1752,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9043106753573972,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.521353222386361e-07,
+      "loss": 1.1237,
+      "step": 8255
+    },
+    {
+      "epoch": 0.9048584104726954,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.458868767078673e-07,
+      "loss": 1.0955,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9054061455879936,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.396729968710835e-07,
+      "loss": 1.1359,
+      "step": 8265
+    },
+    {
+      "epoch": 0.9059538807032919,
+      "grad_norm": 1.25,
+      "learning_rate": 5.334937054471867e-07,
+      "loss": 1.1522,
+      "step": 8270
+    },
+    {
+      "epoch": 0.9065016158185901,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.273490250286173e-07,
+      "loss": 1.1596,
+      "step": 8275
+    },
+    {
+      "epoch": 0.9070493509338884,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.212389780812733e-07,
+      "loss": 1.156,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9075970860491867,
+      "grad_norm": 1.2734375,
+      "learning_rate": 5.151635869444293e-07,
+      "loss": 1.1344,
+      "step": 8285
+    },
+    {
+      "epoch": 0.9081448211644848,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.091228738306497e-07,
+      "loss": 1.1873,
+      "step": 8290
+    },
+    {
+      "epoch": 0.9086925562797831,
+      "grad_norm": 1.375,
+      "learning_rate": 5.03116860825712e-07,
+      "loss": 1.1077,
+      "step": 8295
+    },
+    {
+      "epoch": 0.9092402913950813,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.971455698885263e-07,
+      "loss": 1.1709,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9097880265103796,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.912090228510502e-07,
+      "loss": 1.1968,
+      "step": 8305
+    },
+    {
+      "epoch": 0.9103357616256779,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.85307241418218e-07,
+      "loss": 1.1573,
+      "step": 8310
+    },
+    {
+      "epoch": 0.910883496740976,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.794402471678483e-07,
+      "loss": 1.2055,
+      "step": 8315
+    },
+    {
+      "epoch": 0.9114312318562743,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.7360806155057557e-07,
+      "loss": 1.1618,
+      "step": 8320
+    },
+    {
+      "epoch": 0.9119789669715725,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.6781070588977187e-07,
+      "loss": 1.1254,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9125267020868708,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.620482013814609e-07,
+      "loss": 1.121,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9130744372021691,
+      "grad_norm": 1.25,
+      "learning_rate": 4.5632056909424517e-07,
+      "loss": 1.1182,
+      "step": 8335
+    },
+    {
+      "epoch": 0.9136221723174672,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.506278299692335e-07,
+      "loss": 1.167,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9141699074327655,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.449700048199546e-07,
+      "loss": 1.1577,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9147176425480638,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.393471143322925e-07,
+      "loss": 1.1233,
+      "step": 8350
+    },
+    {
+      "epoch": 0.915265377663362,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.337591790643969e-07,
+      "loss": 1.1454,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9158131127786603,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.2820621944662077e-07,
+      "loss": 1.1687,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9163608478939584,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.226882557814438e-07,
+      "loss": 1.1294,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9169085830092567,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.172053082433858e-07,
+      "loss": 1.1536,
+      "step": 8370
+    },
+    {
+      "epoch": 0.917456318124555,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.117573968789501e-07,
+      "loss": 1.1602,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9180040532398532,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.063445416065415e-07,
+      "loss": 1.1011,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9185517883551515,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.009667622163882e-07,
+      "loss": 1.2093,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9190995234704497,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.9562407837048566e-07,
+      "loss": 1.1483,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9196472585857479,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.9031650960250635e-07,
+      "loss": 1.1597,
+      "step": 8395
+    },
+    {
+      "epoch": 0.9201949937010462,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.850440753177376e-07,
+      "loss": 1.089,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9207427288163444,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.79806794793014e-07,
+      "loss": 1.1409,
+      "step": 8405
+    },
+    {
+      "epoch": 0.9212904639316427,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.7460468717663955e-07,
+      "loss": 1.1505,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9218381990469409,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.6943777148831907e-07,
+      "loss": 1.1877,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9223859341622391,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.6430606661909673e-07,
+      "loss": 1.1677,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9229336692775374,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.5920959133126987e-07,
+      "loss": 1.1571,
+      "step": 8425
+    },
+    {
+      "epoch": 0.9234814043928357,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.541483642583421e-07,
+      "loss": 1.1979,
+      "step": 8430
+    },
+    {
+      "epoch": 0.9240291395081338,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.491224039049379e-07,
+      "loss": 1.1641,
+      "step": 8435
+    },
+    {
+      "epoch": 0.9245768746234321,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.441317286467416e-07,
+      "loss": 1.1128,
+      "step": 8440
+    },
+    {
+      "epoch": 0.9251246097387303,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.3917635673043183e-07,
+      "loss": 1.0991,
+      "step": 8445
+    },
+    {
+      "epoch": 0.9256723448540286,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.3425630627361263e-07,
+      "loss": 1.2116,
+      "step": 8450
+    },
+    {
+      "epoch": 0.9262200799693269,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.293715952647425e-07,
+      "loss": 1.1897,
+      "step": 8455
+    },
+    {
+      "epoch": 0.926767815084625,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.245222415630822e-07,
+      "loss": 1.1805,
+      "step": 8460
+    },
+    {
+      "epoch": 0.9273155501999233,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.197082628986126e-07,
+      "loss": 1.1074,
+      "step": 8465
+    },
+    {
+      "epoch": 0.9278632853152216,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.149296768719834e-07,
+      "loss": 1.0734,
+      "step": 8470
+    },
+    {
+      "epoch": 0.9284110204305198,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.101865009544391e-07,
+      "loss": 1.1775,
+      "step": 8475
+    },
+    {
+      "epoch": 0.9289587555458181,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.054787524877645e-07,
+      "loss": 1.1536,
+      "step": 8480
+    },
+    {
+      "epoch": 0.9295064906611162,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.0080644868420996e-07,
+      "loss": 1.1351,
+      "step": 8485
+    },
+    {
+      "epoch": 0.9300542257764145,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.9616960662643967e-07,
+      "loss": 1.222,
+      "step": 8490
+    },
+    {
+      "epoch": 0.9306019608917128,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.915682432674627e-07,
+      "loss": 1.1448,
+      "step": 8495
+    },
+    {
+      "epoch": 0.931149696007011,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.8700237543057173e-07,
+      "loss": 1.1303,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9316974311223093,
+      "grad_norm": 1.25,
+      "learning_rate": 2.824720198092834e-07,
+      "loss": 1.1381,
+      "step": 8505
+    },
+    {
+      "epoch": 0.9322451662376074,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.7797719296727476e-07,
+      "loss": 1.1853,
+      "step": 8510
+    },
+    {
+      "epoch": 0.9327929013529057,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.7351791133832685e-07,
+      "loss": 1.1622,
+      "step": 8515
+    },
+    {
+      "epoch": 0.933340636468204,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.69094191226259e-07,
+      "loss": 1.1323,
+      "step": 8520
+    },
+    {
+      "epoch": 0.9338883715835022,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.647060488048736e-07,
+      "loss": 1.1172,
+      "step": 8525
+    },
+    {
+      "epoch": 0.9344361066988005,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.603535001178947e-07,
+      "loss": 1.1416,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9349838418140987,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.5603656107891285e-07,
+      "loss": 1.224,
+      "step": 8535
+    },
+    {
+      "epoch": 0.9355315769293969,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.517552474713203e-07,
+      "loss": 1.1628,
+      "step": 8540
+    },
+    {
+      "epoch": 0.9360793120446952,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.4750957494826033e-07,
+      "loss": 1.1395,
+      "step": 8545
+    },
+    {
+      "epoch": 0.9366270471599935,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.4329955903256376e-07,
+      "loss": 1.169,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9371747822752917,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.391252151167001e-07,
+      "loss": 1.1216,
+      "step": 8555
+    },
+    {
+      "epoch": 0.9377225173905899,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.349865584627109e-07,
+      "loss": 1.1221,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9382702525058881,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.308836042021656e-07,
+      "loss": 1.1322,
+      "step": 8565
+    },
+    {
+      "epoch": 0.9388179876211864,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.2681636733609457e-07,
+      "loss": 1.1842,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9393657227364847,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.2278486273494272e-07,
+      "loss": 1.1955,
+      "step": 8575
+    },
+    {
+      "epoch": 0.9399134578517829,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.1878910513851381e-07,
+      "loss": 1.2065,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9404611929670811,
+      "grad_norm": 1.25,
+      "learning_rate": 2.148291091559107e-07,
+      "loss": 1.141,
+      "step": 8585
+    },
+    {
+      "epoch": 0.9410089280823793,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.1090488926548968e-07,
+      "loss": 1.145,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9415566631976776,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.070164598148039e-07,
+      "loss": 1.1382,
+      "step": 8595
+    },
+    {
+      "epoch": 0.9421043983129759,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.0316383502054782e-07,
+      "loss": 1.1571,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9426521334282741,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.993470289685162e-07,
+      "loss": 1.1382,
+      "step": 8605
+    },
+    {
+      "epoch": 0.9431998685435723,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.9556605561353525e-07,
+      "loss": 1.146,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9437476036588706,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9182092877942705e-07,
+      "loss": 1.0946,
+      "step": 8615
+    },
+    {
+      "epoch": 0.9442953387741688,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8811166215895405e-07,
+      "loss": 1.0913,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9448430738894671,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8443826931376474e-07,
+      "loss": 1.0888,
+      "step": 8625
+    },
+    {
+      "epoch": 0.9453908090047652,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8080076367434918e-07,
+      "loss": 1.1671,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9459385441200635,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7719915853999014e-07,
+      "loss": 1.1341,
+      "step": 8635
+    },
+    {
+      "epoch": 0.9464862792353618,
+      "grad_norm": 1.25,
+      "learning_rate": 1.7363346707870877e-07,
+      "loss": 1.1377,
+      "step": 8640
+    },
+    {
+      "epoch": 0.94703401435066,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.701037023272234e-07,
+      "loss": 1.1857,
+      "step": 8645
+    },
+    {
+      "epoch": 0.9475817494659583,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.666098771908986e-07,
+      "loss": 1.1515,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9481294845812565,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.6315200444369406e-07,
+      "loss": 1.1915,
+      "step": 8655
+    },
+    {
+      "epoch": 0.9486772196965547,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.59730096728129e-07,
+      "loss": 1.1233,
+      "step": 8660
+    },
+    {
+      "epoch": 0.949224954811853,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5634416655522343e-07,
+      "loss": 1.1587,
+      "step": 8665
+    },
+    {
+      "epoch": 0.9497726899271512,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5299422630445816e-07,
+      "loss": 1.1867,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9503204250424495,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4968028822373471e-07,
+      "loss": 1.2192,
+      "step": 8675
+    },
+    {
+      "epoch": 0.9508681601577477,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.4640236442931665e-07,
+      "loss": 1.2009,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9514158952730459,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4316046690580178e-07,
+      "loss": 1.1951,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9519636303883442,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.399546075060665e-07,
+      "loss": 1.1862,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9525113655036425,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.367847979512238e-07,
+      "loss": 1.1432,
+      "step": 8695
+    },
+    {
+      "epoch": 0.9530591006189407,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3365104983058873e-07,
+      "loss": 1.1178,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9536068357342389,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.3055337460162632e-07,
+      "loss": 1.1738,
+      "step": 8705
+    },
+    {
+      "epoch": 0.9541545708495371,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2749178358991477e-07,
+      "loss": 1.16,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9547023059648354,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.244662879891012e-07,
+      "loss": 1.1101,
+      "step": 8715
+    },
+    {
+      "epoch": 0.9552500410801337,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.21476898860865e-07,
+      "loss": 1.2003,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9557977761954319,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.185236271348722e-07,
+      "loss": 1.2025,
+      "step": 8725
+    },
+    {
+      "epoch": 0.9563455113107301,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1560648360874005e-07,
+      "loss": 1.1886,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9568932464260284,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1272547894799369e-07,
+      "loss": 1.1359,
+      "step": 8735
+    },
+    {
+      "epoch": 0.9574409815413266,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.0988062368603059e-07,
+      "loss": 1.2033,
+      "step": 8740
+    },
+    {
+      "epoch": 0.9579887166566249,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0707192822408063e-07,
+      "loss": 1.2268,
+      "step": 8745
+    },
+    {
+      "epoch": 0.9585364517719231,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0429940283116613e-07,
+      "loss": 1.1409,
+      "step": 8750
+    },
+    {
+      "epoch": 0.9590841868872213,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0156305764406627e-07,
+      "loss": 1.1499,
+      "step": 8755
+    },
+    {
+      "epoch": 0.9596319220025196,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.886290266728271e-08,
+      "loss": 1.1508,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9601796571178178,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.619894777299632e-08,
+      "loss": 1.168,
+      "step": 8765
+    },
+    {
+      "epoch": 0.9607273922331161,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.357120270103715e-08,
+      "loss": 1.1012,
+      "step": 8770
+    },
+    {
+      "epoch": 0.9612751273484144,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.097967705884558e-08,
+      "loss": 1.1533,
+      "step": 8775
+    },
+    {
+      "epoch": 0.9618228624637125,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.842438032143908e-08,
+      "loss": 1.1602,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9623705975790108,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.590532183137656e-08,
+      "loss": 1.1479,
+      "step": 8785
+    },
+    {
+      "epoch": 0.962918332694309,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.3422510798723e-08,
+      "loss": 1.1496,
+      "step": 8790
+    },
+    {
+      "epoch": 0.9634660678096073,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.097595630101818e-08,
+      "loss": 1.146,
+      "step": 8795
+    },
+    {
+      "epoch": 0.9640138029249055,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.856566728324244e-08,
+      "loss": 1.0557,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9645615380402037,
+      "grad_norm": 1.25,
+      "learning_rate": 7.619165255778327e-08,
+      "loss": 1.1975,
+      "step": 8805
+    },
+    {
+      "epoch": 0.965109273155502,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.385392080440535e-08,
+      "loss": 1.1624,
+      "step": 8810
+    },
+    {
+      "epoch": 0.9656570082708003,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.155248057021502e-08,
+      "loss": 1.1225,
+      "step": 8815
+    },
+    {
+      "epoch": 0.9662047433860985,
+      "grad_norm": 1.3046875,
+      "learning_rate": 6.928734026963258e-08,
+      "loss": 1.1341,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9667524785013967,
+      "grad_norm": 1.25,
+      "learning_rate": 6.705850818436111e-08,
+      "loss": 1.1975,
+      "step": 8825
+    },
+    {
+      "epoch": 0.9673002136166949,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.486599246335212e-08,
+      "loss": 1.1769,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9678479487319932,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.270980112278113e-08,
+      "loss": 1.2048,
+      "step": 8835
+    },
+    {
+      "epoch": 0.9683956838472915,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.058994204601765e-08,
+      "loss": 1.0985,
+      "step": 8840
+    },
+    {
+      "epoch": 0.9689434189625897,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.850642298359188e-08,
+      "loss": 1.173,
+      "step": 8845
+    },
+    {
+      "epoch": 0.9694911540778879,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.6459251553169226e-08,
+      "loss": 1.104,
+      "step": 8850
+    },
+    {
+      "epoch": 0.9700388891931861,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.444843523952581e-08,
+      "loss": 1.158,
+      "step": 8855
+    },
+    {
+      "epoch": 0.9705866243084844,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.2473981394515205e-08,
+      "loss": 1.1602,
+      "step": 8860
+    },
+    {
+      "epoch": 0.9711343594237827,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.0535897237044e-08,
+      "loss": 1.0855,
+      "step": 8865
+    },
+    {
+      "epoch": 0.9716820945390809,
+      "grad_norm": 1.328125,
+      "learning_rate": 4.863418985304735e-08,
+      "loss": 1.157,
+      "step": 8870
+    },
+    {
+      "epoch": 0.9722298296543791,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.6768866195460175e-08,
+      "loss": 1.2024,
+      "step": 8875
+    },
+    {
+      "epoch": 0.9727775647696774,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.4939933084192646e-08,
+      "loss": 1.1865,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9733252998849756,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.3147397206106945e-08,
+      "loss": 1.1473,
+      "step": 8885
+    },
+    {
+      "epoch": 0.9738730350002739,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.1391265114990584e-08,
+      "loss": 1.1184,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9744207701155722,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.967154323153089e-08,
+      "loss": 1.1739,
+      "step": 8895
+    },
+    {
+      "epoch": 0.9749685052308703,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.79882378432983e-08,
+      "loss": 1.1479,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9755162403461686,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.634135510471537e-08,
+      "loss": 1.0821,
+      "step": 8905
+    },
+    {
+      "epoch": 0.9760639754614668,
+      "grad_norm": 1.25,
+      "learning_rate": 3.473090103703891e-08,
+      "loss": 1.1344,
+      "step": 8910
+    },
+    {
+      "epoch": 0.9766117105767651,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.315688152833896e-08,
+      "loss": 1.1173,
+      "step": 8915
+    },
+    {
+      "epoch": 0.9771594456920634,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.161930233347099e-08,
+      "loss": 1.1412,
+      "step": 8920
+    },
+    {
+      "epoch": 0.9777071808073615,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.0118169074061507e-08,
+      "loss": 1.1319,
+      "step": 8925
+    },
+    {
+      "epoch": 0.9782549159226598,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.8653487238488044e-08,
+      "loss": 1.1994,
+      "step": 8930
+    },
+    {
+      "epoch": 0.978802651037958,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.7225262181849177e-08,
+      "loss": 1.1638,
+      "step": 8935
+    },
+    {
+      "epoch": 0.9793503861532563,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.5833499125957896e-08,
+      "loss": 1.1377,
+      "step": 8940
+    },
+    {
+      "epoch": 0.9798981212685546,
+      "grad_norm": 1.2890625,
+      "learning_rate": 2.447820315931382e-08,
+      "loss": 1.125,
+      "step": 8945
+    },
+    {
+      "epoch": 0.9804458563838527,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.3159379237087666e-08,
+      "loss": 1.1463,
+      "step": 8950
+    },
+    {
+      "epoch": 0.980993591499151,
+      "grad_norm": 1.2890625,
+      "learning_rate": 2.1877032181102376e-08,
+      "loss": 1.1219,
+      "step": 8955
+    },
+    {
+      "epoch": 0.9815413266144493,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.063116667981757e-08,
+      "loss": 1.1693,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9820890617297475,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.942178728830957e-08,
+      "loss": 1.1101,
+      "step": 8965
+    },
+    {
+      "epoch": 0.9826367968450458,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8248898428253615e-08,
+      "loss": 1.1785,
+      "step": 8970
+    },
+    {
+      "epoch": 0.9831845319603439,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.71125043879139e-08,
+      "loss": 1.1803,
+      "step": 8975
+    },
+    {
+      "epoch": 0.9837322670756422,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.601260932212245e-08,
+      "loss": 1.172,
+      "step": 8980
+    },
+    {
+      "epoch": 0.9842800021909405,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4949217252262505e-08,
+      "loss": 1.1578,
+      "step": 8985
+    },
+    {
+      "epoch": 0.9848277373062387,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3922332066262923e-08,
+      "loss": 1.1933,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9853754724215369,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.2931957518570459e-08,
+      "loss": 1.2329,
+      "step": 8995
+    },
+    {
+      "epoch": 0.9859232075368352,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1978097230149755e-08,
+      "loss": 1.2313,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9864709426521334,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1060754688460018e-08,
+      "loss": 1.1615,
+      "step": 9005
+    },
+    {
+      "epoch": 0.9870186777674317,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.017993324744615e-08,
+      "loss": 1.1141,
+      "step": 9010
+    },
+    {
+      "epoch": 0.98756641288273,
+      "grad_norm": 1.25,
+      "learning_rate": 9.335636127528746e-09,
+      "loss": 1.182,
+      "step": 9015
+    },
+    {
+      "epoch": 0.9881141479980281,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.527866415586338e-09,
+      "loss": 1.129,
+      "step": 9020
+    },
+    {
+      "epoch": 0.9886618831133264,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.75662706495095e-09,
+      "loss": 1.1161,
+      "step": 9025
+    },
+    {
+      "epoch": 0.9892096182286246,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.021920895391443e-09,
+      "loss": 1.1489,
+      "step": 9030
+    },
+    {
+      "epoch": 0.9897573533439229,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.323750593106859e-09,
+      "loss": 1.1763,
+      "step": 9035
+    },
+    {
+      "epoch": 0.9903050884592212,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.6621187107153145e-09,
+      "loss": 1.1108,
+      "step": 9040
+    },
+    {
+      "epoch": 0.9908528235745193,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.037027667246231e-09,
+      "loss": 1.1569,
+      "step": 9045
+    },
+    {
+      "epoch": 0.9914005586898176,
+      "grad_norm": 1.25,
+      "learning_rate": 4.44847974812701e-09,
+      "loss": 1.1746,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9919482938051158,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.896477105179708e-09,
+      "loss": 1.1355,
+      "step": 9055
+    },
+    {
+      "epoch": 0.9924960289204141,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.381021756612146e-09,
+      "loss": 1.1379,
+      "step": 9060
+    },
+    {
+      "epoch": 0.9930437640357124,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.9021155870079255e-09,
+      "loss": 1.1436,
+      "step": 9065
+    },
+    {
+      "epoch": 0.9935914991510105,
+      "grad_norm": 1.25,
+      "learning_rate": 2.459760347320872e-09,
+      "loss": 1.181,
+      "step": 9070
+    },
+    {
+      "epoch": 0.9941392342663088,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.053957654871708e-09,
+      "loss": 1.1355,
+      "step": 9075
+    },
+    {
+      "epoch": 0.9946869693816071,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.6847089933358373e-09,
+      "loss": 1.1293,
+      "step": 9080
+    },
+    {
+      "epoch": 0.9952347044969053,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3520157127444589e-09,
+      "loss": 1.1847,
+      "step": 9085
+    },
+    {
+      "epoch": 0.9957824396122036,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0558790294745713e-09,
+      "loss": 1.1692,
+      "step": 9090
+    },
+    {
+      "epoch": 0.9963301747275017,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.96300026248975e-10,
+      "loss": 1.1458,
+      "step": 9095
+    },
+    {
+      "epoch": 0.9968779098428,
+      "grad_norm": 1.25,
+      "learning_rate": 5.732796521262796e-10,
+      "loss": 1.1276,
+      "step": 9100
+    },
+    {
+      "epoch": 0.9974256449580983,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.8681872250534437e-10,
+      "loss": 1.1752,
+      "step": 9105
+    },
+    {
+      "epoch": 0.9979733800733965,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.369179191141768e-10,
+      "loss": 1.1254,
+      "step": 9110
+    },
+    {
+      "epoch": 0.9985211151886948,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.235777900154833e-10,
+      "loss": 1.1485,
+      "step": 9115
+    },
+    {
+      "epoch": 0.999068850303993,
+      "grad_norm": 1.25,
+      "learning_rate": 4.679874959556685e-11,
+      "loss": 1.1661,
+      "step": 9120
+    },
+    {
+      "epoch": 0.9996165854192912,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.5810785732089985e-12,
+      "loss": 1.1323,
+      "step": 9125
+    },
+    {
+      "epoch": 0.9999452264884702,
+      "eval_loss": 1.1566358804702759,
+      "eval_runtime": 1068.2466,
+      "eval_samples_per_second": 15.133,
+      "eval_steps_per_second": 1.892,
+      "step": 9128
+    },
+    {
+      "epoch": 0.9999452264884702,
+      "step": 9128,
+      "total_flos": 1.8565750537228124e+18,
+      "train_loss": 1.1714502769954156,
+      "train_runtime": 42415.0484,
+      "train_samples_per_second": 3.443,
+      "train_steps_per_second": 0.215
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 9128,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.8565750537228124e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}